Flying Blind: The Politics of the U.S. Strategic Bomber Program 9781501733567

Citation preview

Flying Blind

?

Cornell Studies edited by

Strategic Nuclear Targeting, edited

by Desmond

Ball

in Security Affairs

Robert and

J.

Art and Robert Jervis

Jeffrey Richelson

japan Prepares for Total War: The Search for Economic Security 1919-1941, by Michael A.

Barnhart The German Nuclear Dilemma, by Jeffrey Boutwell Flying Blind: The Politics of the U.S. Strategic Bomber Program, by Michael E. Brown Citizens and Soldiers: The Dilemmas of Military Service, by Eliot A. Cohen Great Power Politics and the Struggle over Austria, 1945-1955, by Audrey Kurth Cronin Military Organizations, Complex Machines: Modernization in the U.S. Armed Services, by Chris C. Demchak Nuclear Arguments: Understanding the Strategic Nuclear Arms and Arms Control Debate, edited by Lynn Eden and Steven E. Miller Innovation and the Arms Race: How the United States and the Soviet Union Develop New Military Technologies, by Matthew Evangelista Men, Money, and Diplomacy: The Evolution of British Strategic Foreign Policy, 1919-1926, by John Robert Ferris A Substitute for Victory: The Politics of Peacekeeping at the Korean Armistice Talks, by Rosemary Foot The Wrong War: American Policy and the Dimensions of the Korean Conflict, 1950-1953, by

Rosemary Foot The Soviet Union and the Politics of Nuclear Weapons in Europe, 1969 - 1987 by Jonathan Haslam The Soviet Union and the Failure of Collective Security, 1934-1938, by Jiri Hochman The Warsaw Pact: Alliance in Transition edited by David Holloway and Jane M. O. Sharp The Illogic of American Nuclear Strategy, by Robert Jervis The Meaning of the Nuclear Revolution, by Robert Jervis Nuclear Crisis Management: A Dangerous Illusion, by Richard Ned Lebow The Search for Security in Space, edited by Kenneth N. Luongo and W. Thomas Wander The Nuclear Future, by Michael Mandelbaum Conventional Deterrence, by John J. Mearsheimer Liddell Hart and the Weight of History, by John J. Mearsheimer Inadvertent Escalation: Conventional War and Nuclear Risks, by Barry R. Posen The Sources of Military Doctrine: France, Britain, and Germany between the World Wars, by Barry ,

Posen Winning the Next War: Innovation and the Modern Military, by Stephen Peter Rosen Israel and Conventional Deterrence: Border Warfare from 1953 1 97°> by Jonathan Shimshoni Fighting to a Finish: The Politics of War Termination in the United States and japan, 1945, by Leon R.

V. Sigal

The Ideology of the Offensive: Military Decision Making and the Disasters of 1914, by Jack Snyder Myths of Empire: Domestic Politics and International Ambition, by Jack Snyder The Militarization of Space: U.S. Policy, 1945-1984, by Paul B. Stares Making the Alliance Work: The United States and Western Europe, by Gregory E. Treverton The Origins of Alliances, by Stephen M. Walt The Ultimate Enemy: British Intelligence and Nazi Germany, 1933-1939, by Wesley K. Wark The Tet Offensive: Intelligence Failure in War, by James J. Wirtz

Flying Blind THE POLITICS OF THE U.S. STRATEGIC BOMBER

PROGRAM Michael

E.

Brown

Cornell University Press Ithaca and London

—

Copyright

©

1992 by Cornell University

All rights reserved. Except for brief quotations in a review, this book,

or parts thereof,

must not be reproduced

in

any form without permis-

sion in writing from the publisher. For information, address Cornell

University Press, 124 Roberts Place, Ithaca, First

New

York 14850.

published 1992 by Cornell University Press. Printed in the United States of America

© The paper

in this

book meets the minimum requirements

American National Standard for Information Sciences Permanence of Paper for Printed Library Materials.

of the

ANSI

Z39. 48-1984.

Library of Congress Cataloging-in-Publication Data

Brown, Michael

(Michael Edward), 1954Flying blind: the politics of the U.S. strategic bomber program

Michael p.

E.

Brown. cm. (Cornell studies

/

E.

—

in security affairs)

Includes bibliographical references and index

ISBN 0-8014-2285-X

(cloth

:

alk.

—

.

paper)

— —

United States. Air Force Procurement History. 2. Bombers United States History. 3. United States Military policy. 1.

—

I.

Title.

II.

UGii23.B76

Series.

1991

358.4' 283 '00973

— dc20

91-55064

To

my

parents

Digitized by the Internet Archive in

2016 with funding from

China-America

Digital

Academic

Library

(CADAL)

https://archive.org/details/flyingblindpolitOObrow

Contents

Preface

Abbreviations

A

Note on Sources

1.

Introduction

2.

and Doctrinal Setting Building a jet Bomber: The B-45, B-46, B-47, and B-48 The First Intercontinental Bombers: The B-35, B-36, B-49, B-52, and B-60 The Push to Develop Supersonic Capabilities: The B-58 The Nuclear-Powered Bomber and the B-70 Low-Altitude Penetration: The B-i The Politics of Stealth: The B-iB and B-2 The Origins and Outcomes of Weapon Acquisition Programs

3. 4.

5. 6. 7.

8. 9.

Historical, Organizational,

ix xiii

xv

1

29 68

108 161

193

230 268 305

Appendix: Evolution of U.S. Air Force Weapon Acquisition Organizations

Index

[vii]

348 350

Preface

book about the weapon acquisition process. In it, I look at how American weapon acquisition programs begin and why they turn out the way they do. These are important issues because they affect both national and international security. The book's empirical foundation is a series of case studies of weapon acquisition in the U.S. strategic bomber program. I examine every major American bomber program of the postwar period, a total of fifteen programs ranging from the B-35 of the 1940s to the B-2 of the 1980s. My This

is

a

weapon

within established mission areas. Such mainstream programs have been neglected by most stufocus

is

acquisition activity that

falls

dents of weapon acquisition and arms racing because innovative programs, such as the first ballistic missile programs, are more glamorous. In this study I examine the value of technological, economic, bureaucratic, and strategic explanations about the origins of weapon programs. I find no significant support for technological and economic arguments, which are extremely prominent in the literature on the subject. I find some support for bureaucratic accounts and a great deal of support for strategic explanations of how weapon development efforts begin. Some of our ideas about the driving forces behind the Soviet-American arms race and the American military-industrial complex should therefore be reconsidered.

examine competing explanations of the cost, schedule, and performance problems that plague contemporary acquisition efforts in the United States. One school of thought holds that these problems are rooted in the American military's fascination with "high-tech" weaponI

also

fix]

Preface

Another school maintains that problems occur when development and production activities are allowed to overlap, that is, when they take place concurrently. Each group is half right: acquisition programs run into trouble when they try to advance technology and employ concurry.

rency

at the

same

time. In the period studied in this book, powerful

American military organizations to set their performance requirements far beyond the state of the art and to push their programs as fast as possible. The result, all too frequently, was disastrous from an acquisition standpoint. The U.S. Air Force and its predecessors, the Army Air Corps and Army Air Forces, were flying blind when they initiated many of their bomber programs. In many cases, emerging operational threats were far strategic

and bureaucratic

from clear

when

forces led

they set performance requirements for

addition, the technological possibilities for

were

new

systems. In

weapon system development

rarely well understood, because the Air Force decision

make

makers

thorough assessment of the technological horizon before they launched new ventures. They routinely compounded the unknowns their programs faced by setting performance requirements far beyond the state of the art. The Air Force and its predecessors were also flying blind when they imposed concurrency on programs that were intended to make a great leap forward technologically. They rushed headlong into projects that needed to proceed in a more orderly manner. As a result, many bomber programs were acquisition accidents waiting to happen. For gentle but unflagging encouragement during this project's formative stages, I owe an enormous intellectual and professional debt to Richard Rosecrance and George Quester, both then members of the Department of Government at Cornell University. also thank Arch Dotson and Woody Kelley, of the same department, who were always generous with their time and energy. John Mearsheimer and Jack Snyder read several drafts and offered the tough critiques only good friends would give. Kim Scheppele wrote forty-five pages of single-spaced comments, which she promises never to publish, on my penultimate draft. Kim also suggested the title; the Air Force and I both thank her. Robert Art, Robert Jervis, and Harvey Sapolsky read the manuscript for Cornell University Press and forced me to tighten up both my analysis and my prose. Stanley Brown, Michael Clough, Jacques Gansler, Thomas McNaugher, Judith Reppy, and Robert Spitzer also read the manuscript at various stages and contributed in important ways. Dora Davey Brown was present at the creation; hope she is as happy as I am to see the book finished. Renee de Nevers read the manuscript carefully and provided much-needed support as the endgame unfolded. To all my friends and colleagues, I am deeply grateful. generally failed to

a

I

I

[X]

Preface

For financial and institutional support, I thank the Peace Studies Program, the Walter S. Carpenter Endowment, the Government Depart-

ment, and the Center for International Studies, all at Cornell University; the Institute for the Study of World Politics, which gave me a grant to

around the country; the U.S. Arms Control and Disarmament Agency, which gave me a Hubert H. Humphrey fellowship; the Foreign Policy Studies Program at the Brookings Institution, where spent a year researching the case studies and working on the first draft; the Center for Science and International Affairs at Harvard University; the Committee on Research at Vassar College, where I taught travel to Air Force bases

I

for several enjoyable years; the

Center

for International Affairs at

Har-

vard; the Center for Strategic and International Studies in Washington;

and the International

London. am Brookings, Samuel Hunting-

Institute for Strategic Studies in

particularly grateful to John Steinbruner at

I

ton at Harvard, and Francois Heisbourg at the IISS for providing stim-

and supportive places

work. Portions of Chapter 8 originally appeared as "The Strategic Bomber Debate Today," Orbis (Summer 1984), and "B-2 or Not B-2: Crisis and Choice in the U.S. Strategic Bomber Programme," Survival (July-August 1988). I am indebted to the editors of these journals for permission to use some of that material in this book. I am also indebted to the Smithsonian Institution, the Boeing Company, the U.S. Air Force, and the U.S. Department of Defense for permission to use their photographs. John Baker allowed me to photocopy his files on the B-i. Karyn McCarthy provided valuable research assistance. Hussain Rafi Mohamed provided professional artwork. Helen Forrest and Kate Wilson, resident computer wizards at the IISS, made special efforts on my behalf and saved me weeks of time, at a minimum. Sheelagh Urbanoviez and Christine Zibas of the IISS spent many hours helping me proofread. My parents, Florence and Stanley Brown, introduced me to the excitement of learning and taught me the value of education. They have supported me in every way, and I will always be profoundly grateful to them. Given all this support, one can only wonder what took so long. ulating

to

Michael London

[xi|

E.

Brown

Abbreviations

AAC AAF

Army Army

AFLC AFSC

Air Force Logistics

ALCM

air-launched cruise missile

AMC

Air Materiel

AMSA ANP ARDC

Advanced Manned

ASD

Aeronautical Systems Division

ATSC

Air Technical Service

CBO

Congressional Budget Office Congressional Research Service

CRS

DD DDA DDP DORDP

Air Corps Air Forces

Air Force Systems

Command Command

Command Strategic Aircraft project

Aircraft Nuclear Propulsion project

Air Research and Development

Directorate of Directorate of Directorate of Directorate of

Command

Command

Development Development and Acquisition Development Planning Operational Requirements and Development

Plans

GAO GEBO

General Accounting Office Generalized Bomber study

ICBM

intercontinental ballistic missile

NACA NARS

National Advisory Committee for Aeronautics National Archives and Records Service

NDRC

National Defense Research Committee

NEPA

Nuclear Energy for Propulsion Aircraft project

Abbreviations

RAF SAC

Royal Air Force Strategic Air Command

SAM

surface-to-air missile

SLBM TAC USAF VDT

submarine-launched

WADC

Tactical Air

ballistic missile

Command

United States Air Force variable discharge turbine Wright Air Development Center

(xiv]

A

The data

for these case studies

classified Air Force

documents and

Note on Sources

came from

five

main sources:

(1)

de-

studies; (2) aerospace industry docu-

ments and

studies; (3) interviews with policy makers and program participants; (4) congressional hearings and reports; and (3) trade journals and the secondary literature. Together, these sources provided a

examined in this book, and they kept the case studies from becoming dependent on any single source of

detailed record of the historical period

information.

Some

came from approximately 3,000 pages of classified Air Force memoranda, reports, conference minutes, requirements statements, and command histories. Most of this material was declassified by the Air Force before of the

most interesting information

in the case studies

I

began

my

sified

documents

Force gave sift

also obtained approximately 750 pages of clasthrough the Freedom of Information Act. The Air

research, but

me

through

I

was allowed to records before submitting the documents

complete access

its

historical

to its classified archives;

I

I

needed for a declassification review. This access proved to be a doubleedged sword. It allowed me to satisfy myself that no archival stone had been left unturned and no filing cabinet left unopened. On the other hand, it took sixteen months to examine all the relevant files held at several widely scattered locations. The Air Force's Office of Security Review withheld no photocopied document from me, although it edited some of the notes took from classified documents with a pair of scissors; information deleted from these notes was not central to this study. Documents were examined and collected at the following locations: I

Modern vice,

Military Records Branch, National Archives

Washington, D.C. |xv)

and Records

Ser-

A Simpson

Note on Sources

Historical Research Center, Air University,

Maxwell Air Force

Base, Alabama. Office of Air Force History, Bolling Air Force Base, Washington, D.C.

History Office, Aeronautical Systems Division, Wright-Patterson Air Force Base, Ohio.

Command, Andrews

History Office, Air Force Systems

Air Force Base,

Maryland. History Office, Air Force Logistics

Command,

Wright-Patterson Air

Force Base, Ohio.

History Office, Air Force

Museum,

Wright-Patterson Air Force Base,

Ohio.

Museum, Washington, D.C.

History Office, National Air and Space

Access to aerospace industry records was uneven. Rockwell's and Northrop's historical files were not open to the general public for proprietary reasons. Several General Dynamics executives were kind

enough

to give

personal

files,

me

copies of documents and historical studies from their

but corporate

files,

were not open to the extensive and well-organized again,

public.

Boeing allowed me to examine its files on every program it had been involved in since the 1930s. Since Boeing was involved in most air force bomber programs in one way or another, these archives proved to be extremely valuable. Archival material was supplemented by approximately forty interviews with program participants including military and civilian decision makers as well as engineers and executives from Boeing, Rockwell, and General Dynamics. These interviews lasted from one to five hours. All interviewees were guaranteed anonymity and assured that they would not be quoted directly in this book. This guarantee seemed to encourage forthrightness.

Annual congressional authorization and appropriations hearings conducted by the House and Senate Armed Services and Appropriations committees were useful, as were special hearings focused on specific programs. Committee reports and studies conducted by the Congressional Budget Office, the Congressional Research Service, and the General Accounting Office tended to be more detailed and better organized. The trade press provided a wealth of technical information, although plowing through years (in some cases, decades) of unindexed journals was frequently tedious. The most useful trade journals were Aerospace Daily Armed Forces Journal International Aviation Week (now known as ,

,

Aviation Week and Space Technology), Defense Electronics, Defense News, Defense Systems Review, Defense Week, Flight International, International Defense Review, Jane's Defence Weekly,

and

Space! Aeronautics.

[xvi]

Flying Blind

[xvii]

[

1]

Introduction

book analyze the origins and outcomes of weapon acquisition programs A thorough understanding of these issues is essential to the promotion of both national and international security. The origins of weapon acquisition programs are important because those driven by parochial bureaucratic or commercial interests rarely serve national interests well. Unneeded weapons inevitably divert resources from legitimate military programs as well as vital social and economic activities. National security can be undermined as a result: directly, because deployment of the wrong kinds of weapons can weaken a nation's deterrent posture or lead to disaster on the battlefield; indirectly, because excessively high levels of defense spending can affect 2 a nation's long-term economic competitiveness The initiation of a weapon development program can also lead other powers to engage in weapon development efforts, creating an arms race. Arms races are always expensive propositions, and they can generate dangerous instabilities. At a minimum, they raise the level of tension between and among states. More ominously, if one of the participants in an arms race believes that the balance of power is moving against it, In this

I

1

.

.

The weapon acquisition process includes several distinct activities: research, develuse the term "acquisition" to opment, production or procurement, and deployment. denote the process as a whole and "development" and "production," for example, to refer to more specific sets of activities. 2. The impact of high levels of defense spending on a nation's long-term international 1.

1

analyzed in Richard Rosecrance, The Rise of the Trading State (New York: Basic Books, 1986); Paul Kennedy, The Rise and Fall of the Great Powers (New York: Random House,

position

is

1987).

W

Flying Blind

incentives for preventive

weapons

war may be

created.

If

an arms race features

preemptive or offensive action,

that favor

crisis instabilities

be aggravated. 3

may

how weapon development programs begin because they acquire a momentum quickly and become difficult to cancel. Policy makers consequently do not have as much Finally,

it

is

essential to

understand

control over the acquisition pipeline as they should, to

be

and arms

races tend

difficult to control.

Program outcomes are important as well. Technologically advanced programs aggravate qualitative arms races, which have proven to be especially difficult to control because it is hard to devise and verify limitations on the technical characteristics of weapons. Fast-paced programs can accelerate arms races and overwhelm slow-moving arms control

negotiations.

weapon

programs also have serious national security implications. Most weapon programs in the United States, for example, are plagued by a pernicious combination of cost overruns, schedule slippages, and performance deficiencies. When unit costs climb, production runs are scaled back to control total program costs. As a result, fewer weapons are bought for the force structure. When programs are delayed, weapons enter the force structure later

The outcomes

of

acquisition

than expected. They consequently face a more formidable operational environment than they otherwise would have, which impinges on their effectiveness in the short run and makes them obsolete sooner. Finally, when weapons fail to meet their performance specifications, the opera-

weakened. Long-term acquisition trends in the United States are ominous. Each new generation of weapons takes longer to develop and costs more than its predecessor. If current cost and budgetary trends continue, it has been estimated that the Department of Defense will be able to afford only one new aircraft in tional effectiveness of the force structure

the year 2054. 4 Unfortunately, the origins

is

and outcomes

grams are poorly understood, because

of

weapon

acquisition pro-

existing explanations tend to be

3. For a discussion of the risks and dangers associated with arms races, see Samuel P. Huntington, "Arms Races: Prerequisites and Results," in Robert J. Art and Kenneth N. Waltz, eds.. The Use of Force (Boston: Little, Brown, 1971), pp. 365-401; Colin S. Gray, "The Arms Race Phenomenon," World Politics 24 (Oct. 1971), 39~79- For a through overview of the literature on preventive war, see Jack S. Levy, "Declining Power and the Preventive Motivation for War," World Politics 40 (Oct. 1987), 82-107. The relationship between offensive military capabilities and crisis instability is analyzed in George H. Quester, Offense and Defense in the International System (New York: Wiley, 1977); Robert Jervis, "Cooperation under the Security Dilemma," World Politics 30 (Jan. 1978), 167-214; Stephen W. Van Evera, The Causes of War (Ph.D. Dissertation, University of California at Berkeley, 1984), chaps. 1-3. 4. See the discussion in Jacques S. Gansler, Affording Defense (Cambridge: MIT Press,

1989), pp. 169-179.

[2]

Introduction

narrow and contradictory.

My main goal in this book,

the validity of existing theories about

weapon

therefore,

acquisition

and

to

is

to test

develop

integrated explanatory frameworks for analyzing both the origins and the outcomes of

weapon

do this, I address two main sets of questions. First, how do weapon development programs begin? Where do the ideas for these programs originate? What triggers the process in the first place? Second, why do programs turn out the way they do? Why do many push the technological state of the art and move at a fast pace? How do they acquire momentum? Why are some contemporary programs successful in meeting their cost, schedule, and performance targets while most are strikingly unsuccessful? In attempting to answer these questions, analyze in detail the fifteen major U.S. strategic bomber programs of the postwar period: the B-35, acquisition programs. To

I

B-36, B-45, B-46, B-47, B-48, B-49, B-52, B-58, B-60, nuclear-powered

bomber, B-70, B-i, B-iB, and B-2. 5 These programs, none of which has been the subject of serious scholarly examination, are grouped into six case studies. 6 Others including fighter, helicopter, tank, and missile programs are discussed in passing, primarily in the later stages of the

—

—

book.

Program Origins philosophers and international relations theorists have long debated the causes of state behavior. 7 In recent years, the debate has been carried by those who focus on the external or systemic determinants of state behavior, on the one hand, and those who focus on the internal or domestic determinants of state action, on the other. Within one systemic, three domestic have this debate, four main arguments Political

—

—

programs had their origins during World War II. Obviously, some bombers are missing from this numerical sequence. The B-38 and B-40, for example, were modified versions of the B-17, which was designed and built in the 1930s and produced in quantity during the war. Similarly, the B-39, B-44, B-50, and B-54 were advanced versions of the B-29, another World War II bomber. The B-42, B-43, B-51, B-53, B-57, and B-66 were tactical bombers; their range and payload capabilities were not suitable for strategic missions. The B-55 and B-59 were simply design studies. Finally, many "bomber numbers" were assigned to missile projects. Examples include the B-61 (Matador), B-62 (Snark), B-63 5.

Some

of these

(Rascal), B-64 (Navaho), B-65 (Atlas), B-68 (Titan), B-72 (Quail), B-75 (Thor), B-77

(Hound

Dog), B-78 (Jupiter), B-80 (Minuteman), and B-87 (Skybolt). After the cancellation of the B-70 in the early 1960s, the Air Force broke with the past and began renumbering its bomber programs; hence, the B-i was the successor to the B-70. 6. The method of comparative case study is outlined in Alexander L. George, "Case Studies and Theory Development: The Method of Structured, Focused Comparison," in Paul Gordon Lauren, ed.. Diplomacy (New York: Free Press, 1979), pp. 43-68. 7. The classic survey of this debate is found in Kenneth N. Waltz, Man, the State, and War (New York: Columbia University Press, 1959).

[31

Flying Blind

been advanced ly,

to explain military policy in general

the origins of

weapon

and, more specifical-

acquisition programs: strategic explanations,

bureaucratic explanations, economic explanations, and technological ex-

planations

8 .

Strategic Explanations

Strategic explanations,

which include balance-of-power theories and

accounts of international relations, maintain that the wellspring of state behavior is the character of the international system within which states operate 9 According to this line of argument, the most important

realist

.

system is its anarchic nature; since national security cannot be guaranteed by enforceable international laws or any supranational agency, states must look to their own devices to safeguard or promote vital national interests. As a result, states are said to be acutely aware of international developments that have the potential to change the global balance of power and adversely affect national security. In theory, any of the following developments could disrupt the balance and pose a threat to national security: the emergence of a new adversarial (or potentially adversarial) power; the formation of a new adversarial (or potentially adversarial) alliance; the expansion of an adversary's power through conquest, alliance, or devotion of more resources to military activities; the appearance of new military technologies or weapon systems in an adversary's arsenal; the collapse or defeat of a friendly power; a withdrawal from or disintegration of friendly alliance structures; or a military defeat that reduces access to vital natural resources or forward bases or leads to the loss of territorial buffers. States can respond to changing strategic conditions such as these in one of two basic ways. One option is external: they can try to form new alliances or augment existing ones. The other option is internal: they can 10. devote more national resources to military activities 10 Changing strategic circumstances might therefore lead policy makers to initiate new weapon development programs. characteristic of the international

.

8.

Many have argued

shaping

its

that a state's internalized foreign policy goals are instrumental in international behavior. In particular, some states have expansionistic, ag-

gressive foreign policies or foreign policies guided by missionary ideologies of one kind or another. Aggressive states such as these might initiate military programs in order to intimidate opposing states or to prepare for conquest. Although several internal determinants of policy are analyzed in this book, explaining the behavior of genuinely malignant

such as Nazi Germany is beyond the scope of this study. See Waltz, Man, the State, and War, chap. 6; Waltz, Theory of International Politics (Reading, Mass.: Addison-Wesley, 1979), chap. 6; Hans Morgenthau, Politics among Nations (New York: Knopf, 1978); Barry R. Posen, The Sources of Military Doctrine (Ithaca: Cornell states 9.

University Press, 1984), pp. 59-67. Posen, Sources of Military Doctrine, pp. 61-63.

[4]

Introduction

Samuel Huntington has observed that internal balancing mechanisms have assumed greater importance over the course of the past hundred years because (1) the number of great powers in the system (that is, the

number of potential alliance partners) has declined; (2) forces-in-being have become more important than territory for national security; and (3)

the industrial

significantly.

11

and technological

have increased

Barry Posen has argued that politically isolated states and

superpowers are especially the

capacities of states

reliant

on

internal balancing

mechanisms,

in

case, because allies are difficult to find and, in the second,

first

because alliances with smaller powers are not likely to enhance a superpower's security position substantially. 12 Balance-of-power theorists would argue that, since both the United

and the Soviet Union were highly dependent on internal balancing mechanisms during the Cold War, the most likely response to the appearance of a new weapon on one side was deployment of a new weapon on the other. This response could take the form of a weapon designed to counter or neutralize the other's new weapon, thereby reStates

storing the operational balance.

Alternatively,

the responder could

weapon, 13 allowing him to claim that the force structures of the two sides were still comparable and providing him with a useful bargaining chip for arms control negotiations. Many Soviet and American weapon decisions, it is said, can be explained by this "action-reaction" process. 14 Since it generally takes five to fifteen years to develop, produce, and field a weapon system, each superpower was reluctant to wait until its adversary deployed a new weapon before initiating an appropriate response; neither wanted to be confronted by a long and potentially dangerous capabilities gap or window of vulnerability. So, each superpower tried to anticipate what the other had in the acquisition pipeline and respond accordingly. Therefore, it might be more accurate to characmatch the

initiator

and build

a similar kind of

process as an "anticipation-reaction" process, where reactions are driven by discrete developments as well as nagging strategic uncertainties. Former U.S. Secretary of Defense Robert terize

11.

this interactive

Asa

tury; see

result,

arms

races have

become more frequent

since the early nineteenth cen-

Huntington, "Arms Races," pp. 367-372.

Posen, Sources of Military Doctrine, pp. 61-67. 13. Graham T. Allison (who is not known as a balance-of-power theorist) discusses these arguments in "Questions about the Arms Race: Who's Racing Whom? A Bureaucratic Perspective," in Robert Pfaltzgraff, ed.. Contrasting Approaches to Strategic Arms Control (Lexington, Mass.: Lexington Books, 1974), pp. 31-72. 14. In discussing the Soviet-American arms race, Herbert York and G. Allen Greb argued that the action-reaction cycle "played a very fundamental role in determining the 12.

course of events"; see "Military Research and Development: the Atomic Scientists 33 (jan. 1977), 13. [5]

A

Postwar History,"

Bulletin of

Flying Blind

McNamara

"What

understand here is that the Soviet Union and the United States mutually reinforce one another's strategic plans. Whatever their intentions or our intentions, actions or even realistically potential actions on either side relating to the buildexplained:

is

essential to

—

—

up

of nuclear forces necessarily trigger reactions

Some would argue ligence process

that,

on the other side ." 15

given the uncertainties inherent in the

and the tendency

of policy

intel-

make worst-case adversaries, weapon de-

makers

to

assumptions about the military capabilities of velopment decisions are driven by an "anticipation-overreaction" process, where each overreaction feeds successive rounds of overreactions 16 Although this chain of events might produce an arms race that serves neither side's interests well, the process would nonetheless be driven by legitimate national security concerns which grow out of devel.

opments

in the strategic arena.

Bureaucratic Explanations

and balance-of-power explanations have been criticized on several grounds over the past twenty-five years, mainly because they neglect the internal determinants of state behavior. One of the most prominent critiques of balance-of-power theory argues that one cannot Strategic

understand

without understanding the important role that bureaucracies play in the policy-making process. Bureaucratic-politics analysts maintain that it is inaccurate to characterize state behavior as purposive and policy making as a rational process of analysis and choice designed to maximize national interests and goals. They argue that policy is the product of a political process of pulling and hauling among bureaucratic players with different interests, different stakes in different issues, and varying amounts of influence

15.

Robert

59. Similarly,

state behavior

McNamara, The George

Essence of Security (New York: Harper and Row, 1968), pp. 58Rathjens has written that "the action-reaction phenomenon, with the

reaction often premature and/or exaggerated, has clearly been a major stimulus of the strategic arms race"; see "The Dynamics of the Arms Race," in Bruce M. Russett and Bruce

G.

Anns

Control? (San Francisco: Freeman, 1979), p. 37. 16. Albert VVohlstetter has argued that the United States routinely underestimated Soviet strategic force levels in the 1960s, but he does not dispute that U.S. decisions were influenced by intelligence projections about Soviet capabilities. For more on the overestimation/underestimation debate, see Wohlstetter, "Is There a Strategic Arms Race?" Foreign Policy no. 15 (Summer 1974), 3-20; Wohlstetter, "Rivals, but No Race," Foreign Policy, no. 16 (Fall 1974), 48-81; Michael Nacht, "The Delicate Balance of Error," Foreign Blair, eds., Progress in

no. 19

(Summer

1973), 163-177; Wohlstetter, Foreign Policy, no. 20 (Fall 1975), 170-198. Policy,

"Optimal Ways

to

Confuse Ourselves,"

[

6

]

Introduction

over the outcomes of these internal negotiations. It is especially important, they argue, to recognize that parochial bureaucratic interests can differ substantially from national interests broadly defined 17 According to Morton Halperin, core bureaucratic interests include preserving main missions; expanding mission capabilities; enlarging departmental budgets; increasing organizational autonomy from central authorities; improving organizational morale by staying involved in activities of national importance and by providing attractive career prospects; and increasing the organization's political influence, which allows it to pursue all its goals more effectively 18 Strategic bombardment, for example, is a core mission of the U.S. Air Force, one that has defined its essence since its inception. Bureaucratic-politics analysts would consequently predict that the Air Force would place a high priority on having bomber .

.

whenever possible. analysts would also predict that

projects in the acquisition pipeline Bureaucratic-politics

military organi-

zations tend to be highly successful in defending core missions and

promoting

institutional interests. Military organizations are

powerful

because they have large budgets and because they supervise many large acquisition programs. As a result, the American military, for example, has influential constituents in industry and labor as well as loyal allies in Congress. In addition, the military services in the United States rarely challenge the core missions of the other services, and civilian policy

makers frequently defer

As

to the military's professional

many

judgment on de-

who

cannot be characterized as bureaucratic-politics analysts, have concluded that military organizations play a decisive role in shaping national policy. Robert Heilbroner, for example, has argued that the military is "a selfcontained entity, capable of impressing its views and imposing its will not only on the civil establishment to which it pays ritual obeisance, but over a section of the economy in which the language of private enterprise is merely a fiction to hide its absolute authority." John Kenneth Galbraith has concluded that "the military services, not their industrial power" in the militarysuppliers, are the prime wielders of industrial complex, and that, because of their high degree of depenfense issues.

a result,

scholars, including several

.

.

.

Brown, 1971), chap. 5; Graham T. Allison and Morton H. Halperin, "Bureaucratic Politics: A Paradigm and Some Policy Implications," World Politics 24 (Spring 1972), 40-79. The best survey of the bureaucraticpolitics literature is found in Robert J. Art, "Bureaucratic Politics and Foreign Policy: A 17.

Graham

T.

Allison, Essence of Decision (Boston: Little,

Critique," Policy Sciences 4 (Dec. 1973), 467-490. For a pointed critique of bureaucraticpolitics analyses, see Stephen D. Krasner, "Are Bureaucracies Important? (Or Allison Wonderland)," Foreign Policy, no. 7 (Summer 1972), 159-179. 18. Morton Halperin, Bureaucratic Politics and Foreign Policy (Washington: The Brookings Institution, 1974), chap.

[7]

3.

Flying Blind

dence on defense contracts, many companies are "captive contractors" 19 of different military services .

Moreover, because they are relatively permanent fixtures on the policy-making scene, military organizations are said to play a particularly important role in shaping weapon acquisition programs. According to Graham Allison and Frederic Morris, "the services and their subunits are the primary actors in weapon development. Consequently, force posture is shaped by the goals and procedures and especially the missions and weapon systems to which services (and subunits) are committed. Political officials might disturb this process; only rarely do they control

20 it ."

Bureaucratic-politics analysts, therefore, expect military organizations to play the leading role in initiating

most weapon development pro-

grams. The mainstream commands of the services are expected to take the lead within established mission areas, while bureaucratic entrepreneurs, mavericks, insurgents, and separatists are expected to take 21 In the case of the the lead in areas where no established mission exists former, the military is expected to build weapons designed to imple.

ment

existing operational doctrines. In the case of the latter, operational

doctrines might not yet be well formed. But, in each case, the military's motivation is promoting bureaucratic interests rather than responding to strategic

developments.

Economic Explanations

Another school

on the military-industrial complex argues secondary and essentially reactive role in the

of thought

that the military plays a

policy-making process. Gabriel Kolko, for example, has stated that the military tends to be "docile" and that it is "unquestionably among the most restrained of those in power ." 22 According to this line of think19.

Heilbroner and Galbraith quoted in Charles C. Moskos,

Military-Industrial

1974 )/ 5 ° 3

jr.,

Complex: Radical Critique or Liberal Bogey?"

"The Concept

of the

Social Problems 21 (April

-

and Frederic Morris, "Armaments and Arms Control: Exploring the Determinants of Military Weapons," in Franklin A. Long and George W. Rathjens, eds., Arms, Defense Policy, and Arms Control (New York: Norton, 1976), p. 123. See also Allison, "Questions about the Arms Race," pp. 42-51; Morton H. Halperin, "The Decision to Deploy the ABM: Bureaucratic and Domestic Politics in the Johnson Administration," 20.

World

Graham

Allison

25 (Oct. 1972), 62-95. 21. For more discussion of the roles played by bureaucratic entrepreneurs and insurgents in promoting new programs, see Vincent Davis, The Politics of Innovation: Patterns in Naiy Cases (Denver: University of Denver, Graduate School of International Studies Monograph, 1967); Frederic A. Bergerson, The Army Gets an Air Force (Baltimore: Johns Hopkins Politics

University Press, 1980). 22.

Kolko quoted

in

Moskos, "Concept of the Military-Industrial Complex,"

p.

also pp. 498-512.

[

8

]

506; see

Introduction

behind defense policy in general and weapon acquisition in particular is the defense industry, which in turn is driven by its own economic interests 23 Many American companies, to be sure, have long been dependent on defense contracts 24 Some analysts argue that, because of compelling economic interests, defense contractors have generally been responsible for initiating new weapon development programs. According to this line of analysis, most contractors invest a great deal of time and money on ing, the driving force

.

.

research and development activities of their own, in the hope that fund-

come out

able projects might

shops.

When

of their laboratories

and engineering

interesting ideas are generated, they are passed along to

which is usually interested in pursuing them for parochial reasons of its own. Later, corporate wealth and influence are said to play a role in securing support for programs at higher levels in the executive branch. Congress, and the media. As one corporate executive put it, the military "depends on companies like ours to tell them what they need ." 25 Another executive observed: the appropriate military service,

At the

aircraft

corporation where

military contracting

seemed

I

worked

related not so

as an engineer for

much

many

years,

to national security as to

who

and it is their interest which must be stimulated. Certain segments of the work force, therefore, are assigned to discover new "ventures." An improved

corporate security. ...

version of an existing

It is

the

weapon

armed

services

are the customers,

or even something dramatically

new may be

The reaproposed to the military by the defense-supported contractor. son for all these new weapons and weapon systems is the profit motive, a matter of keeping business going rather than of protecting the country 26 .

.

.

.

Ronald Fox argued in his examination of American weapon acquisition that "defense contractors are profoundly influential in the origination and development of new program ideas ." 27 Mary Kaldor concluded that

J.

and Paul M. Sweezy, Monopoly Capital (New York: Monthly Review, 1966); "No Business like War Business," The Defense Monitor 16 (1987), 18. Historical analyses of the U.S. military-industrial complex can be found in Benjamin 23.

See, for example, Paul A. Baran

Franklin Cooling, ed.. War, Business, and American Society (Port Washington, N.Y.: Kennikat Press, 1977). For a comparative perspective, see Cooling, ed., War, Business, and World Military-Industrial Complexes (Port Washington, N.Y.: Kennikat Press, 1981). 24. For a dispassionate analysis of the U.S. defense industry's dependence on defense contracts, see Jacques S. Gansler, The Defense Industry (Cambridge: MIT Press, 1980), chaps. 2-3. See also the comparative case studies in Nicole Ball and Milton Leitenberg, eds., The

(New York: St. Martin's, 1983). The Baroque Arsenal (New York: Hill and Wang, 1981), p. 69. Kaldor, 25. Quoted in 26. Robert C. Aldridge, "How Defense Industries Keep the Business Coming," Bulletin

Structure of the Defense Industry

Mary

Atomic Scientists 32 (May 1976), 44-46. 27. J. Ronald Fox, Arming America (Boston: Harvard University, Graduate School of Business Administration, 1974), p. 101.

of the

[9]

Flying Blind

the existence of large corporations

which are dependent on government

contracts creates an "industrial imperative" for

ment programs

new weapon develop-

States 28

United James Kurth has argued that new development and production contracts tend to be awarded to important contractors whenever old contracts are about to expire or be canceled. Similarly, new contracts tend to be awarded to important contractors who are on the brink of bankruptcy. Strategic and bureaucratic explanations for what he calls the "follow-on imperative" and the "bail-out imperative" are unsatisfactory, in his view. Kurth maintains that economic considerations play a central role in decisions about contract awards 29 in the

.

.

Technological Explanations

A

argument about the origins of weapon development programs emphasizes the importance of technological factors in the decisionmaking process. According to this argument, the original impetus for weapon development is the emergence of new discoveries in laboratories, which communicate word of their technological breakthroughs to the military. The military, it is said, reacts to this news by devising military rationales and requirements for these technologically irresistible ideas. Once rationales have been devised, the military lobbies for funding and political support. Programs build up momentum as time goes by and eventually move into full-scale development and production. Because the ideas that emerge from weapons laboratories are so irresistfinal

ible, a

"technological imperative"

is

often said to be behind the origins of

weapon development programs and

arms race itself. This argument has been extremely prominent for decades. In their classic study of the American weapon acquisition process, Morton Peck and Frederic Scherer concluded that the precipitating factor in a decision to

develop a

discovery

30 .

new weapon system

Solly

the

is

Zuckerman, former

usually a technical or scientific

chief science adviser to the British

Ministry of Defence, has elaborated on the same theme. "Ideas for a new weapon system derive in the first place, not from the military, but

from different groups of

scientists

and technologists," he wrote, and

Mary

Kaldor, "The Weapons Succession Process," World Politics 38 (July 1986), 593. See James R. Kurth, "A Widening Gyre: The Logic of American Weapons Procurement," Public Policy 19 (Summer 1971), 373-404; Kurth, "Why We Buy the Weapons We Do," Foreign Policy, no. 11 (Summer 1973), 33-56. Arnold Kanter and Stuart Thorson argue that several other factors are more important in production decisions than follow-on im28. 29.

peratives; see

"The Weapons Procurement Process: Choosing among Competing Theo-

ries," Public Policy

20 (Fall 1972), 479-520. 30. Morton Peck and Frederic Scherer, The Weapons Acquisition Process (Boston: Flarvard University, Graduate School of Business Administration, 1962), p. 226. [10]

Introduction

went on

who by

convention are a country's official advisors on national security, as a rule merely serve as the channel through which the men in the laboratories transmit their views ." 31 John Steinbruner and Barry Carter have observed that military requirements come out of "a process that is driven by technical and scientific developments. A major advance in technology in itself provides the impetus for applying it to a weapon; military requirements out that "military chiefs,

to point

tend to flow from

that, rather

than from a prior judgment of actual

need ." 32

weapon

and arms racing emphasize the roles played by scientists and engineers. Although scientists and laboratories might have a vested interest in promoting In short, technological explanations of

their technological breakthroughs, the discoveries

responsible for triggering the

acquisition

themselves are held

weapon development

process.

It

is

no

exaggeration to say that this line of thinking constitutes the conven-

wisdom on

tional

the subject in the arms control and scientific

commu-

nities 33 .

Analyzing the Origins of Weapon Acquisition Programs

Many analysts combine

several lines of argumentation in their studies

Unfortunately, their analysis generally becomes a result. For example, in addition to emphasizing actionreaction processes in his analysis of arms races, George Rathjens notes that "the simple desire to bring to fruition an interesting and elegant

weapons muddled as

of

issues.

technological concept

who

is

known

is

also important ." 34 Similarly,

Hans Morgenthau,

and wide as one of the leading proponents of realism, the nuclear arms race is driven by technological devel-

far

has written that

Zuckerman, Nuclear Illusion and Reality (New York: Viking, 1982), pp. 103, 105. 32. John Steinbruner and Barry Carter, "Organizational and Political Dimensions of the Strategic Posture: The Problems of Reform," in Long and Rathjens, eds., Anns, Defense Policy, and Arms Control, p. 143. DeWitt, "Labs Drive 33. To get a sense of how prominent this argument is, see Hugh E. the Nuclear Age Assessing eds.. McGuire, the Arms Race," in Len Ackland and Steven (Chicago: Educational Foundation for Nuclear Science, 1986), pp. 101-106; Marek Thee, 31.

Solly

"Science-Based Military Technology as a Driving Force behind the Arms Race," in Nils Petter Gleditsch and Olav Njolstad, eds., Arms Races: Technological and Political Dynamics (London: Sage, 1990), pp. 103-120; Harvey Brooks, "The Military Innovation System and the Qualitative Arms Race," in Long and Rathjens, eds., Arms, Defense Policy, and Arms Control, pp. 75-98; Marvin L. Goldberger, "Does the Technological Imperative Still Drive the Arms Race?" in Roman Kolkowicz and Neil Joeck, eds., Arms Control and International Security (Boulder, Colo.: Westview, 1984), pp. 63-67; Deborah Shapley, "Technology Creep and the Arms Race," Science 201 (Sept. 1978), 1102-1105; Ralph Lapp, Arms beyond Doubt (New York: Cowles, 1970). See also widely used university textbooks such as Dietrich Schroeer, Science, Technology, and the Arms Race (New York: Wiley, 1984), chap. 13. 34.

Rathjens, "Dynamics of the

In]

Arms

Race,"

p. 37.

Flying Blind

opments

that are "politically

and

militarily irrational.

.

.

.

What seems

to

be technologically possible is put into practice for no better reason but because it can be done ." 35 Morton Halperin claims that decisions about

weapon development programs

"are usually

made by

the military ser-

based on their interests and what is technologically possible ." 36 In all these cases, nothing is said about the relative importance of strategic, bureaucratic, economic, and technological factors or the conditions under which each factor is especially important. To date, the most sophisticated attempt to integrate these factors into a coherent explanation of weapon acquisition has been made by Matthew Evangelista. He argues that new technological discoveries trigger the weapon development process in the United States, and that scientists take the lead in suggesting military applications for these discoveries. Scientists seek support for their new ideas among their military associates and, as programs get under way, allies are sought in industry, higher levels of the executive branch, and Congress. According to Evangelista, it is only in the middle and later stages of the process that external factors, such as the identification of military threats, play a role, and they are important only in that they are needed to justify full-scale vices,

development and production. The centralized, rigid, secretive nature of the Soviet system, on the other hand, has acted to inhibit innovation. Technological initiatives have been routinely stifled and little progress made until emerging strategic threats attract the attention of high-level civilian policy makers. Defense priorities are then reassessed, resources reallocated, and crash programs devoted to developing new and important military technologies. According to Evangelista, strategic factors have thus played a central role in Soviet weapon acquisition 37 The main limitation of Evangelista's study is that it analyzes only innovative programs programs that involve fundamental changes in military missions, operational doctrines, and other entrenched organizational arrangements. His generalizations may not apply to programs that fall within established mission areas, which he carefully notes. This is an important limitation, though, because as Evangelista acknowledges, mainstream programs constitute the main activity of military .

—

35. Hans Morgenthau, "Some Political Aspects of Disarmament," in David Carlton and Carlo Schaerf, eds.. The Dynamics of the Arms Race (London: Croom Helm, 1975), p. 62. 36. Morton Halperin, "the Limited Influence of the Military-Industrial Complex," in Morton H. Halperin, Jacob A. Stockfisch, and Murray L. Weidenbaum, The Political Econo-

my

of the Military-Industrial Complex (Berkeley: University of California, Institute of Business and Economic Research, 1973), p. 4, emphasis added. 37. Matthew Evangelista, Innovation and the Arms Race (Ithaca: Cornell University Press, 1988), pp. 22-82 227-228, 239-240. ,

[

12 ]

Introduction

research and development 38

They have certainly played a critical role in driving the Soviet-American arms race. The main focus of this book is weapon acquisition activity that falls .

within established mission areas. Programs of this type have been neglected by most students of weapon acquisition and arms racing because 39 Only a few scholinnovative programs appear to be more important .

40 Alhave focused on mainstream development efforts though my arguments about the origins of mainstream programs are developed in more detail in later chapters, they can be sketched out

arly studies

.

here.

Major programs that fall within established mission areas are generally initiated by the military services. Military organizations are highly sensitive to threats

— strategic as well as bureaucratic — to their core mis-

and bureaucratic developments are generally responsible for triggering mainstream programs. The military typically sets performance requirements for new systems at the outset of its development efforts, and these requirements are frequently unattainable sions. Therefore, strategic

given available technologies. In

many

when

many

cases, these requirements are not

unforeseen technological breakthroughs finally materialize. As a general rule, technological developments are not the driving forces behind these kinds of programs; military demands usually outpace technological supply by a wide margin. Similarly, defense contractors usually take a back seat in the early stages of these programs. They play a greater role in suggesting component and model improvements once the process is well under way. In short, strategic and bureaucratic factors are especially important in the early stages of programs that fall within established mission areas.

met

38.

for

years,

totally

Ibid., pp. 12-13, 5 1 ’ 2 45-

Studies of innovative programs include Warner R. Schilling, "The H-Bomb Decision: How to Decide without Actually Choosing," Political Science Quarterly 76 (March 1961), 24Rand Corporation Research Memoran46; Robert L. Perry, System Development Strategies, dum, RM-4853-PR, August 1966; Davis, Politics of Innovation; Michael H. Armacost, The Harvey M. Politics of Weapons Innovation (New York: Columbia University Press, 1969); Sapolsky, The Polaris System Development (Cambridge: Harvard University Press, 1972); Edmund Beard, Developing the ICBM (New York: Columbia University Press, 1976); Robert Stephen E. Ockenden, "The Domestic Politics of Cruise Missile Development, J. Art and 1970-1980," in Richard K. Betts, ed.. Cruise Missiles (Washington: The Brookings Institution, 1981), pp. 359-413; Evangelista, Innovation and the Arms Race. Military (Boston: Little, Brown, 40. See Robert J. Art, The TFX Decision: McNamara and the Program (Ph.D. Disserta1968); Richard G. Head, Decision-Making on the A-7 Attack Aircraft The F-111 and the tion, Syracuse University, 1971); Robert F. Coulam, Illusions of Choice: Problem of Weapons Acquisition Reform (Princeton: Princeton University Press, 1977); D. Douglas Dalgleish and Larry Schweikart, Trident (Carbondale, 111.: Southern Illinois Decision (BoulUniversity Press, 1984); Lauren H. Holland and Robert A. Hoover, The 39.

MX

der, Colo.:

Westview, 1985).

Flying Blind

The bomber programs examined

in this

book are

classic

examples of

such programs. Detailed analyses of the decision-making processes in these cases pinpoint the impact different factors had during different stages of the acquisition process.

Most

of these

programs were begun

in

response to strategic developments: a change in the operational environment threatened to make existing bombers militarily ineffective. In some cases, bureaucratic motivations were also important. Many analysts have argued that military requirements are after-the-fact rationalizations for programs inspired by either technological opportunism or economic self-aggrandizement. 41 These case studies do not offer any significant support for technological or economic explanations of how weapon de-

velopment programs begin. The Army Air Corps, Army Air Force, and U.S. Air Force were flying blind in several respects when they initiated their bomber development programs. First, the nature of the operational threat they faced was not always clear because intelligence reports were frequently vague. At the

same

bomber development were organizations often failed to make a

time, the technological possibilities for

not always

clear, since air force

thorough assessment of the technological horizon before initiating development; thus, doctrinal and organizational preconceptions played an important role in shaping key decisions. 42 In every case examined in this book, air force decision makers compounded the unknowns their programs faced by setting performance requirements beyond, frequently far beyond, the state of the art.

Program Outcomes Ihe U.S. Department of Defense has conducted some spectacularly unsuccessful weapon acquisition programs over the years. Cost, schedule, and performance problems have been the rule, not the exception. These problems are both chronic and systemic; they have deep historical roots, and they affect all military services. Two classic studies of weapon

found substantial cost and schedule growth in most programs, with costs exceeding original estimates by an average of

acquisition in the 1950s

See, for example. Brooks, “Military Innovation System," pp. 91-92; Steinbruner Carter, “Organizational and Political Dimensions," p. 143. 41.

and

The importance of cognitive and organizational predispositions in decision making analyzed in James G. March and Herbert A. Simon, Organizations (New York: Wiley, 1958), chap. 6; John D. Steinbruner, The Cybernetic Theory of Decision (Princeton: Princeton University Press, 1974), chaps. 3-5; Robert Jervis, Perception and Misperception in International Politics (Princeton: Princeton University Press, 1976), chaps. 4-5; Posen, Sources of Military Doctrine, chaps. 2, 7; Jack Snyder, The Ideology of the Offensive (Ithaca: Cornell 42.

is

University Press, 1984), chaps.

i, 8;

Coulam,

Illusions of Choice,

chaps.

1, 6.

[14]

,

Introduction

300 percent and schedules slipping by an average of 50 percent. In addition, most programs failed to meet one or more of their primary

performance requirements. 43 A highly regarded Rand Corporation study of acquisition outcomes in the 1960s found that the typical program had a cost overrun of about 40 percent and a schedule slippage of about 15 percent. System performance deviated from original specifications by an average of 30 to 40 percent. 44 Despite a concerted effort to reform the acquisition process in the Pentagon in the early 1970s, programs had cost overruns of 20 percent and schedule slippages of around 45 Serious problems continued to plague 15 percent during that decade. acquisition programs in the 1980s. The Sergeant York antiaircraft gun was canceled in 1985, for example, because it was tremendously expensive

and technically

defective.

This situation has not gone unrecognized. Indeed, the weapon acquisition process has been scrutinized by an impressive array of presidential commissions, congressional committees, and defense analysts since the 1950s. 46 Several common themes run through these studies,

with a strong consensus that the acquisition process in general would

work

better

if

six steps

were taken.

First, acquisition regulations,

pro-

cedures, and organizations should be streamlined, reducing red tape and clarifying channels of communication and areas of responsibility.

Second, engineering changes and budgetary fluctuations should be A. W. Marshall and W. H. Meckling, "Predictability of the Costs, Time, and Success of Development," in Richard R. Nelson, ed.. The Rate and Direction of Inventive Activity (Princeton: Princeton University Press, 1962), pp. 461-475; Peck and Scherer, Weapon Ac43.

425-444. Robert Perry et al.. System Acquisition

quisition Process, pp.

44.

PR/ARPA, June 45.

1971, pp.

Edmund Dews

et

R-2516-DR&E, Oct. 1979,

v,

Strategies,

Rand Corporation

Report, R-733-

1-11.

al..

Acquisition

Policy

Effectiveness,

Rand Corporation Report,

p. 2 7.

46. For the conclusions of various high-level study groups, see Commission on the Organization of the Executive Branch of the Government (the Hoover commission). Report on Military Procurement June 1955; Blue Ribbon Defense Panel (the Fitzhugh commission). Report to the President and the Secretary of Defense on the Department of Defense, July 1970; Commission on the Organization of Government for the Conduct of Foreign Policy (the

Murphy commission). Report, Vol. 4, pt. II, June 1975; Blue Ribbon Commission on Defense Management (the Packard commission), A Quest for Excellence, June 1986; Secretary of Defense Richard Cheney, Defense Management, July 1989. In addition, see congressional reports such as Weapons Acquisition Policy and Procedures, Report of the Special Panel on Defense Procurement Procedures, House Armed Services Committee, 97th Cong., 1st sess., Dec. 1981; Defense Organization, Staff Report, Senate Armed Services Committee, 99th Cong., 1st sess., Oct. 1985. Studies by independent defense analysts and scholars include U.S. Defense Acquisition (Washington: Center for Strategic and International Studies, March 1987); J. Ronald Fox, The Defense Management Challenge (Boston: Harvard Business School Press, 1988); Gansler, Affording Defense; Thomas L. McNaugher, New Weapons, Old Politics (Washington: The Brookings Institution, 1989); Alan R. Yuspeh, "The Acquisition Process," in James A. Blackwell, Jr., and Barry M. Blechman, eds., Making Defense Reform Work (Washington: Brassey's, 1990), pp. 215-236.

Flying Blind

minimized, promoting program stability. Third, independent cost estimates should be used more widely to prevent contractors and the services from misrepresenting program costs. Fourth, independent tests of new weapons should be conducted on a more regular basis, providing decision makers with better information about technical progress and operational effectiveness. Fifth, program management should be improved; this could be accomplished by providing acquisition personnel with better training, better pay, more opportunities for advancement, and longer program assignments. Sixth, criminal activity including should be invesfraud, overbilling, and misuse of insider information tigated and prosecuted aggressively. Although defense analysts agree about these general issues, they disagree about why some programs turn out worse than others. One school of thought holds that programs with ambitious development objectives are especially prone to cost, schedule, and performance problems. Another maintains that troubles arise when development and production activities take place at the same time, that is, when concurrent procurement strategies are adopted.

—

—

Development Objectives

Defense analysts have long suspected that technologically adventurous programs encounter more than their share of acquisition problems 47 As Congressman William Dickinson has argued, "for every ten acquisition programs that have problems, perhaps nine are caused by overly-ambitious 'requirements' that push the technology too hard ." 48 Several members of the military reform movement a loose, bipartisan coalition of defense analysts, former military officers, journalists, congressional staffers, and several dozen members of both houses of Congress have seized on this issue. According to James Fallows, "the distinguishing feature of modern American defense has been the pursuit of the magic weapon." This "wonder-weapon mentality," as he puts it, is based on the misguided belief that the pursuit of high technology leads to high-quality weapons and a sound defense 49 In fact, he and others argue, technologically ambitious and complex weapons generate a wide variety of developmental and operational problems. One problem is the high unit costs of such weapons; the military can afford to buy relatively few of each. There are obvious operational problems associated with shrinking force structures. Congressman Newt .

—

—

.

and Meckling, "Predictability of the Costs," p. 475. Dickinson Letter to Boxer on Concurrency," Aerospace Daily, June

47.

See, for example, Marshall

48.

From "Text

of

11,

1987, p. 408. 49.

James Fallows, National Defense (New York: Random House,

1981), pp. 35, 60-61.

[16]

Introduction

Gingrich summed up the thinking of many military reformers on this point when he said, "high technology limits size and, as Lord Nelson annihilate ." 50 In addition,

complex weapons are said to be less reliable than their simpler counterparts. Even if they are ready for action, complex weapons are difficult to operate, and considerable training is required before they can be used proficiently. They are also more expensive to operate and maintain than simpler weapons.

warned

us,

Finally,

it

is

numbers

said that technologically ambitious

weapons slow down

force structure modernization because they take longer to build than

simpler weapons

What

51 .

needed, Fallows suggests, is the development of "simple, reliable, flexible tools that can be produced quickly ." 52 In other words, national security would be better served by weapon acquisition programs that, on the whole, featured more modest development objecis

tives.

Procurement Strategies

development objectives, one of two basic procurement strategies can be employed: sequential or concurrent. Sequential strategies are based on the assumption that research and development programs are laced with technological uncertainties that are best resolved by proceeding in an orderly, sequential manner. The Given

a set of

traditional stages of the acquisition process

— research, design develop-

ment, engineering or full-scale development, production, and deployment are not compressed, nor are they allowed to overlap. A conscious effort is made to develop systems fully and test them thoroughly before making a decision to proceed with production. Simple prototypes are built and tested during the development phase of the process and, if possible, competing prototypes are tested against one another. Final designs and program schedules are kept flexible while developmental testing takes place. These tests help to identify residual development problems, and they provide decision makers with solid sets of facts on which to base their design, source selection, and production decisions. Funding is kept to a minimum during the development phase of sequen-

—

Gingrich, Congressional Record 127 (Sept. 11, 1981), H6149. For a more scholarly analysis of this issue, see Michael Handel, "Numbers Do Count: The Question of Quality versus Quantity," journal of Strategic Studies 4 (Sept. 1981), 225-260. 51. See Fallows, National Defense, pp. 35-75; Franklin C. Spinney, Defense Facts of Life (Boulder, Colo.: Westview, 1985), pp. 5-111. For a pointed critique of these arguments, see William Perry, "Fallows' Fallacies," International Security 6 (Spring 1982), 174-182. For a review of the military reform debate as a whole, see Asa A. Clark et al., eds.. The Defense 50.

Reform Debate (Baltimore: Johns Hopkins University Press, 1984). 52.

Fallows, National Defense,

p. 29.

,

Flying Blind

programs in order to keep sunk costs and procurement momentum under control. As a result, a bold line is drawn between development and production, and a discrete production decision follows the conclusion of the development phase of the process. 53 Concurrent strategies try to speed up the acquisition process by initiating production activities while development is still under way. Such strategies are based on the assumption that acquisition programs are fundamentally predictable and that residual technological uncertainties can be resolved without relying on prototype testing in the development phase of the process. Instead, paper studies are used to map out and tial

confirm the designs of entire systems, including all attendant subsystems. Once these studies are completed, designs are frozen and highly compressed development and production schedules are drawn up. Production materials and hard production tooling are purchased during the development phase, and the first production units are relied on for hardware testing. In concurrent programs, therefore, hardware testing cannot begin until production lines are up and running. Since it is expensive to support two contractors under these conditions two production lines would have to be set up competition in concurrent programs usually ends prior to full-scale development. Even so, concur-

—

—

programs involve substantial financial commitments early in the development process because of the expenses associated with moving into production. In short, de facto and sometimes explicit production decisions are made during the development phase of concurrent prorent

grams. 54 Concurrent strategies have dominated U.S. weapon acquisition throughout most of the postwar period. This has led many to conclude that concurrency should be deemphasized, if not eschewed altogether,

and that sequential strategies should be strongly favored. A long series of Rand Corporation studies dating back to the late 1950s has taken this position. 55 In 1970, the president's Blue

Ribbon Defense Panel recom-

53. See B. H. Klein, T. K. Glennan, Jr., and G. H. Shubert, The Role of Prototypes in Development Rand Corporation Research Memorandum, RM-3467/1-PR, April 1971; Robert Perry, A Prototype Strategy for Aircraft Development, Rand Corporation Research Memorandum, RM-5597-1-PR, July 1972; G. K. Smith et al.. The Use of Prototypes in Weapon System Development Rand Corporation Report, R-2345-AF, March 1981; Perry, System Acquisition ,

Strategies.

See the sources cited in note 53. See also Perry, System Development Strategies, pp. 101-111; U.S. Congressional Budget Office (CBO), Concurrent Weapons Development and 34.

Aug. 1988. See, for example, B. H. Klein, W. H. Meckling,

Production,

and

G. Mesthene, Military Research and Development Policies, Rand Corporation Report, R-333, Dec. 1958; Klein, Role of Prototypes; Perry, System Acquisition Strategies; Perry, Prototype Strategy; Smith, Use of Prototypes; Michael Rich and Edmund Dews, Improving the Military Acquisition Process, Rand Corporation Report, R-3373-AF/RC, Feb. 1986. 55.

E.

[18]

Introduction

development and production." 56 In the early 1970s, Deputy Secretary of Defense David Packard instituted a "fly before you buy" strategy that was suc-

mended implementing

cessfully applied to

"a general rule against concurrent

some programs. 5/

Packard's reforms withered after

he left the Pentagon, though, and the 1986 Blue Ribbon Commission on Defense Management, which he headed, was forced to sing a familiar refrain: the Department of Defense should "require the testing of prototype systems and subsystems before the authorization of full-scale development." 58 When asked about concurrency in 1987, Les Aspin, chairman of the House Armed Services Committee, said, "I think it's always a mistake." 59

Analyzing the Outcomes of Weapon Acquisition Programs I

argue that

procurement

it is

on

simplistic to focus

either

development objectives or

strategies in analyzing acquisition outcomes; the success of

program is a function of the interaction between its development objectives and the procurement strategy on which it is based. Certain combinations of objectives and strategies are likely to be successful. Others are likely to generate disastrous cost, schedule, and

a

weapon

acquisition

performance outcomes. For the moment, it might be useful

program's development objectives as being either modest or ambitious, although in the real world they might be found anywhere along a continuum. A program with modest objectives relies heavily on off-the-shelf components and proven technologies. It seeks few technological advances. On the other hand, a program with ambitious objectives seeks major advances in the state of the art. Similarly,

it

to think of a

might be useful

to think of procure-

strategies as being either sequential or concurrent, although many programs contain elements of both. These simple distinctions lead us to

ment

the matrix depicted in Figure 1. Sequential strategies are particularly appropriate, even essential, for programs with ambitious development objectives (combination A in FigBlue Ribbon Defense Panel, Report to the President, p. 8. analytic work done at Rand and the recommen57. Packard's approach was based on the dations of the Blue Ribbon Defense Panel. For expositions of Packard's views, see his testimony in Policy Changes in Weapon System Procurement, Hearings before the House Government Operations Committee, 91st Cong., 2d sess., Sept. 1970 PP- ^~4 2 2 &7~3 2 4> Advanced Prototype, Hearings before the Senate Armed Services Committee, 92d Cong., 1st 56.

'

sess.. Sept. 1971, pp. 1-57. grams is analyzed in Smith, 58. p.

The application Use

>

of Packard's strategy to several aircraft pro-

of Prototypes.

Blue Ribbon Commission on Defense Management,

A

Formula for Action, April 1986,

32. 59.

Aspin quoted

York Times,

March

in

John H. Cushman,

4, 1987.

Jr.,

'

Build-Now-Finish-Plan-Later Plan,

Neu

’

Flying Blind

ambitious

A

D

modest

B

C

sequential

concurrent

Development objective

Procurement strategy Figure

ure

1.

Development objectives and procurement

Programs that

strategies

on major technological advances need extensive prototype testing to ensure that all major technological questions are resolved before system designs are frozen and production decisions made. It is important to resolve major technological uncertainties before making commitments to production, and these uncertainties are es1).

rely

pecially formidable in ambitious programs. Sequential strategies provide the best way of reducing developmental uncertainties. In addition, pro-

grams

that are

efforts

fail.

committed to major technological advances might run into snags. Sequential programs are inherently resilient because system designs and acquisition schedules are not firm, and competitive programs provide fallback options in the event that some development keep sunk costs to a minimum until explicit production decisions are made. This is important because some technologically ambitious programs fail and therefore need to be Finally, sequential strategies

canceled.

Sequential strategies

may

not be necessary, however,

when

development objectives are relatively modest (combination B). Extensive hardware testing may not be needed when most of a system's technologies are within the state of the art. Testing already proven hardware is wasteful and time-consuming.

Introduction

Given modest development objectives, concurrent strategies may be optimal (combination C). Programs that do not involve major technological advances may not need extensive prototype testing. he final I

designs of these kinds of systems may indeed be fairly predictable, so it is not unreasonable to begin buying production materials and tooling

during the development phase of the acquisition cycle. For programs with modest objectives, concurrent strategies may deliver operational systems quickly and inexpensively. Concurrent strategies are particularly inappropriate, however, for programs with ambitious development objectives (combination D). Programs with new and exotic technologies probably need extensive testing before system designs can be frozen. More important, it is risky to make production commitments to these kinds of programs early in the development process. Under concurrency, however, one has to begin buying production tooling and materials before prototype testing can take place. As a result, considerable investments have to be made in highly speculative ventures. This combination of objectives and strategies is consequently very risky. Contrary to what some

reform movement have argued, weapons without incurit is possible to build technologically ambitious ring cost overruns, schedule slippages, and performance shortfalls. "High technology" programs are not doomed to fail. The key is to employ sequential strategies in these cases. It is also important to note that, contrary to what many analysts have argued, sequential and concurrent strategies are,

optimal

some

in the military

by themselves, neither good nor bad. Each strategy of the time; neither

is

optimal

all

is

of the time.

think of development objectives and strategies as falling along two continua, our simple matrix can be replaced by Figure 2. Programs If

we

with appropriate combinations of objectives and strategies are likely to have optimal outcomes. Other combinations tend to produce less satisfactory outcomes: some are wasteful, others are risky. elaborate set of definitions is needed to make this framework useful. The nine levels of technological ambitiousness defined in Table 1 60 enable us to categorize development objectives with some precision

A more

.

To categorize procurement strategies

we

can use the eight

criteria out-

highly sequential programs exhibit none of the earmarks 61 Given of concurrency; highly concurrent programs exhibit all eight these two sets of variables, we can categorize programs according to the lined in Table

2:

.

For more discussion of the methodological issues associated with this kind of categoSubjective Assessment of rization, see S. James Press and Alvin J. Harman, Methodology for April 1975. Technological Advancement Rand Corporation Report, R-1375, are not rank-ordered. All strategies 61. The eight criteria for categorizing procurement 60.

,

are considered to be equally important.

ambitious

Risk

Development

Optimal

objective

Waste

modest concurrent

sequential

Procurement strategy

Figure

2.

Development

objectives,

procurement

9X9 matrix depicted in Figure 3 and,

strategies,

from

that,

and program outcomes

make some

predictions

about program outcomes. For example. Program X (Figure 3) involves some technological advances and is based on a procurement strategy that contains both sequential and concurrent elements. Its procurement strategy is well suited to its level of technological risk, so it will probably be relatively successful in meeting the other hand.

Table 1.

2.

3. 4. 5. 6.

1.

New New New

its

and performance technologically ambitious and

cost, schedule,

Program Y

is

it

On

features a

Levels of technological ambitiousness

and

radically different system design is needed. technology must be developed to meet system needs. technology must be developed to meet subsystem needs. Several major subsystems require major improvement. One major subsystem requires major improvement. Several subsystems require some improvement.

One subsystem

8.

some improvement. Contemporary' technologies must be integrated

9.

No new

7.

targets.

requires

technology or hardware

is

needed; off-the-shelf components can be

Adapted from Robert Perry et al.. Systems Report, R-733-PR/ARPA, June 1971, pp. 10-14. Source:

into the program.

Acquisition Strategies,

utilized.

Rand Corporation

[22]

Introduction

Table 2. Criteria for categorizing

procurement

strategies

Sequential strategy

Concurrent strategy

System design

Flexible during develop-

Frozen as soon as possi-

Military subsystems

ment phase Not initially integrated

Fully integrated into de-

into system design

sign from beginning

Element

Source of

critical

Paper studies

Prototype testing

pro-

ble

gram information Initial

tooling

(first

development tooling (development pro-

pro-

Soft

totype)

Hard production tooling (production prototype)

totype)

Development and pro-

Open-ended and

duction schedules Competition

Preserved through

Funding during development phase Production decision

flexible

full-

Preplanned and

Ended

at

design phase

scale development Minimal

Substantial

Discrete decision follows

De

full-scale

rigid

facto decision

made

during development

development

highly concurrent procurement strategy. This program will probably encounter serious problems. Program success and failure also need to be defined more precisely. My focus is on the success of programs as acquisition programs, and so I

compare the

actual cost, schedule,

62 question to their original targets.

and performance It is

of the systems in

also important to consider

how

resolving the technological uncertainties originally associated with them. Therefore, I also consider the amount of that is, the extent to which retrofitting associated with each program

effective

programs were

in

—

critical

subsystems had

main system left the which extensive modifications had to be made

to

be installed

after the

and the extent to to bring the weapon up to its design specifications. The set of cases examine in this book is well suited factory

I

to a

study of

contains significant variation across all relevant variables. Some of these programs were extremely ambitious technologically (e.g., B-35, B-47, B-58, B-70, B-2), while others were only moderately ambitious (B-36, B-52, B-i, B-iB). Some used highly concur-

acquisition outcomes;

rent

procurement

it

strategies (B-58, B-70, B-iB, B-2), while

highly sequential strategies,

at least in their early

two featured

stages (B-47, B " 5 2 )-

many Rand Corporation

studies over the years, including Perry, System Acquisition Experience; Dews, Acquisition Policy Effectiveness. For a more detailed discussion of how such assessments can be made, see Alvin J. Harman and 62. This

approach has been used

in

Susan Henrichsen, A Methodology for Cost Factor Comparison and Prediction, Rand Corporation Research Memorandum, RM-6269-ARPA, Aug. 1970.; G. K. Smith and E. T. Friedmann, An Analysis of Weapon System Acquisition Intervals, Past and Present, Rand Corporation Report, R-2605-DR&E/AF, Nov. 1980.

ambitious 1

2

Y

Risk

3 4

Development

Optimal

X

^

objective 3 6 7

Waste

8 9

01

modest

2345678 concurrent

sequential

Procurement As defined in Table 1. b Categorized according to the

strategy b

a

Table

of concurrent elements exhibited, as defined in

Development objectives, procurement framework for analysis

Figure a

number

2.

3.

Finally,

some programs were

strategies,

and program outcomes:

quite successful in meeting their cost,

schedule, and performance targets, while others were unmitigated di-

The use of case studies enables us to see how development objectives and procurement strategies interacted to bring these outcomes about. One could not develop this insight through a simple statissasters.

tical

A

analysis of a set of programs. close examination of detailed case studies also allows us to address

what

development objectives and procurement strategies? One of my main arguments is that a powerful set of strategic and bureaucratic forces pushed these programs in the direction of extremely ambitious development objectives and highly concurrent procurement strategies simultaneously. As a result, many of these efforts had serious problems. This pattern is widespread in the United States, and it has proved resistant to repeated reform efforts, suggesting that the forces that shape acquisition programs are powerful indeed. Although develop these arguments in detail in later chapters, I can outline them here. a critical question:

led to the adoption of these

I

Many

U.S. Air Force leaders believed that technology

warfare and that

it

was consequently important

to

was

push the

decisive in state of the [24]

Introduction

weapon development programs. Many also believed that technological innovation was the United States' area of comparative advan-

art in

tage in the strategic competition with the Soviet Union. In addition, it was widely held that defense planners should make pessimistic assump-

mance requirements

for

new

when

they established perforweapons. These strategic considerations

tions about future operational threats

were strongly reinforced by the Air Force's bombardment doctrine, which placed enormous emphasis on highly capable aircraft. Air Force thinking dictated that its bombers have exceptional range, speed, altitude, payload, and defensive armament capabilities. Air Force decision makers were able to push several performance parameters at once and ignore the trade-offs inherent in their requirements by assuming that technological advances would eventually save the day. They were encouraged in this by the technical commands in the Air Force and by the various aerospace companies involved in these projects, none of which had a vested interest in discouraging new programs, and all of which were infected with a "can do" attitude toward technological challenges such as these. Finally, the Air Force needed to sell its new programs to the civilian leadership in the executive branch and Congress. Thus, its new programs had to promise a substantial improvement in capabilities over existing systems. In short, mutually reinforcing strategic and bureaucratic forces pushed Air Force performance requirements beyond the state of the art, and the Air Force's bomber programs frequently had extremely ambitious development objectives. At the same time, there seemed to be a compelling need to get new into the force structure as soon as possible. This sense of urgency, justified or not, loomed over American weapon acquisition efforts for most of the postwar period. As a result, there seemed to be a

bombers

legitimate strategic reason for adopting concurrent procurement strategies, which held out the promise of deploying weapons quickly. The

Air Force also preferred concurrency for internal reasons. First and foremost, concurrent programs were difficult to cancel because they moved into production quickly

and because

a great deal of

money was spent on

early in the acquisition process. Concurrent programs were also easier to sell to the civilian leadership in the defense establishment be-

them

cause they appeared to be cheaper than their sequential counterparts. Finally, concurrency enabled the Air Force to sidestep a painful trade-off: its technologically ambitious bombers would take a long time to develop

compress the acquisition cycle. Why did the civilian leadership allow the Air Force to pursue these policy preferences? For one thing, civilians rarely became involved in the

and build unless something was done

to

requirements formation process because it was part of the military s professional domain. In any event, many civilians also believed that it

Flying Blind

push technology for reasons of national security. As far as procurement strategies were concerned, many civilians joined military leaders in believing that it was important to deploy new strategic

was important

to

systems without delay. Others, such as Secretary of Defense Robert McNamara, believed that paper studies could substitute for prototype testing. It is not surprising that concurrency flourished in the Pentagon under McNamara's stewardship in the 1960s. There have, however, been three discrete periods since 1945 when some programs have been guided by sequential strategies: in the late 1940s, when budgetary constraints prevented the services from pushing

programs into production; in the early 1970s, when Deputy Secretary of Defense Packard championed the cause of prototyping and austere, sequential development; and in the late 1980s, when some programs were reoriented in the wake of the Packard commission's strong endorsement of sequential strategies. 63 In each case, there were special circumstances surrounding the employment of sequential procurement strategies. In the first case, severe budgetary constraints impinged on the acquisition process, and, in the last two cases, direct, high-level civilian intervention accounted for whatever policy shifts took place. Concurrency has been the dominant strategy during the rest of the postwar period. One of the main arguments of this book is that military organizations prefer concurrent procurement strategies for a variety of strategic and bureaucratic reasons. Moreover, military organizations generally but not always are left to themselves to pursue these preftheir

—

—

erences. 64

The Air Force was

flying blind

when

it

repeatedly issued performance

requirements that demanded major technological advances. The feasibility of these requirements was not adequately assessed in most cases. Totally unforeseen technological breakthroughs had to take place before the requirements could be met. The Air Force was not being bold and farsighted

when

it

set these extraordinarily

ambitious development ob-

Air Force

was

also flying blind

when

it

was

The imposed concurrency on its

jectives; its disregard of technological uncertainties

reckless.

development and production programs. Concurrency's arbitrary assumptions about technological feasibility were not examined closely more discussion on the use of sequential strategies in earlier eras, see Smith, Use of Prototypes, pp. 1-4. The Air Force's Advanced Tactical Fighter program and the Navy's Advanced Tactical Aircraft program were restructured to include more competitive prototyping after the Packard commission's recommendations came out in 1986; see James 63.

For

Kitfield, "ATF: Playing by New Rules," Military Logistics Forum, Sept. 1986, pp. 13-18. I discuss this in more detail in Chapter 9. 64. Posen analyzes organizational preferences toward certain kinds of operational doctrines as well as the conditions under which civilian intervention is likely to intrude on these policy preferences; see Sources of Military Doctrine, chaps. 2 and 7.

[26]

Introduction

when

was applied to technologically ambitious development programs. The lack of developmental testing and the compressed development and production schedules that were part and parcel of concurrency only compounded the risks that these programs faced. It is not it

surprising, then, that concurrency proved to be badly counterproductive in these cases.

Chapter 2 provides general background on the evolution of American air forces in the first half of this century, in particular, the development

bombardment doctrine and strategic bomber forces years leading up to World War II. The fifteen programs examined of strategic

in the in this

book are organized into six case studies (Chapters 3-8). Snapshots of these programs are provided in Table 3 and Figure 4. Chapter 3 focuses on the medium-range jet bomber competition of the late 1940s, which included the B-45, B-46, B-47, and B-48. The B-47 won this competition and was produced in great numbers in the 1950s. Chapter 4 is an analysis of the B-35, B-36, B-49, B-52, and B-60 programs, the long-range bomber programs of the 1940s and early 1950s. Chapter 5 covers the

Table 3

.

U.S. strategic

Program

Ended b

1941 1941

1944 1954 d 1 94^ 1948

1944 1944 1944 1944 1945 1946

B -49 B-52 B-58 B-60

1951

A

1950 1951 1953 1970

B-iB

1981

B-2

1981

f

B-70 B-i

acquisition

Began 3

B-35 B-36 B-45 B-46 B-47 B-48

ANP

bomber

programs since 1941

Number

Contractor

built

1952 1956 1962

2

Northrop Convair North American Convair Boeing Martin Northrop Boeing Convair e Convair

0

several

2

1977 1988

100

North American North American Rockwells Rockwell Northrop

15c

366 142 1

1957 1948

U923

1949 1962 1962

3

—

2

744 116

4

75

h

major development contract. of production cancellation or end of production run. c Three were converted into B-49S. d The decision to halt the B-45 production run was made in 1948, although production actually continued into the 1950s. c Convair became the Fort Worth Division of General Dynamics in 1961. Aircraft Nuclear Propulsion (nuclear-powered bomber) program. sNorth American Aviation merged with the Rockwell Standard Corporation in 1967The new company was known as North American Rockwell until 1973' when its name was a

Date of

first

b Date

f

changed to Rockwell International. b Planned production run.

Flying Blind

Figure

4.

U.S. strategic

bomber

acquisition

programs since 1941

and evolution of the B-58, the supersonic successor to the B-47. Chapter 6 examines the nuclear-powered bomber program, also known as the aircraft nuclear propulsion (ANP) program, as well as the B-70. Both were to be long-range bombers, but the ANP program was canceled in the mid-1950s and the B-70 was canceled in the early 1960s. origins

Chapter 7 is an analysis of the B-i bomber, now known as the B-iA, from its conception in 1961 until its cancellation in 1977 by President Carter. Chapter 8 presents a look at the B-iB, a modified version of the B-i resurrected by the Reagan administration in 1981, and the B-2, which moved quickly from development to production in the 1980s. I conclude in Chapter 9 with some general observations about the origins and outcomes of weapon acquisition programs and the implications of this study for weapon acquisition reform. An appendix provides a brief overview of the evolution of U.S. Air Force weapon acquisition organizations since the early 1940s. [28]

Historical , Organizational,

and Doctrinal Setting

People thought about dropping

bombs from

the sky long before there

were airplanes. In 1670, the Italian Jesuit Francesco Lana speculated that airships might one day be able to bomb targets with impunity, destroying buildings and capsizing ships in the process. He raised this possibility over one hundred years before the first manned balloon flight, which took place in 1783 outside Versailles. Austria was the first country to use hot-air balloons as bombing platforms; the results of its attack on Venice in 1849 were uneven at best. To each of approximately two hundred unmanned balloons Austrian soldiers attached one time-fused bomb. Many of the balloon bombs went astray, some blew back over the Austrian lines, and those that reached the target did little damage. Balloons were used fairly extensively in the early years of the American Civil War, mainly for scouting enemy troop movements, but they were almost completely neglected by the U.S. military for the remainder of the century. The situation was quite different in Europe at this time, where competitive pressures were substantially keener. Each of the great powers in Europe established a balloon corps and was actively engaged in balloon research in the latter third of the nineteenth century. Germany and France, most notably, were starting to build steerable, rigid airships, or dirigibles, as the century came to a close. There was 1

already a great deal of interest in military aviation in

some

quarters,

summary and more discussion on developments prior to 1900, see Basil Collier, A History of Air Power (New York: Macmillan, 1974), pp. 1-42; Charles DeForest Chandler and Frank D. Lahm, How Our Army Greui Wings (New York: Ronald, 1943), chaps. 1.

For this

Flying Blind

when

and Wilbur Wright made the first powered flight in a heavier-than-air craft near Kitty Hawk, North Carolina, on December 17, 1903. therefore,

Orville

Early Years

The U.S. Army, consistent if nothing else, showed little early interest in airplanes, even though the Wright brothers made over one hundred successful flights in 1904 and approached the Army for funding on three separate occasions in 1905. The only step the Army took to get involved in aviation was to create the Aeronautical Division in the Signal Corps in August 1907. Although the new division's charter was "to study the flying machine and the possibility of adapting it to military purposes," it was not given any money to fund experimental studies. 2 Moreover, the new Aeronautical Division had a staff of only three, one officer and two enlisted men, one of whom deserted shortly after receiving his new assignment. Funding for airplane projects was not forthcoming until the Aero Club of America brought the Army's apathy to the attention of President Theodore Roosevelt, who personally intervened in the case and directed the Army to begin investigating aeronautical activities more vigorously. This led the Army to issue a request for aircraft proposals in December 1907, which in turn led to a $25,000 contract with the February 1908. 3 Even so, technical progress continued to be slow because, year after year. Congress refused to set aside funds specifically for aeronautical development. There seemed to be no compelling strategic rationale for a major research and development program in this area, since the Atlantic

Wright brothers

in

and Pacific oceans protected the continental United States from European and Asian air forces, and Congress had no intention of getting involved in a war in Europe, should one break out. Army aviation programs consequently had to be financed out of the department-wide experimental fund, where they had to compete with more conventional programs put forward by powerful bureaucratic sponsors. Only in 1911 did Congress, under growing public pressure, appropriate $125,000 for

2.

Quoted

Army

in

aviation. This enabled the Aeronautical Division to

George

E.

Stratemeyer, "Administrative History of the U.S.

Forces/' Air Affairs 1 (Summer 1947), 510. 3. The most detailed account of this episode

Grew Wings, chap.

is in

Army

Air

Chandler and Lahm, Hoiv Our Army

10.

[30]

4

Historical Organizational, and Doctrinal Setting

establish a flying school

and gradually increase

its

number

of certified

pilots.

Over the next few

years.

Army

aviators

conducted experiments

to

develop the military value of the airplane. Many of these experiments, such as testing bombsights and bomb-dropping devices, were designed to

enhance the

airplane's potential as a

bombing

platform.

From the means of

beginning. Army aviators saw the airplane "not only as a observation and liaison, but as a striking arm against forces in the field ^ and supporting facilities to the rear." The meager appropriations set aside for military aviation nonetheless left the United States far behind every major European power, all of which made a concerted effort to develop aviation forces. France, for

example, spent $7.4 million on aviation in 1913 alone; Germany and Russia spent $5 million each. The United States, on the other hand, set aside only $125,000 for aviation in 1913, less than one-third of what Mexico devoted to the same activity. Inevitably, American air forces were smaller and less advanced than those of nearly every other great power. France, for example, fielded 260 military aircraft on the eve of the Great War, whereas the United States had only six serviceable aircraft in its

inventory

Some

in

at the time. 6

Congress

felt

that

new

organizational arrangements were

needed to facilitate the development of American air forces. Two bills were proposed in 1913, both of which would have taken aviation activities away from the Signal Corps but still left them under the supervision of the Army. The secretary of war opposed these bills on the nothing but another branch of the service of information, which includes all communication, observa7 Clearly, the Army establishment had an tion, and reconnaissance." extremely narrow view of how air forces should be used. Many pilots also opposed these bills, but on the grounds that aviation was not yet

grounds

that "the aviation service ...

is

developed enough to become a separate service. Younger pilots, however, were more inclined to create a separate service on the spot and

Arthur Sweetser, The American Air Service (New York: Appleton, Weapons (New Haven: Yale University Press, 1919), pp. 10-12; Irving B. Holley, Jr., Ideas and History of the United States Air Force, 1907-1957 1953), pp. 27-28; Alfred Goldberg, ed., A (Princeton: Van Nostrand, 1957), pp. 3-6. in the Army Air Arm, 1917-1941, 5. Thomas H. Greer, The Development of Air Doctrine 4.

See

ibid. Also,

Air University, Dec. 1957, p. 3. 6. See Sweetser, American Air Service, p. 16; Holley, Ideas and Weapons, p. 29. History," p. 511. The War Department 7. Quoted in Stratemeyer, "Administrative became the Department of the Army when the U.S. military establishment was re-

organized

in 1947.

Flying Blind

thus escape superiors who knew end, both bills were defeated.

"little

or nothing" about flying. 8 In the

was, however, widely recognized that military aviation had to be given a more permanent footing and a higher profile. As it stood, the Aeronautical Division was simply the product of an administrative It

order; theoretically,

do

so.

The Army

it

Army chose to change manpower levels

could be dissolved whenever the

certainly

had the authority

to

along with the division's lowly organizational status and the lack of proper compensation for extremely hazardous duty, made recruitment difficult and morale poor. With this in mind. Congress decided to give Army aviation a statutory basis. It authorized personnel at will. This,

slots for 60 officers

and 260

enlisted

men, and

it

raised

pay

levels for

aviators. Legislation creating the Aviation Section of the Signal

was enacted on

July 18, 1914,

Corps ten days before Austria declared war on

Serbia. 9

American Air Forces

in

World War

I

The United States did little to build its military aviation forces as the war in Europe unfolded. Budgets remained low, and acquisition plans remained modest. The only significant American step during the early months of the war was the creation in March 1915 of the National Advisory Committee for Aeronautics (NACA) to sponsor and coordinate 10. scientific research on the problems of flight. NACA was given an annual budget of only $5,000. 10 8.

See Goldberg, History

of the United States Air Force,

Stratemeyer, “AdminChase C. Mooney and Martha E. Layman, Organization of Military Aeronautics, 1907-1935, Historical Study No. 25, Historical Division, Asst. Chief of Air Staff for Intelligence, Dec. 1944, PP- 10-1 7; R. Earl McClendon, The Question of Autonomy for the United States Air Arm, 1907-1945, Air University Documentary Research Study, Nov. 1950, PP- 27-42. 9. See Stratemeyer, “Administrative History," pp. 511-512; Goldberg, History of the United States Air Force, p. 8; Sweetser, American Air Service, pp. 20-21. In addition, American aviators were very much in the dark about how the air war in Europe was progressing. The warring powers were all highly secretive about the performance capabilities of their aircraft and their operational experiences. When asked by the House Military Affairs Committee if the Army's Aviation Section was keeping up on developments in the war, one colonel replied, "I think we are, as far as it is possible to say that we are keeping abreast of conditions that we do not know anything about"; quoted in Sweetser, American Air Service, p. 26. Discussion of the 1915-18 period is based on ibid., chaps. 2, 12; Holley, Ideas and Weapons, chaps. 2-4, 7-8; Goldberg, History of the United States Air Force, chap. 2; James L. Cate, "The Air Service in World War I," in Wesley F. Craven and James L. Cate, eds.. The Army Air Forces in World War II, 7 vols. (Chicago: University of Chicago Press, 1948), vol. 1, pp. 3-16; Historical Section, Army War College, The Signal Corps and Air Service (Washington: Government Printing Office, 1922), pt. 2; Mooney and Layman, Organization of Military Aeronautics, pp. 26-36. istrative History/' p. 512;

p.

8;

Historical, Organizational,

and Doctrinal Setting

the international situation deteriorated, policy makers in Washington became increasingly concerned about the level of military pre-

As

paredness

in the

United States.

for military aviation

An emergency appropriation

was approved

million appropriation in August.

in

March

1916, followed

Germany began

of $500,000

by

a $13.3

unrestricted sub-

marine warfare in February 1917, which led to American entry into the war in April. At the time, there were only two hundred aircraft in the Army's inventory. Most of these were trainers, and all were technologically obsolete.

Many Americans impact U.S.

air

entered the war with naive expectations about the forces would have on the course of the conflict. Many

believed that, once the country's industrial capacities were engaged, "clouds" of American airplanes would break the gruesome stalemate on the European battlefield and bring the war to a quick conclusion. After

some

debate,

it

was announced

that the United States

planned

to build

22,625 airplanes in the first six months of 1918. This was ambitious, to say the least, given that the American aircraft industry had produced

fewer than one thousand airplanes, civilian and military, between 1903 and 1916. Congress nonetheless put this plan in motion in August 1917 with an appropriation of $640 million, the largest made to that time for a single purpose. This plan was doomed to fail. New aircraft designs were needed in order to catch up with the many technical advances that had been made in Europe since 1914. Developing new designs would take months, at a minimum. European designs could be used, but even then it would take time before dozens of production lines could be set up. In any event, the Aviation Section was not experienced enough to supervise a program of this magnitude. In short, there was no way to overcome a decade of neglect in less than one year.

When

it

became

clear in early 1918 that

production output would fall far short of expectations, widespread disillusionment set in and, inevitably, various investigations into the matter were launched. 11 This state of affairs led President Woodrow Wilson to reorganize the Army's aviation activities in May 1918. Aviation was taken away from the Signal Corps and placed under the general jurisdiction of the War Department. The Bureau of Aircraft Production was given responsibility for acquisition activities,

and the Division

was would be

of Military Aeronautics

placed in charge of operational matters. These two offices overseen by the director of the Air Service, who would also be the Only 11,000 aircraft were built in the United States during the war, and the vast majority of these were trainers and reconnaissance aircraft. Most of the bombers built in the United States during the war were of British design, and the first American bomber was still in development when the war came to an end. 11.

Flying Blind

assistant secretary of tion

war

for aviation.

would eliminate some

It

was hoped

that this reorganiza-

of the institutional causes of the

bottleneck as well as head off calls for a

procurement completely independent air

force.

The net effect of these delays was that American aviators saw little combat before the spring of 1918. When American air forces finally did swing into action, their operations were constrained by the equipment they had on hand and by existing command structures, both of which were shaped by prevailing organizational arrangements. The main mission of the Signal Corps, of which the Aviation Section was a part until mid-1918, was gathering information about enemy movements and passing this information along to Army commanders on the ground. Predictably, the Signal Corps saw the airplane, first and foremost, as a new and improved way of fulfilling this mission. As a result,

when

the United States entered the war, 89 percent of the

Army

procurement program was devoted to observation aircraft and pursuit aircraft (now known as fighters), which the Signal Corps believed would be used mainly to protect reconnaissance operations. 12 American advocates of bombardment operations were thus constrained by the fact that few bombers came off American production lines and, of those that were fielded during the war, none had enough range to fly deep into German territory. American advocates of bombardment operations were also constrained by the command arrangements under which they operated. Air squadrons were considered to be integral components of ground units, and final decisions about how air forces were used were made by the commanders of these units. 13 These commanders preferred to use the air forces at their disposal for reconnaissance, air defense, and close air support of ground operations rather than for strikes against distant military and industrial targets. Maj. William Mitchell, who became one of the country's most forceful advocates of strategic bombardment, recommended that American air forces in Europe be split into two main groups. He suggested that some units be assigned to ground forces, as they had always been, but also that some units be set aside for independent, strategic operations against enemy aircraft and materiel. 14 Mitchell's plan to set up an independent striking force was rejected by Gen. John Pershing, commander of the American Expeditionary Forces aircraft

Holley, Ideas and Weapons, p. 134. 13. Greer, Development of Air Doctrine, pp. 4-5. 14. Maj. William Mitchell, Memorandum for the Chief of Staff, U.S. Expeditionary Force, June 13, 1917, in Maurer Maurer, ed.. The U.S. Air Service in World War l, 4 vols. (Washington: Government Printing Office, 1978-1979), vol. 2, p. 111. 12.

[

34

]

Historical, Organizational,

in

Europe, and American

and Doctrinal Setting

air forces

consequently engaged

in

few sus-

bombing campaigns during the war. The most notable exception to this pattern took place in the final three months of the war, when Mitchell was put in charge of bombing campaigns against German forces at St. Mihiel and Meuse- Argonne. Several raids were carried out in September, October, and November 1918. The

tained

which were assigned to ground commanders and two-thirds of which were under Mitchell's direct control. Although they sustained heavy losses, these large formations dominated the airspace over the battlefield and were able to execute their attacks on German ground forces, air fields, and communications posts. It is important to note that these attacks were not directed at the German economic infrastructure or other targets far behind the front lines; that is, they were not "strategic" campaigns in the sense of the term used in later years. They were tactical operations 15 In any event, designed to support army operations on the ground. they contributed significantly to the success of the Allied offensive, and largest of these involved 1,500 aircraft, only one-third of

they reinforced air power advocates' belief in the effectiveness of independently organized, highly concentrated bombardment operations. Had the war continued beyond November 1918, the United States

undoubtedly would have built more and better bombers, and American air forces in Europe undoubtedly would have engaged in more of the sustained bombing campaigns that were successful in the last few months of the conflict. Indeed, American aviators planned to attack German manufacturing and transportation centers in order to break supply lines to the German front. The few bombardment campaigns that took place during the war, however, were purely tactical in nature, and American bombers penetrated no more than 160 miles behind German lines

during these operations. 16

American Air Power

in

the Interwar Years

The development of American air power in the interwar years revolved around three related issues: the crusade for bureaucratic autonomy waged by Army aviators; their efforts to develop and promote a distinct strategic bombardment doctrine; and their campaign to deploy 17 These issues were closely relatlarge numbers of long-range bombers. See Cate, "Air Service," pp. 10-16; Greer, Development of Air Doctrine, pp. 4-7; Goldberg, History of the United States Air Force, pp. 23-27. "Air Service, p. 15. 16. Greer, Development of Air Doctrine, p. 11; Cate, Craven, "The Army Air Arm between Two World Wars, 17. James L. Cate and Wesley F. Air Forces in World War II, vol. 1, p. 17. 1919 - 1939 /' in Craven and Cate, eds.. Army 15.

Flying Blind

ed because, although air power advocates genuinely believed that the United States needed strategic bomber capabilities to defend itself, their case for bureaucratic autonomy rested on the claim that strategic bombardment was a distinct military mission that advanced bombers could carry out successfully. Bombardment doctrine and bombers were seen as strategic ends unto themselves, but they were also means to a parochial bureaucratic end. 18 Accordingly,

posed

the military establishment op-

growing band of bomber advocates for both strategic and bureaucratic reasons. The Army and Navy genuinely believed, contrary to the claims of radical air power theorists, that ground and naval forces were still needed to provide for the common defense, but they also recognized that the arguments made by bomber advocates challenged their own claims on the scarce budgetary resources of the interwar this

period.

As

the interwar period progressed.

Army

were gradually given more bureaucratic autonomy. This came about through a series of small steps, each of which was intensely debated by the American military establishment. 19 In addition, strategic

refined in this period,

and by the

aviators

bombardment doctrine was

early 1930s a highly stylized concep-

bombardment operations had begun to emerge from the Corps Tactical School. 20 Although this doctrine was not embraced by

tion of strategic

Air

18.

Perry

Hopkins

McCoy

Smith, The Air Force Plans for Peace, 1943-7945 (Baltimore: Johns

Press, 1970),

p.

17.

My

discussion of the struggle for autonomy waged by air power advocates and the changing institutional relationship between the Army and its aviators in the interwar period is based on McClendon, Question of Autonomy; Smith, Air Force Plans for Peace, chap. 2; Harry Howe Ransom, “The Politics of Air Power: A Comparative Analysis," Public Policy 87-119; Cate and Craven, "Army Air Arm," pp. 17-71; Mooney and Layman, ( 1 95^)/ Organization of Military Aeronautics, 1907-1935; Chase C. Mooney, Organization Military 19.

of

Aeronautics,

1935-1945, Historical Office, Headquarters Army" Air Forces, April 1946; Maurer Maurer, Aviation in the U.S. Army, 1919-1939 (Washington: Office of Air Force History, 1987); Goldberg, History of the United States Air Force, chap. 3; Robert W. Krauskopf,

“The Army and the Strategic Bomber, 1930-1939," Military Affairs 22 (Summer 1958), 8394, and 22 (Winter 1958-59), 208-215. 20. The classic treatise on the uses of air power is Giulio Douhet, The Command of the Air

(New

York: Coward-McCann, 1942), originally published in Italy in 1921. The evolution of William Mitchell's highly influential thinking can be tracked in his Our Air Force (New York: Dutton, 1921); Winged Defense (New York: Putnam's, 1925); and Skyioays (Philadelphia:

Lippincott, 193°)-

The best scholarly studies on the development

bombardment doctrine

of

American

strategic

interwar period are Greer, Development of Air Doctrine; Robert T. Finney, History of the Air Corps Tactical School, 1920-1940, Research Studies Institute, Air University, March 1955; Robert F. Futrell, Ideas, Concepts, Doctrine: A History of Basic Thinking in the United States Air Force, 1907-1964, 2 vols. Aerospace Studies Institute, Air University, June 1971,; Holley, Ideas and Weapons, chap. 10; James L. Cate, “Development of Air Doctrine," Air University Quarterly Review 1 (Winter 1947), 11-22; Cate and Craven, "Army Air Arm," pp. 33-54; Bernard Brodie, Strategy in the Missile Age (Princeton: Princeton University Press, 1959), chap. 3; Russell F. Weigley, The American Way of War (Bloomington: Indiana University Press, 1973), chap. 11; Edward Warner, "Douhet, Mitchell, Seversky: Theories of Air Warfare," in Edward M. Earle, ed.. Makers of Modern Strategy (Princeton: in the

[

36

]

Historical, Organizational,

and Doctrinal Setting

—

—

American military establishment as a whole far from it it provided air power advocates with the conceptual framework they needed to formulate more specific performance requirements in their bomber the

acquisition programs.

mid-1950s.

aviators

was not began

a coincidence that,

to play a

much more

starting in the

assertive role in

and shaping these programs. They soon developed

initiating

that

Army

It

met

their doctrinal requirements, but

even

a

bomber

in the late 1930s

they 21

had serious difficulty in securing funding for their acquisition efforts. Although they came a long way over the course of the interwar period, bomber enthusiasts still found themselves in the U.S. Army when the United States entered World War II in late 1941. The interwar period can be usefully divided into four phases, each capped by a formal change in the relationship between the Army and its aviators. The first phase ended with the passage of the Army Reorganization Act of 1920, which established the Air Service on a statutory basis. The second concluded with the passage of the Air Corps Act of 1926, which transformed the Air Service into the somewhat more autonomous Army Air Corps. The third ended with the establishment of an independent striking Force, in 1935.

ment

force,

known

The fourth and

of the semi-independent

final

as the General Headquarters Air

phase concluded with the establish-

Army

Air Forces in 1941. 22

Lessons from the War: 1918-1920

ended before the efficacy of strategic bombardment could be put to a critical test. The American Air Service had no direct experience with strategic bombing, and, although British, French, and German air forces engaged in some attacks on industrial and urban centers, these attacks were scattered and did not have a decisive impact on the outcome of the war. The lack of clear-cut evidence about the military value World War

I

Princeton University Press, 1943), pp. 485-503; Ronald Schaffer, Wings of judgment (New York: Oxford University Press, 1985), chap. 2; Michael S. Sherry, The Rise of American Air Power (New Haven: Yale University Press, 1987) chaps. 2-3; George Quester, Deterrence Force Plans for Peace, chap. 3. before Hiroshima (New York: Wiley, 1966), chap. 8; Smith, Air interwar period are Mary R. in the programs 21 The best sources on American bomber Bombardment Aircraft, 1917Self, History of the Development and Production of USAF Heavy Historical Office, Air Materiel Command (HO/AMC), Dec. 1950; Jean H. DuBuque .

1949,

Gleckner, Development of the Heavy Bomber, 1918-1944, Historical Division, Aviation Air University, Aug. 1951; Cate and Craven, "Army Air Arm,"pp. 54-7 1 Maurer, Aircraft of World War II," in in the U.S. Army, chaps. 8, 12, 19; Alfred Goldberg, "AAF Craven and Cate, eds., Army Air Forces in World War 11, vol. 4, pp. 193-227; Goldberg, — Peter M. History of the United States Air Force, chaps. 3 4 Gordon Swanborough and Bowers, United States Military Aircraft since 1908 (London: Putnam, 1963). Army Air 22. This chronological framework was suggested by Cate and Craven,

and Robert

F.

;

'

Arm,"

p. 23.

(3 7]

Flying Blind

bombardment allowed bomber enthusiasts and Army traditionalists to believe what they wanted to believe about the bomber's ability to transform warfare. As one study correctly observed, air power's potential was "sensed rather than tested" in 1918. 23 Bomber enthusiasts believed that strategic bombardment would have played a significant role in the war had hostilities continued into 1919. They also believed that it would be decisive in future conflicts. They of strategic

argued that, since infantry and artillery officers were mainly concerned about the tactical situation on the battlefield, bombardment operations would have to be organized by airmen if they were to be strategic in character and as effective as they could be. Some suggested that air forces be administered separately as part of a new Department of Defense. Others

went so

far as to

recommend

establishing a cabinet-level

department of aeronautics equal in stature to the War and Navy departments and comparable to Britain's Royal Air Force (RAF), which had been established as an independent service in January 1918. 24 Army traditionalists held fundamentally different views. Secretary of War Newton Baker believed that the just-concluded bombardment campaign had "no appreciable effect" on Germany's war-making capacity or, indeed, on the outcome of the war itself. 25 General Pershing, the Army's field commander during the war, argued that "an air force acting independently can of its own account neither win a war at the present time nor, so far as we can tell, at any time in the future." He believed that the air force's main mission should be "to drive off hostile airplanes and procure for the infantry and artillery information concerning the enemy's movements." 26 Predictably, Army traditionalists strongly opposed the idea of giving the Air Service any more operational or administrative autonomy. A total of eight bills that would have established an independent department of aeronautics were introduced in Congress in 1919-1920, and three panels of experts were convened to review the institutional relationship between the Army and its aviators. Two of the panels, the Menoher board and the Dickman board, supported the status quo. The Cate and Craven, "Army Air Arm," p. 34. McClendon, Question of Autonomy, chap. 4; Cate and Craven, "Army Air Arm," pp. 17-24; Mooney and Layman, Organization of Military Aeronautics, chap. 2; Goldberg, 23.

24.

History of the United States Air Force, chap. 3. 25. Baker quoted in Holley, Ideas and Weapons,

p. 170. Given that the Army built few bombers and refused to allow the Air Service to bomb German industrial targets, it is not surprising that American air forces had little impact on Germany's war-making capacity. Under the circumstances, it was disingenuous to criticize the Air Service on these grounds. 26. Pershing quoted in Goldberg, History of the United States Air Force, p. 29, and in Krauskopf, "Army and the Strategic Bomber," p. 84.

[38]

Historical

,

Organizational, and Doctrinal Setting

findings of the Crowell board, which favored the establishment of a

department of aeronautics, were deliberately suppressed by Secretary of War Baker, who had convened the panel in the first place. 27 The Army's position ultimately carried the day in Congress. The Army Reorganization Act of 1920 established the Air Service as a branch of the Army on a statutory basis, and it specified that aviation would be the fourth regular combatant arm along with the infantry, cavalry, and artillery. Although the Reorganization Act required the Army to place

command

Army

General Staff in charge of all major operational decisions. And, although the act gave the Air Service more day-to-day control over its acquisition programs, the War Department had the final word on the fate of these programs as well as on the size and shape of the Air Service's budget. The Air Service, moreover, would have less influence than ever over pilots in

of tactical aviation units,

it

left

the

these decisions, because the position of assistant secretary of aviation,

war

for

which was then vacant, was abolished.

statements about operational doctrine reflected the Army point of view during this period. A 1919 Air Service manual, for example, accepted the view that armies determined the outcomes of wars and that infantry was the heart of the army. According to this Official Air Service

manual, "when the infantry

loses, the

Army

loses." 28 Therefore, the

mission of the Air Service and the other combatant arms was to support the infantry. An Air Service textbook written by the Army Command and General Staff School in 1920 observed that "teamwork with the ground troops" was "the basic idea" underlying the organization of aviation units. 29 Observation and pursuit aircraft were consequently held to be

more important than bombers, and bomber

target selection

was to be made by the regular Army staff. A strategic striking force would be composed of whatever units were left after the tactical requirements of the ground forces had been satisfied. 30 As for hardware, most of the Air Service's acquisition programs were canceled when defense budgets were slashed after the war. Even so, it was allowed to keep some of its British-designed DH-4 bombers in service and to continue work on the MB-2, a new bomber being developed by the Martin aircraft company in the United States.

27.

According to Cate and Craven, "Army Air Arm," p. 24. Col. Edgar S. Gorrell, Air Service Notes on Recent Operations, June

18, 1919, pp. 1-2; Air Doctrine, quoted in Greer, Development of p. 15. 29. E. L. Naiden, Air Service (Fort Leavenworth, Kans.: General Service Schools Press,

28.

1920), p. 6. 30.

As discussed

in Holley, Ideas

and Weapons,

p.

172.

Flying Blind

The Air Service: 1920-1926

power advocates naturally resented being bridled by the Army in the early 1920s, and none was more bitterly resentful of this than Billy Mitchell, who had risen to the rank of brigadier general and the position of assistant chief of the Air Service. Mitchell became the country's most vocal proponent of air power and a thorn in the sides of his superiors, such as Maj. Gen. Charles Menoher, an artillery officer who had been Air

made

chief of the Air Service. Mitchell believed that "changes in military

systems come about only through the pressure of public opinion or disaster in war," and, since the United States was unlikely to become involved in any major wars in the near future, he decided to mobilize public support for his crusade to pry the air force away from the Army. 31

One

of Mitchell's

most extravagant claims was

that "aviation will

com-

pletely drive surface ships off the water in the next war," a contention that the Navy naturally rejected. 32 Mitchell convinced Congress to allow

him

conduct some tests of bomber effectiveness and ship survivability, which he began in July 1921. In a series of attacks involving DH-4 and MB-2 bombers against captured German ships, Mitchell's planes sank a German destroyer and a cruiser and then the supposedly to

unsinkable battleship Ostfriesland. 33 When word of Mitchell's success was leaked to the press, Menoher decided that the Air Service was not big enough for both of them. When the secretary of war refused to sack

who had

Congress, Menoher resigned. He was replaced as chief of the Air Service by Maj. Gen. Mason Patrick, an Army engineer who said in 1918 that he had never seen an airplane, "save Mitchell,

allies in

casually." 34 Mitchell's

bombers went on to sink a retired U.S. Navy battleship in September, which led him to proclaim that land-based aircraft were needed for coastal defense. The Navy argued that tests against defenseless ships proved nothing and that strong naval forces still

constituted the nation's

first line

of defense. 35

Although Mitchell championed the cause of air power in the early 1920s, he did not argue that the Army's air force should be composed primarily of bombers, contrary to what one might have expected. Mitchell recognized that the bombers then in service and on the drawing boards were technologically primitive; they would not be able to survive attacks by high-speed interceptors. He concluded that a large force of pursuit aircraft would be needed to sweep aside enemy air defenses in Mitchell quoted in Ransom, “Politics of Air Power," p. 113. 32. Mitchell quoted in Greer, Development of Air Doctrine, p. 36. 31.

33. 34. 35.

See Maurer, Aviation in the U.S. Army, pp. 113-121. Patrick quoted in Collier, History of Air Power, p. 77. Greer, Development of Air Doctrine, pp. 30-36. Air Service bombers also sank two

retired U.S.

Navy

battleships in

September

1923.

Historical, Organizational,

and Doctrinal Setting

would also be needed for air defense purposes. Mitchell consequently recommended that 60 percent of the air force be composed of pursuit aircraft, 20 percent bombers, and 20 percent attack aircraft designed to support operations on the the

stages of a war. Pursuit aircraft

first

ground. 36 Mitchell's vision of how this air force should be organized differed sharply from the plan the Air Service put together in 1921, which called for a force of twenty-seven squadrons, nineteen of which would be observation or surveillance units, four of which pursuit outfits, and four of which bomber units. Reconnaissance was emphasized at 37 the expense of both pursuit and bombardment.

American bombers in service in the early 1920s had limited capabilities. The Martin MB-2, the most advanced bomber in the inventory, was designed during the war and, like most of the bombers of this era, was a two-engine biplane. It had a top speed of only 99 mph, a ceiling of only 10,000 feet, and a total range of 550 miles. The Barling NBL-i bomber, which first flew in 1923, was a much larger six-engine aircraft. This underpowered airplane had a top speed of only failed to get over the 95 mph, and its ceiling was so low that it Appalachian Mountains in flight tests. Although bomber enthusiasts in the Air Service hoped to deploy a large, multiengine bomber some day, the NBL-i was a disappointment and only one prototype was built. The Curtiss B-2, which first took to the air in 1924, was another twoengine biplane, and it too had a top speed of around 100 mph. Pursuit aircraft did not have to carry heavy payloads or fly long distances, which allowed aircraft designers to maximize speed and altitude capabilities in these systems. As a result, pursuit aircraft of the mid-i920S had top speeds of 180 mph and service ceilings of 22,000 feet. These capabilities It

was

certainly true that the

would have given air defenses a decisive advantage over unescorted bombers in an actual engagement. 38 Bomber development lagged because it was inherently difficult, but also because it was not one of the Army's top priorities, to put it mildly. Little money was set aside for aviation activities in general in the mid-i920S. Air Service expenditures for fiscal years 1924-26 averaged only $12.5 million per year, approximately 2 percent of the total spent on national defense. 39 Although the Air Service set aside almost 25 percent of its budget for research and development during this period, most of 36.

Holley, Ideas and Weapons,

3U 36-3 8

p. 167.

See also Greer, Development

.

37.

Goldberg, History

38.

The MB-2 was

of Air Doctrine, pp.

30-

„

of the United States Air Force , p. 30.

as the NBS-i, and the B-2 was also known as the NBS-4; see ibid., pp. 16, 30, 32; Maurer, Aviation in the U.S. Army, pp. 81, 126, 214; Self, History of 10-11; Cate and the Development and Production of USAF Heavy Bombardment Aircraft, pp. also

known

Craven, "Army Air Arm," p. 58. Power, 39. Derived from Ransom, "Politics of Air

p. 116.

,

Flying Blind

was spent on

and observation aircraft. 40 Inevitably, this affected the progress of bomber technology. Small expenditures also affected the number of bombers the Air Service could buy for the active inventory. In fact, the Air Service failed to build a single bomber in 1923 and 1924, and it had only 39 bombers in service in 1924. It managed to add only two bombers to the force in 1923 and one in 1926. 41 The Army also exercised as much control as it could over doctrinal matters during this period. The 1923 Field Service Regulations mainthis

pursuit, attack,

tained that "the ultimate objective of all military operations struction of the enemy's armed forces by battle" and that "the

is

the de-

combined employment of all arms is essential to success." It continued, "The coordinating principle which underlies the employment of combined arms is that the mission of the infantry

the general mission of the entire force. The special missions of the other arms are derived from their powers to contribute to the execution of the infantry mission." Observation, close is

support, and pursuit aircraft were therefore needed for obvious reasons. Bombers had an important contribution to make, but mainly in air

terms of attacking tactical targets beyond the range of artillery. 42 As a textbook on the Air Service prepared for the Army's Command and General Staff College declared, tactical bombing was "a necessity" and

bombing

strategic

"a luxury." 43

These doctrinal discussions took place

in the context of

another debate

over the Air Service's institutional relationship with the Army. Mitchell continued to agitate for more autonomy, and two high-level review boards convened in the mid-i920S agreed with him. The Lassiter board of 1923

recommended

creating an independent force of

pursuit aircraft under General

Headquarters

bombers and

Command;

this

force

would concentrate more on strategic operations. This recommendation was not implemented by the War Department. The Lampert committee of the House of Representatives went further in 1923 when it recom-

mended

creating an administratively independent air force, as well as a Cabinet-level department of defense that would supervise the activities of

all

the military services. 44

40.

Goldberg, History of the United States Air Force p. 33. Holley, Ideas and Weapons, p. 172; Goldberg, History of

41. p.

the United States Air Force,

32.

42.

Field

Sendee Regulations (Washington: U.S. Army, 1923), pp. 11, 22-23, 77- Similar in Fundamental Conceptions (Washington: Air Service, 1923),

arguments were outlined pp. 1-2. 43. U.S.

Army Command and General Staff College, Corps and the Army Air Sendee (Fort Leavenworth, Kans.: General Service Schools Press, 19 22), p. 26. 44. See McClendon, Question of Autonomy, chap. 5; Mooney and Layman, Organization Military Aeronautics, chap. 3; Cate and Craven, "Army Air Arm," of pp. 24-29; Greer, Development of Air Doctrine, pp. 26-29; Goldberg, History of the United States Air Force, p. 36^ [42I

Historical, Organizational,

and Doctrinal Setting

At the same time, several factors worked against Mitchell and his cohorts. Many people remembered the bold claims made in 1917 about the impact American air power would have on the war in Europe; this undermined the credibility of the sweeping claims made by air power advocates less than ten years later. Given the isolationist and pacifist sentiments that dominated the American policy debate in the mid-i920S, of it was also hard to argue for military forces that would be capable engaging in offensive military operations, which bomber forces would certainly be able to do. And given that the most advanced aircraft then in service anywhere had an effective operating radius of only 300-350 miles, the United States faced no immediate threat from the air; it was 45 In any case, hard to argue for stronger air forces on these grounds. despite Mitchell's success in sinking old battleships in carefully staged exercises, most policy makers still assumed that the Navy would provide the nation with its first line of defense against overseas attack. Air

was not a sufficient reason for creating an expensive, independent service. Another high-level review panel, the Morrow board, recommended in late new 1925 that the Air Service not be given its independence and that a forces might constitute a second line of defense, but this

department of defense not be created. It favored the status quo, although it noted that steps should be taken to raise the Air Service's status and to increase its level of representation on the Army General Staff and in the War Department. Above all, the Army and Navy, which had strong vested interests in the status quo and were powerful political actors in Washington, were opposed to the idea of spending a great deal of money on Army aviation forces, redefining the Air Service's military mission in any fundamental way, or granting the Air Service much operational or administrative autonomy. Ultimately, the Air Corps Act of July 1926 made only a few changes in the Army's relationship with its aviators. It elevated the status of the Army's air force from a service to a corps, and it reestablished the position of assistant secretary of war for aviation, which gave the new Army nificant,

which

War Department decisions. More sigaviators more control over their budget, long run allowed them to devote more effort to developing

(AAC) more input it gave the Army's

Air Corps

in the

into

the kind of highly capable, long-range bombers unofficial air force thinking called for. Finally, the Air Corps Act marked a turning point in steps official thinking on the uses of air power when it noted that these

should strengthen "the conception of military aviation as an offensive, 46 striking arm rather than an auxiliary service.' These points are developed

Ransom,

Politics of

Air Power,

pp. creation The Force, 36. Air States United p. 46. Quoted in Goldberg, History of the provisions of the Air Corps Act are discussed in the sources cited in note 44. 45.

[

43

]

in

and

Flying Blind

power advocates had already started to stake out radical positions on bomber operations. Mitchell was court-martialed in late 1925 after he accused the Army and Navy high commands of "incompetency, criminal negligence, and almost treasonous administration of the national defense," and he resigned his commission in February 1926 In fact, air

rather than suffer through a five-year assignment in San Antonio. 47 Freed completely from bureaucratic constraints, he began to place more

emphasis on the industrial centers

bomber forces to destroy the enemy's vital quickly and he began to play down the role of pursuit

ability of

aircraft in offensive operations. 48

Two

textbooks published by the Air Service Tactical School in 1926 also stated what were, from an official standpoint, heretical positions on how

bomber operations should be conducted. The Employment of Combined Air Force manual began by declaring that the primary objective of air operations was not to assist in the destruction of the enemy's ground forces but to take the lead in destroying the enemy's capacity and will to continue. It went on to say that the best way to bring about an enemy's economic collapse and a quick end to a war was "heavily striking vital points rather than gradually wearing down an enemy to exhaustion." 49 The textbook for the bombardment course at the tactical school stated that these industrial targets must be selected with care. According to this

manual,

it

would be "wrong

to

send out planes simply to drop their Employment manual ar-

bombs when over gued that bomber offensives would be difficult to stop in the future because bombers would have the edge over pursuit aircraft. It therefore a large target." 50 Finally, the

considered bombers to be the most important element in the air force. 51 These themes the independent and decisive nature of strategic bombardment operations, the importance of targeting vital economic cen-

—

—

and the ascendancy of bombardment over pursuit would become main pillars of the strategic bombardment doctrine that was refined in the Air Corps Tactical School in subsequent years. Although the Air Corps Act of 1926 addressed several chronic organiters,

zational issues,

did not resolve the underlying doctrinal dispute that kept the Army and its aviators at odds. In fact, it aggravated this dispute by giving credence to the claim that aircraft should be used as indepenit

dent striking forces and not just as adjuncts

to infantry operations.

Mitchell quoted in Goldberg, History of the United States Air Force, p. 32. 48. See Weigley, American Way of War, chap. 11; Cate and Craven, “Army Air Arm," PP- 33 43 Schaffer, Wings of judgment, pp. 25-27. 49. Air Service Tactical School, Employment of Combined Air Force (Washington: Government Printing Office, 1926), pp. 3-4. See also Greer, Development of Air Doctrine, pp. 40-43. 50. Air Service Tactical School, Bombardment (Washington: Government Printing Office, 47.

'>

1926), p. 63. 51.

Employment

of

Combined Air

Force, pp.

9-11, 23-24.

— Historical, Organizational,

Doctrinal controversy

and Doctrinal Setting

would become even more intense once bombers

developed the capacity

to carry out long-range offensive missions. 52

The

Army

Air Corps: 1926-1935

high-performance bombers in the late 1920s, but they were constrained by three things. First, the technological state of the art prevented high-speed aircraft from carrying heavy payloads long distances. The bombers of the late such as the Curtiss B-2 and the Keystone B-3, B-4, B-5, and B-6 1920s were all biplanes, and they could not fly very fast or very far. The B-4, 53 for example, had a top speed of 120 mph and a range of 850 miles. Second, the Army continued to supervise the AAC acquisition program, and it tried to keep the AAC from building highly specialized aircraft optimized for long-range bombardment missions. In 1929, for example, the Army shelved an AAC proposal to develop a high-speed, heavily armed bomber. Instead, it forced the AAC to buy a mediocre allpurpose plane that was also designed to perform reconnaissance and

Bomber enthusiasts were anxious

to build

—

close air support missions. 54 This short-range, lightly

the Douglas O-35, also

known

as the B-7

armed

aircraft

— was incapable of flying long-

range missions over heavily defended targets, which the

AAC

hoped

to

do.

AAC's annual expenditures increased to an average of $22 million in the late 1920s, this was not enough to sustain a full55 Priorities had to be estabscale modernization of the entire air force. lished, and bomber modernization was at the bottom of the Army's Third, although the

priority

list.

1930 when the Army permission to launch a design competition for a spe-

Bomber development reached gave the

AAC

a

watershed

in

high-performance bomber. The fact that Army budgets were rising sharply at the time seems to have influenced its decision to allow some leeway on this issue. 56 In any event, the AAC wasted little time in

cialized,

circulating a request for proposals to the aircraft industry. Six

companies

responded by building experimental prototypes, the most promising of which were two twin-engine aircraft, Boeing's B-9 and Martin's B-10. incorporated several technological advances. Most important, they were all-metal monoplanes with retractable landing gear.

Both

aircraft

Discussed in Greer, Development of Air Doctrine, p. 43. 53. Maurer, Aviation in the U.S. Army, pp. 214-215. 54. Cate and Craven, "Army Air Arm," p. 59. Rutkowski, The 55. Ransom, "Politics of Air Power," p. 116; Edwin H. 52.

Politics of Military

Aviation Procurement, 1926-1934 (Columbus: Ohio State University Press, 1966), chap. 56. Ransom, "Politics of Air Power," p. 116.

I45]

2.

Flying Blind

These improvements reduced aerodynamic drag and enabled the aircraft to slice through the air relatively cleanly and efficiently. This, in turn, dramatically improved their performance. The B-9, which first flew in April 1931, had a top speed of 188 mph, more than 60 mph faster than the biplane bombers then in service. The B-10, which first flew in July 1932, was even more advanced in that it featured an internal bomb bay. The B-10's top speed was over 200 mph, faster than any pursuit aircraft in the U.S. inventory.

than the B-4, and

AAC

its

could fly at 21,000 feet, 50 percent higher production version had a range of 1,240 miles. The It

ultimately bought seven B-9S

and 152 B-ios

of various

makes and

models. 57

Other important technological developments were also taking place around this time. Most notably, the advanced Norden Mark XV bombsight was tested in October 1931. The AAC conducted extensive tests of this new bombsight in 1932 and placed a production order for the Norden system in 1933. To keep up the competitive pressure and maintain an alternative source of supply, the AAC also placed an order with the Sperry Gyroscope Company in 1933 for models of its advanced bombsights. 58

Meanwhile, the AAC continued to push ahead with long-range bomber development. It issued a request for proposals in December 1933, calling for a

bomber capable

of carrying a 2,000-pound payload on a 5,000-mile mission at a speed of 200 mph. These were extremely de-

manding requirements, given this distance

that the B-10 could fly only a fraction of

with a comparable payload. Boeing

won

development contract in 1935 to build the B-15, which first flew in October 1937. Flight testing subsequently revealed that it could not meet the AAC's original performance requirements. It could fly 5,000 miles and reach speeds of a

mph

without a payload, but with a full payload it could fly only 3,500 miles at 145 mph; it would be too slow to survive against highspeed air defense interceptors. The B-15 was simply too large and too heavy for the engines that were available, even though it was powered by four engines. These technical problems, in conjunction with Army 195

opposition to building large numbers of long-range bombers, led to a decision to build only one B-15 prototype. 59

57. The B-i2, B-13, and B-14 were advanced versions of the B-10; see DuBuque and Gleckner, Development of the Heavy Bomber, pp. 70-72; Swanborough and Bowers, United States Military Aircraft, pp. 85, 375-378. 58. James L. Cate, "Plans, Policies, and Organization," in Craven and Cate, eds., Army Air Forces in World War II, vol. 1, pp. 598-599.

DuBuque and Gleckner, Development of the Heairy Bomber, pp. 83-86; Self, History of Development and Production of USAF Heavy Bombardment Aircraft, 16-17; Krauskopf, pp. "Army and the Strategic Bomber," pp. 91-92; Goldberg, "AAF Aircraft," p. 202; Cate and 59.

the

[

46

]

Historical, Organizational,

and Doctrinal Setting

Another development program was launched in May 1934 when the AAC announced a design competition for a high-speed, long-range, multiengine bomber. Boeing was the only contractor to submit a proposal for a four-engine aircraft, which gave it a decisive advantage over the competition; two-engine bombers had comparatively limited speed

and payload capabilities. Boeing won a contract to build a prototype of what became known as the B-17. The first flight of this aerodynamically clean aircraft took place in July 1935, and the new bomber quickly went on to demonstrate impressive performance capabilities. It had a top speed of over 250 mph, a service ceiling of over 30,000 feet, and a maximum payload of over 10,000 pounds. In one test it flew 2,100 miles at an average speed of 232 mph. The heavily armed bomber also had five machine gun turrets, which led people to call it the Flying Fortress. he B-17 was fh e bomber of the AAC's doctrinal dreams, and bomber enthusiasts lobbied intensively to buy 65 B-17S straight away; the War Department cut this request back to 13 aircraft. The rest of the decade saw a continuing tug of war between the AAC and the War Department over B-17 production. The AAC wanted to buy enough B-17S to field a legitimate strategic striking force, while the War Department wanted the AAC to buy less capable but more versatile aircraft. 60 All these programs were shaped by the AAC's prevailing conception of air power, which as early as 1926 explicitly emphasized bombardment operations against selected economic and industrial targets. Operations of this type could not be conducted, however, given the bombers in service in the late 1920s. The AAC needed bombers capable of flying long distances, penetrating stiff air defenses, and delivering large payloads accurately. The bombers of the late 1920s could do none of these things, let alone all of them. There was, therefore, an emerging doctrinal imperative for a bomber with impressive range, speed, altitude, defensive armament, payload, and bomb delivery capabilities. I

This doctrinal imperative, in

became

substantially

more

its

formative stages in the late 1920s,

explicit in the early 1930s. In

many

respects,

important refinements appeared in AAC thinking before technological developments took place that would make them practicable. In many respects, doctrine called for new technology and new weapon systems; it did not flow from them. For example, the primacy of strategic bombardment operations was firmly established at the Air Corps Tactical School

Craven,

“Army

Aircraft, p. 540. 60. DuBuque

Air Arm," pp. 63-66;

Swanborough and Bowers,

United States Military

Heavy Bomber, pp. 74-80; Self, History of Goldberg, the Development and Production of USAF Heavy Bombardment Aircraft, pp. 17-20, “AAF Aircraft," pp. 203-208; Cate and Craven, "Army Air Arm," pp. 66-69; Swanborough

and Gleckner, Development

and Bowers, United [47]

of the

States Military Aircraft, pp. 87-96.

Flying Blind

long before a real strategic bomber had been built. Some instructors at the school argued in 1930 that pursuit aircraft would not be able in 1930,

to stop

two

bombardment

operations, one year before the B-9 took to the air, years before the B-10 began its testing program, and roughly five

years before the B-17

made

Textbooks published at the Tactical School in 1931 stressed that new bombsights would be needed to achieve the accuracy necessary to carry out effective bombardment operations. They also began to argue in 1931 that bombardment operations would have to be conducted in daylight if precise attacks were to be successful. These arguments were developed before the Norden and Sperry bombsights were fully tested and before a bomber capable of its first flight.

surviving daylight operations, such as the B-10 or B-17, had flown. 61 In all these respects, doctrine demanded new technologies and better

bombers. Given that

had already begun to take shape bv the early 1930s, it is not surprising that the AAC guided the requirements formation process for the B-15 and B-17 with a firm hand. At the same time, it would be simplistic to argue that AAC doctrine

was immune

AAC

1930s.

this doctrine

to the influence of technological

was

thinking

developments

in the early

formative stage during this period, and technological developments played an important role in reinforcing emerging doctrinal currents and suggesting doctrinal refinements. For example, the emergence of the B-9 and B-10 in the early 1930s buttressed the claims made by bomber enthusiasts about the feasibility of bombard-

ment

operations.

It

still

in a

also suggested that high-speed

bombers might not

need pursuit escorts, which was still a bone of contention at the Air Corps Tactical School. The development of the Norden bombsight led instructors at the school to look more closely at the possibility of bombing targets with pinpoint accuracy.

It

was not

a coincidence that the Air

Corps Tactical School began to stress the importance of targeting specific economic chokepoints in 1933, after the capabilities of the Norden bombsight became clear. And the appearance of the high-speed, heavily armed B-17 ended the debate over pursuit escorts for the rest of the decade. Received wisdom subsequently held that advanced bombers would not need escorts. 62 It is

came

generally agreed that the AAC's theory of strategic bombardment together by 1935. This theory consisted of five main proposi-

tions. 63

and most generally, the AAC argued that air power was decisive warfare and that bombardment was the dominant form of air power.

First

in

61. 62.

See Finney, History of the Air Corps Tactical School, pp. 31-32. See ibid., pp. 3 1— 33 3^; Greer, Development of Air Doctrine, '

History of the United States Air Force, 63.

This discussion

is

pp. 57—59; Goldberg,

p. 44.

based on the sources cited

in

note 20.

[48]

Historical, Organizational,

and Doctrinal Setting

This view was based on the conviction that bombardment operations would determine the course of future military conflicts, but it was also true that the AAC needed to establish for purely bureaucratic reasons that it had a distinct, autonomous military mission to perform. Its case for institutional air forces

independence from the

were more than adjuncts

to

bureaucratic motivations also led the

Army

rested

on the claim

that

ground operations. Therefore,

AAC

to

adopt bombardment as

its

primary military mission 64 Second, the AAC argued that bombardment would be most effective when it was directed at an enemy's economic infrastructure and warmaking capacity. Attacking an enemy's military forces was held to be important only in that one had to destroy enemy bombers and air de.

fenses to establish

and naval

forces

command

of the

air.

The

AAC

believed that ground

would be highly vulnerable to attack from the wars could be won more quickly and more

air,

but

it

easily by maintained that destroying an enemy's economic support system. This strategic argument was reinforced by a bureaucratic concern. If bombardment operations focused on battlefield forces, they would tend to become tactical in

character, auxiliary in nature,

and subject

to the influence of

Army com-

manders on the ground. In defending the strategic character of bombardment operations, the AAC was also safeguarding its operational independence. In addition, the

AAC

rejected the idea of targeting urban residential

did not believe, as the RAF did, that civilian morale 65 The AAC argued that it should be a main focus of strategic operations would be easier to destroy an enemy's war-making capacity than to

areas per se.

It

.

undermine civilian morale. This strategic argument was reinforced by moral and practical considerations. Not only was the idea of deliberately attacking civilians abhorrent to

many Americans,

including

many

air-

men, but from a practical standpoint a doctrine that called for attacking civilians was unlikely to become national policy and an organization that advocated operations of this type was unlikely to receive strong political support from any quarter 66 This was an important political consideration given that the AAC, unlike the RAF, was still trying to win its independence from the Army. .

This point is developed in Smith, Air Force Plans for Peace, chap. 2; Ransom, "Politics of Air Power," pp. 107-117. differed from the 65. For more on British strategic bombardment doctrine and how it (London: Groom Slide-Rule without American conception, see Barry D. Powers, Strategy Helm, 1976), chaps. 5-6; Williamson Murray, "British and German Air Doctrine between the Wars," Air University Review 31 (March-April 1980), 39 5 8 Cate and Craven, "Army Air Arm," pp. 33-34; Quester, Deterrence before Hiroshima, chaps. 5, 7-8. War II, see 66. For a discussion of the moral choices faced by American airmen in World 64.

;

Schaffer's excellent study. Wings of Judgment.

I49]

Flying Blind

Third, the

AAC

argued that the most effective way of crippling an enemy's economy was to bomb key industries thoroughly. No doubt influenced by the unfolding events of the Great Depression, AAC strategists believed that modern economies were highly interdependent and consequently "brittle." They argued that a country's "industrial web" contained a few key industries or "vital centers," the destruction of

which would paralyze the economy as a whole, and they concluded that the challenge for air force planners was to identify a country's economic "solar plexus" so that focused but intense attacks could create systemwide "bottlenecks." The suggested that petroleum and steel indus-

AAC

tries,

power-generating

tions centers

were

facilities,

and transportation and communica-

likely targets 67 In this respect, too, .

AAC

thinking

—

from that of the RAF, which favored more widespread some would say indiscriminate— attacks on large industrial and urban areas. differed

The

AAC

believed that scattered attacks on large areas were unlikely to

induce economic collapse. This strategic argument was also buttressed by practical considerations.

peared

One to

of strategic

bombardment's main

make wars

attractions

AAC

of attrition things of the past; forced this impression especially well. In addition, the

economic targeting plans appeared scientific exercise, which enhanced

was

that

it

ap-

doctrine rein-

AAC's elaborate

be the product of a rigorous, their credibility. Moreover, it appeared that aerial blitzkrieg attacks could be carried out by a relatively small fleet of bombers, an important consideration given that defense budgets were tight in the mid-i93os. AAC plans therefore seemed to be both economical and affordable. to

Fourth, the AAC argued that air forces had to be capable of conducting attacks with pinpoint accuracy in order to destroy well-defined economic targets; thus, it favored daylight bombing operations. Again, AAC thinking differed from that of the RAF. British strategists found in

World War that bombing accuracy was poor even during the day; they also argued that daylight operations were hazardous because air defenses were especially stiff during the day. The AAC maintained, however, that technological advances since World War had improved bombing accuracy significantly and that modern, high-performance I

I

67.

AAC

thinking diverged from that of Douhet,

who advocated indiscriminate attacks about target selection. See Finney, History of the Air Corps Tactical School, pp. 30-34; Greer, Development of Air Doctrine, pp. 57-60; Futrell Ideas Concepts, Doctrine, pp. 55-86; Cate and Craven, "Army Air Arm," pp. 33-54; Cate, "Development of Air Doctrine," pp. 19-21; Smith, Air Force Plans for Peace, pp. 27-58; Weielev o y> American Way War, on urban centers and said

of

little

pp. 236, 337.

[50]

Historical, Organizational,

bombers would be much

and Doctrinal Setting

less vulnerable to air

defenses than their dis-

tant ancestors 68 .

AAC's strategic arguments were reinforced by pragmatic considerations. American military strategy had a strong defensive orientation in the 1930s, and American military and political leaders were unlikely to adopt an offensive scheme like strategic bombardment in the immediate future. At the time, the AAC's only official long-range mission was providing for land-based coastal defense. According to an agreement reached between the Army and the Navy in 193 1 AAC aircraft were to help defend American coasts and important sea lanes from

Once

again, the

'

naval attack, the only real threat the United States faced at the time. To do this, the AAC was allowed to buy a small fleet of long-range bombers (B-ios and, later, B-17S). These bombers had to be capable of attacking

moving

ships; this required accurate

bomb

delivery. So, the best

way

for

had on hand was to adopt an operational doctrine that could be carried out by a 69 relatively small fleet of highly accurate bombers Fifth, the AAC argued that high-performance bombers would be able the

AAC

to build a strategic striking force out of the force

it

.

without the benefit of pursuit escorts. It believed that many enemy air defense aircraft would be caught on the ground and that many enemy air defense bases would be to attack targets far

behind enemy

lines

rounds of the next war. Moreover, it believed that high-performance bombers would be capable of defending themselves against remaining air defenses: bombers would avoid air defenses by flying at high altitudes, beyond the reach of antiaircraft guns and low-altitude interceptors; fast-flying bombers would outrun the interceptors that could climb to high altitudes; and large formations of heavily armed bombers would outgun the aircraft they

knocked out

of

commission

in the early

could not outrun. The AAC's belief in the invincibility of the bomber led in the late 1930s, a decision it to put little effort into escort development 70 many bomber pilots would rue in 1942 and 1943American air power strategists were critical of the British approach to strategic bombardment even in the years immediately after the war; see the narrative summary to World War l, the U.S. Bombing Survey for World War in Maurer, ed., U.S. Air Service in views after the Battle vol. 4, pp. 495-503. British strategists became more adamant in their operations during daylight in participate of Britain in 1940, and they were reluctant to World War II even though long-range escort capabilities improved as the war wore on; see Weigley, American Way of War, pp. 354 35 ^68.

1

Because of these constraints, the AAC invariably justified requests for long-range bombers in terms of coastal defense and reinforcement of outlying possessions. This was History of the United certainly true for the B-10, B-15, and B-17 in the 1930s; see Goldberg, 69.

States Air Force, pp. 38-42. 70. Finney, History of the Air Corps Tactical School, p. 32.

"The Search

for a

Long Range

See also Bernard

Escort Plane, 1919-1945," Military Affairs 30

L.

Boylan,

(Summer

1966),

Flying Blind

Bureaucratic considerations also led the AAC to neglect escort aircraft in the 1930s. The case for the strategic bomber would have been weak-

ened if bombers were seen as vulnerable to air defenses and dependent on pursuit escort, and spending money on escort development and procurement would have diverted resources away from bomber programs, which were already underfunded as far as the AAC was concerned. The AAC was convinced, in any event, that it would be extremely

not impossible, to build a long-range escort. This strategic bombardment doctrine could be carried out only if very advanced bombers could be built. These bombers would need impressive performance capabilities in no less than six areas: range, aldifficult,

if

speed, defensive armament, payload, and accuracy. 71 There were inherent trade-offs in these doctrinally driven performance requirements, however. For example, if everything else was equal, one titude,

had

to sacrifice aircraft

speed

to get

more range, and

vice versa. Similar-

armed bomber could not carry as much payload as a lightly armed aircraft, and it was hard to fly a heavily armed bomber with a big payload far. In addition, bombing accuracy decreased when bombers flew higher and faster. The AAC tended to ignore these trade-offs and push performance requirements in many areas simultaneously, as it did in the B-15, B-17, and later programs. Although the AAC's strategic bombardment doctrine was fully developed by the mid-i93os, it still had to justify bomber procurement in terms of coastal defense requirements. American military and political leaders were not yet convinced that the country needed a bomber force per se. The Baker board of 1934, chaired by former War Department Secretary Newton Baker and composed of four ground officers and only a heavily

ly,

one

maintained that "the idea that aviation, acting alone, can control the sea lanes, defend the coast, or produce decisive results in any other general mission contemplated under our policy is visionary, as is the idea that a very large and independent air force is necessary to defend our country against air attack." 72 The Baker board believed that the AAC should remain part of the Army, under General Staff control, but it recognized that public sentiaviator,

One

developments AAC strategists did not foresee in the 1930s radar, which would alert air defenses to approaching bombers and provide more time for interception. Nor did the AAC expect the speed capabilities of air defense aircraft to improve as much as they did in the late 1930s. As Greer, Development of Air Doctrine, 57-67.

of the technological

was

pointed out, the effectiveness of 1930s,

when AAC

strategic

was at an all-time low bombardment doctrine came together. air

defense

aircraft

The RAF, on

p. 58, in the early

71. the other hand, emphasized nighttime operations, so its bombers did not have to be particularly fast or heavily armed. RAF bombers did, however, carry big

payloads 72.

to compensate for their lack of accuracy. Quoted in Krauskopf, “Army and the Strategic Bomber,"

pp. 88-89.

Historical, Organizational,

and Doctrinal Setting

an independent air force. 73 Twelve bills calling for an independent department of aeronautics and seventeen bills calling for an independent air force under the supervision of a department of defense had been considered by Congress since 1926. More ominous, the Federal Aviation Commission's Howell board was expected to recommend setting up the AAC as a separate department when it issued its report in early 1935. With this in mind, the Baker board offered a compromise: a General Headquarters Air Force that would be composed of all AAC combat units and trained as a homogenous organization but would still be under the control of the Army chief of staff in peacetime and Army field commanders in time of

was moving

merit

in the direction of establishing

war. These limitations notwithstanding, AAC leaders welcomed the General Headquarters Air Force as a step forward; it provided for the operational unification of Army air forces and creation of a bomber force that could be used in strategic operations. The Howell board subse-

quently decided to give the new organizational experiment a fair trial rather than recommend more radical reforms, and the General Head74 quarters Air Force was activated in March 1935.

Preparations for War: 1935-194.1

The

AAC continued

to press for

in the late 1930s, especially for

bomber development and production

long-range systems that could be used for

this either coastal defense or strategic bombardment. It was opposed in had been "led astray by the by the Army, which believed that the allurement of a quest for the ultimate in aircraft performance at the

AAC

expense of practical military need." As far as the Army was concerned, the most important of these practical needs included close air support and tactical bombardment capabilities, neither of which required longrange aircraft. The AAC was also opposed by the Navy, which resented 75

threatened by the AAC's growing interest in coastal defense, national traditionally one of the Navy's most important contributions to

and

felt

security. 76

One Its

of the

goal

AAC's most ambitious programs

was

to build

an

aircraft

in the 1930s

was

the B-19.

capable of carrying a 2,400-pound

Army

William A. Goss, "Origins of the Army Air Forces," in Craven and Cate, Air Forces in World War II, vol. 6, p. 3. Strategic Bomber," pp. 88-90; Cate and Cra74. Based on Krauskopf, "Army and the chap. 6; Mooney and ven, "Army Air Arm," pp. 29-32; McClendon, Question of Autonomy, Layman, Organization of Military Aeronautics, chap. 4; Maurer, Aviation in the U.S. Army, eds..

73.

chaps. 16-18.

Goldberg, History of the United States Air lone, p. Air Doctrine, pp. 89-91. 76. As discussed in Greer, Development of 73.

According

to

42.

Flying Blind

payload on an 8,000-mile mission, 60 percent farther than the B-15 was to fly with a smaller payload. The B-19 program was launched in October 1935, when the Douglas Aircraft Corporation began design development. The program soon attracted the attention of the Army General

which opposed it on the grounds that it would be "an airplane of aggression. 77 In August 1936, the AAC was given permission to build one B-19 prototype, but only because the program was already under way and the Army did not want to risk damaging the government's Staff,

reputation as a reliable research and development partner. The Army made it clear, however, that further development of the new bomber

was

and production was simply out of the question. As it turned out, Douglas had difficulty building an aircraft capable of meeting the AAC's extremely demanding requirements. The first flight of the unlikely,

B-19 did not take place until June 1941, and, although the B-19 could fly 7,800 miles without a payload, it could fly only 5,200 miles with a 2,500pound payload. And, as in the case of the B-15, the lumbering B-19 would have been highly vulnerable to high-speed air defense aircraft. 78 The Army and the AAC also disagreed over the size and pace of the B-17 production program. The AAC's running battle with the Army to build more B-17S came to a climax in May 1938, after three B-17S intercepted the Italian liner Rex 725 miles east of New York City in an offshore exercise. The Navy, outraged by the AAC's encroachment on its sea-lane defense mission, pressed the Army to renegotiate their understanding on the AAC's offshore operations. The Army, anxious to corral the AAC for its own reasons, agreed. The Army chief of staff and the chief of naval operations subsequently decided that AAC coastal defense operations would no longer extend beyond 100 miles of the U.S. coast. This decision swept away the AAC's official justification for building the B-17 and longer-range bombers. In June, Secretary of War Harry

Woodring would be allowed to buy only 13 B-17S in fiscal year 1939 and only light, medium, and attack bombers— no B-17S— in 1940. The AAC was left with a total of 52 B-17S on order, as well as 350 of the told the

AAC

that

it

shorter-range, twin-engine B-i8s the Army General Staff preferred. According to the Army's deputy chief of staff, only a relatively small number of B-i 7 s was needed to reinforce Hawaii, Alaska, and Panama, and no plane larger than the B-17 was needed. The Army consequently

Quoted

Krauskopf, "Army and the Strategic Bomber," p. 85. 78. See DuBuque and Gleckner, Development of the Heavy Bomber, pp. 87-88- Self History of the Development and Production of USAF Heavy Bombardment Aircraft, pp. 21-22Krauskopf, "Army and the Strategic Bomber," pp. 85/92-93; Cate and Craven, "Army Air Arm, p. 69; Greer, Development of Air Doctrine, pp. 96-97; Swanborough and Bowers United States Military Aircraft, p. 556. See also "XB-19," April 30, 1943, Aircraft Projects ° berS )' RG 341 HQ/U s Air Force (USAF), National Archives and Records Service K A oc (NARS). 77.

in

'

m

'

-

T

[54]

Historical

,

Organizational and Doctrinal Setting ,

turned down the AAC's request B-17, an aircraft that would be

changed its mind in 1940. 79 The deteriorating situation

in

to build a longer-range version of the

known

as the B-29

when

Europe soon led American

the

Army

political

and

and procurement priorities. Although Germany's rearmament program had been under way for some time, its significance was not driven home to most American observers until September 1938, when Britain and France acquiesced at Munich to Hitler's demand that Czechoslovakia military leaders to reassess the country's military preparedness

cede the Sudetenland. President Roosevelt believed that the threat implied by the existence of a powerful German air force played an important role in Hitler's success at Munich, and Roosevelt subsequently asked the Army if the United States was in a position to build 15,000 aircraft a year, should it become necessary to do so. In his State of the January 1939, Roosevelt stated that the AAC's "antiquated force" of 1,800 aircraft was "utterly inadequate," given the military arsenals that were being built elsewhere, and he urged Congress to embark on an immediate buildup designed to reinforce hemispheric defense. He also observed that the findings of the 1934 Baker board were now "completely out of date." 80 In April, Congress responded by appro-

Union address

in

priating $300 million for aircraft acquisition, which to field a force of 5,500 aircraft. The fact that the

would allow the AAC AAC was assigned a

major role in hemispheric defense was especially important because it 81 gave the AAC a military justification for building long-range bombers. The situation became even more acute in September, when the war in Europe began with the German invasion of Poland. At the time, the AAC had a total of 13 B-17S in service and another 39 on order. In the spring of 1940, Norway, Belgium, and the Netherlands fell to Germany in quick succession, and French forces were reeling from the German blitzkrieg. Roosevelt consequently asked Congress in May to fund a force of 50,000 aircraft and to develop the country's industrial capacity to a point where it could build 50,000 aircraft per year. To put this in context, the U.S. aircraft industry had only built 40,000 aircraft since 1903, and it was then producing around 2,000 aircraft per year; Roosevelt wanted the capacity to build 4,000 per month. In June, British forces Krauskopf, "Army and the Strategic Bomber," pp. 208-215; Greer, Development of Air Doctrine, pp. 91-100; Maurer, Aviation in the U.S. Army, pp. 402-412; Goldberg, "AAF Aircraft," pp. 203-208. 80. Roosevelt quoted in Goldberg, History of the United States Air Force, pp. 44-45, and in Goss, "Origins of the Army Air Forces," p. 10. 81. See Goss, "Origins of the Army Air Forces," pp. 8-9; Cate and Craven, "Army Air Forces," pp. 116-126; Greer, Development of Air Doctrine, pp. 76-77, 100-101; Irving B. Holley, jr., Buying Aircraft (Washington: U.S. Army Office of the Chief of Military History, 79.

1964), chap. 8.

[55]

Flying Blind

were driven

off the

continent and France

fell

to

Germany. Congressional

appropriations for military programs of all kinds subsequently skyrocketed. In the twelve-month period beginning in July 1940, the AAC

spent $100 million on research and development alone, and $42 million of this was devoted to long-range bomber programs. 82

Once

American military buildup began in 1939-1940, the Army stopped being the main constraint on AAC bomber procurement. The problem became the aircraft industry's limited capacity to engage in development and production at breakneck speed. The growing national emergency led the AAC to set aside its traditional acquisition practices, in which research, design development, engineering development, and the

production took place

in

an orderly, sequential manner. The

AAC began

compress the acquisition process, allowing development and production to overlap. 83 The B-24, B-26, and B-29 were the first major acquisition programs to be guided by this concurrent procurement strategy. Each of these bombers was technologically ambitious, and each experienced procurement problems. to

The B-24, Consolidated's counterpart to Boeing's B-17, was designed to fly a 3,000-mile mission at 300 mph with a 9,000-pound payload. Roosevelt's 1939 State of the Union speech led to the first B-24 development contracts in March of that year. Significantly, a production contract for 43 aircraft preceded the first flight of a prototype, which took place in December 1939. A production contract for 500 B-24S was signed in the autumn of 1940 even though development of the aircraft was still continuing. Ultimately, a total of 18,190 B-24S were built during the war, the vast majority of which had to go through modification centers before

being fielded in order to incorporate the hundreds of changes that were made in the design of the aircraft after it went into production. 84 Extensive retrofitting soon became a hallmark of concurrency. Martin's B-26 program also got under way in the months after Roosevelt's 1939 speech. Performance requirements for the twin-engine bomb-

were issued in March, and a design competition led to an AAC decision to buy 201 B-26S in July on the basis of drawing board plans alone. This off-the-shelf procurement strategy, as it was called, involved even more concurrency than the procedure used for the B-24; a bigger comer

Army

82.

Goss, "Origins of the

83.

For

84.

Greer, Development of Air Doctrine,

Air Forces,"

p. 13; Krauskopf, "Army and the Strategic Bomber," pp. 211, 214; Holley, Buying Aircraft, chap. 11; James L. Cate and E. Kathleen Williams, "The Air Corps Prepares for War," in Craven and Cate, eds.. Army Air Forces in World War II, vol. 1, pp. 101-150.

more discussion

of the use of concurrency and the problems it caused during World War II, see Tom Lilley et al., Problems of Accelerating Aircraft Production during World War II (Boston: Harvard University Graduate School of Business Administration" p. 118;

Goldberg,

"AAF

1946). Aircraft," pp. 206-207;

Holley, Buying Aircraft, pp. 518-538.

[56]

Historical

,

Organizational, and Doctrinal Setting

development process. 85 Flight testing of the B-26 began in November 1940, and it soon demonstrated serious design problems. Most notably, it had to take off and land at high speed because its relatively small wing generated little lift. This was a dangerous proposition, and five of the first six aircraft delivered to the AAC crashed. Even though extensive modifications were eventually made in the B-26's design— its wingspan and wing area were the "Widow Maker" continued to have a high increased, for example accident rate. Mechanical defects led a board of officers to ground the entire B-26 fleet at one point during the war. One wartime review identified the program's basic problem: "While it was generally admitted that

mitment

to

production was

made

earlier in the

—

extensive testing of the B-26 prior to the initiation of a production contract would have furnished a better production airplane, the immediate needs of the Air Corps demanded faster, if more hazardous, procure-

ment." 86 A total of 5,157 medium-range B-26S were built before the need for longer-range bombers in the Pacific led to a cutback in B-26 procurement. The B-29 program, which the AAC tried to start up in 1938, finally got under way in January 1940, when performance requirements were issued for a heavily armed aircraft that could carry a 2,000-pound payload (which would give it a 2,000-mile radius of action) at a speed of 400 mph. A design competition led to an August contract to Boeing to build two B-29 prototypes. A production contract for 500 B-29S was signed in May 1941, and three other manufacturers were brought into the program in February 1942 to begin setting up production lines for the 1,644 aircraft that were then on order. At this point, flight testing of the new bomber had not yet begun. A prototype took to the air in September 1942, which coincided with a production order for 1,000 more B-29S. Although the B-29 did not have serious design flaws comparable to the B-26's, it was nonetheless rushed into

on

a 5,333-mile mission

production before development of the system and its military subsystems was complete. The B-29, moreover, was technologically ambitious in several respects. For example, it featured a pressurized cabin, remote-control turrets, as well as new engines and propellers. As a result, over 900 design changes had to be made between the time production contracts were

first

awarded and the

flight of the

first

B-29

See Capt. C. M. Thomas, Materiel Command, "Circular Proposal," March 11, 1939; of the Air Corps, July 25, Brig. Gen. G. H. Brett, Chief, Materiel Division, Letter to Chief NARS. of the AAF, 1939; both in B-26 Project, RG 18, Records 85.

McMurtrie and Paul M. Davis, History of the Army Air Forces Materiel Command, 1926-1941, Historical Study No. 281, Materiel Command, Nov. 1943, p. 88. See 86.

Mary

L.

also ibid., pp. 86-89; ical

Study No.

196,

Edward O.

AMC,

Purtee, Development of Light and

Dec. 1946, pp. 114-121; Goldberg,

Medium Bombers,

"AAF

Histor-

Aircraft," pp. 199-201.

Flying Blind

prototype. Three modification centers were set up to make all the postproduction changes required in the 3,974 B-29S that were built. 87 (Table 4

provides an overview of the AAC's main bomber programs of the 1930s.) Doctrinal development in this period was mainly devoted to fleshing out the AAC's theories in greater detail and identifying particular eco-

nomic target systems for priority attack. By the late 1930s, these theories began to take on a life of their own; combat experiences in Spain, China, and other foreign countries did little to influence the evolution of air doctrine in the United States. When the war in Europe began in 1939, instructors at the Air Corps Tactical School proclaimed that the German air force

was demonstrating

tactical

school theories about the impor-

German air power doctrine was fundaAAC's; the German air force was oriented

tance of air power, even though

mentally different from the

toward

bombing

tactical

rather than strategic operations. 88

When

the Battle of Britain began in August 1940, air power theorists had the clearest test yet of the effectiveness of different kinds of bom-

bardment

became convinced that their original inclinations were correct: daylight operations were hazardous and most attacks would have to be made at night if bomber survival rates were to be kept to tolerable levels. The RAF also demonstrated that a strong defense could have a significant impact on the effectiveness of offensive operations. As a result, the number of daylight attacks on both tactics. British strategists

quickly

sides decreased as the battle progressed. 89

Although the

convinced some in the AAC that its ideas about strategic bombardment might have to be modified, most AAC strategists disagreed. As a leading study noted: Battle of Britain

The dominant view at the Tactical School prior to America's entry into World War II was that daylight bombing was essential to the whole precision idea, that bombers would usually have to fly without escort (because of the limited range of pursuit), and that bombers could provide sufficient defensive fire to permit them to accomplish their missions without high losses. Although it is not surprising that Air Corps theorists developed such ideas in the absence of actual tests, what is remarkable is the tenacity with which they held to them even when these ideas were discredited by the experience of the war. The Air Corps held to its theory of daylight, unescorted .

.

.

See The Superfortress, Historical Study No. 192, Air Technical Service Command (ATSC), April 1945; James L. Cate, "The VLR Project," in Craven and Cate, eds.. Army Air Force in World War II, vol. 5, pp. 3-32; Goldberg, "AAF Aircraft," 208-224; Krauskopf, pp. "Army and the Strategic Bomber," p. 211; Swanborough and Bowers, United States Military Aircraft, pp. 97-103. 87.

Greer, Dei’elopment of Air Doctrine, pp. 101, 109. 89. See E. Kathleen Williams, "The Air War, 1939-1941," in Craven Army Air Forces in World War II, vol. 1, pp. 92-100. 88.

and Cate,

[58]

eds..

XH

»

£ c 2 8 . £ 00 o 2r> ro XI

x

XI

Q_X

>-

l~

XI

00

X

00

o c

£ Xi £ Xi c c o £ o X _ro ro o X

XI

XI

C/5

U *

I—*

a.'

tn

‘C

x u ro >-

nj

Xo _C

ro -*—

i_r

73 •C

X

M

73

X £ o Q x IS 00

xi

"n 41 2 00

ac

ro

0»

£

>

*

(/>

XX x00 Xop > Xop — 5~ — j> 2 00 00 X x x x ^ if if

73

X ro

X c/3

ro

00

c o

U

c X o

c

X ro

oo

00 OCIH 'n OCifi

c

c X X 3 O O O °

'x

C C

Xo

xxQU

X

xX x £ S 3 X 2

73

N

X

LTO

(N

O'

Ox

ro

ri

oo

r-i

^ N ON

in

rro

O

c

o X ON

C/3

T—

.2 t 5 p. ro

2

»1

s 0 0 x i:

73

X

^ ON

rro

O'

ON ON 1

rH

rH

rH

1

ON ON ON rH

rn

rn

ON

(N| rH ,^"' '

0»

X

C/3

£

c/>

ro i~

a

00

£

Xoo

1-1

fN

rroro7

hs

li"3

1-1

roNrro^

rri

7

OnOnOnOnOnOnOnOn

i£

o k.

a X Ih 0/

£ o

X C

D kn

c a*

£ 5 a —QX > X

*-

u ro

—

c o u

O' O O O IA *t in O' r,N 7 7‘ O'O'O'O'O'O'O'O' roNfr

33

0

^

at that relatively

E Holtby, Bomb. Br., Aire. Proj. Sect -' En §- Div Se L P ‘

,

b

AtrC

A KMDc AAE, NARS;

34.

Beall

'

cited hereafter as

'

Holtby

low speed, the still-mysterious

Sect., Eng. Div.,

11 '

much

* 945 /

Memorandum

XB-47 and B-47

to Chief,

Aircraft,

Memorandum.

Memorandum.

[84]

RG

18,

Building a

Jet

Bomber

compressibility effect

was encountered. 35 As

a result, the aircraft s jet

engines were not being used efficiently. In a jet engine, fuel energy is changed into a stream of jet air. For optimal efficiency, the speed of the back of the aircraft should approach that of the jet stream coming out the of engine. Because of the compressibility effect, though, the speeds speed of the aircraft in the wind tunnel were nowhere near the potential put it, if the engine. As George Schairer, Boeing's chief aerodynamicist you go only four hundred miles an hour, you're just pushing air and 36 According to one burning up your fuel. You won't get any range." There Boeing account, "the design effort seemed to be getting nowhere. was even a question as to whether Boeing should drop out of the pic ture." 37

This might have been the end of Boeing's program and the AAF's long 38 Schairer, who was a shot except for some fortuitous coincidences. member of the AAF's Scientific Advisory Board, was part of a team that

and technological accomplishments at making the end of the war. While he was in Washington in mid-1945 new preparations for the trip to Germany, Schairer was told of some The theory, theoretical work on high-speed wings being done at NACA. tests, was that a as yet unconfirmed by experimentation or wind tunnel angle) would thin swept wing (that is, one set at a nonperpendicular swept have much less drag than a straight wing at high subsonic speeds,

was

to

analyze

German

scientific

wings might therefore neutralize the compressibility effect. The Allied operation to absorb Germany's scientific expertise. Operawar in Europe was tion Paperclip, was put into effect the instant the Allies, American scienover. The morning Germany surrendered to the Europe swooped into Germany. tific teams already positioned in assigned to the Schairer was a member of Theodore von Karmen s team across some aeronautical research institute at Brunswick, where he came them. drawings of swept-wing aircraft and some wind tunnel data on

The German data seemed to confirm NACA's theoretical calculations. The next day, von Karman and Schairer questioned Adolf Busemann, status of one of the institute's aerodynamicists, about the origins and reGermany's research in this area. Busemann replied: "Don't you Boeing (New York. Based on the accounts in Harold Mansfield, Vision: The Story 0) B-47 StratoDu ell, Sloan and Pearce, 1966), pp. 147-161; Boeing Company, "The Boeing archives; B-47 Design, pp. 4-5; Perry and jet" Nov. 1, 1959, and Jan. 15, 1965, Boeing principals at Boeing, including Holtbv memoranda. Mansfield interviewed many of the archives. Georee Schairer; transcripts of these interviews are in the Boeing Vision, p. 148. See also Constant, Origins of the Turbojet 36. Schairer quoted in Mansfield, 35

Revolution, pp. 263-265. Nov. 37. "Boeing B-47 Stratojet," 38.

This section

[85]

is

based on

1959* PP- 3~4 Mansfield, Vision, pp. 148-152. 1,

Flying Blind

member? Rome?

Volta Scientific Conference in 1935? You remember my supersonic aerodynamics? It told how sweepback would re-

paper on duce the drag at supersonic speeds. ... No one paid any attention. Finally Messerschmitt ran some tests and everybody got excited. The Me-262 is under construction now with forty-five degrees sweepback."^ It

was

a coincidence that

and high-speed

NACA and German research on swept wings

came to the attention of Boeing's chief aeroBoeing was trying to redesign the B-47. These were and unforeseen technological windfalls clearly not

flight

dynamicist just as entirely fortuitous

AAF when AAF bomber

anticipated by the

—

it

established

its

speed requirements

Nor did NACA's (or Germany's) 1944.

in

requirements influence the course of basic aerodynamic research in any way. The irony of the story is that Schairer himself had known about Busemann's paper earlier and subsequently had forgotten about it. If Schairer's discovery had come even six months later, it is possible

program would have already been canceled or that Boeing would have already irrevocably committed itself to building a straightwing aircraft. As it was, Boeing was in an extremely flexible position; it was not deeply committed to any particular design ^in mid-1945. Andat this stage, both Boeing's and the AAF's sunk costs were minimal due to that the B-47

the sequential strategy that guided the program. Schairer returned from Germany in August 1945, and Boeing immediately set out to elaborate on his findings and redesign its jet bomber. Boeing moved quickly and made several successive proposals

in September, October, and November (see Table 5). Model 446 was the first of Boeing's swept-wing designs; based on Model 432, it had four engines

housed

in the fuselage. Boeing's next proposal.

Model 448, had a swept two additional engines in the aft fuselage. Sweeping the wings and tail dramatically improved aerodynamic efficiency and made it worthwhile to add two more engines to the aircraft (see Figure 3). Some relatively minor changes for improved visibility resulted in Model 448-2-2. The net effect of these changes was an estimated top speed of m ph, quite an improvement over the estimates from the summer's 555 earlier wind tunnel testing and slightly better than the "desired" perfor40 mance specified by the AAF. One final, major change in the design of the B-47 was still needed. The most serious problem with Model 448-2-2 was the dubious legacy of Model 432: the aircraft's engines (and there were now six of them) were buried in the fuselage. As it stood, the proposal was still unacceptable to tail

as well as

39.

Busemann quoted

40.

Perry and Holtby

in ibid., p. 151.

memoranda.

[

86

]

Building a

Jet

Table 5.

Bomber

XB-47 designs

Model number 413 422 424 425

426 432 446 448

Wing

Jan. 1944 early 1944

Straight

4 turbojets

Straight

early 1944 early 1944

Straight

4 turboprops 4 turbojets

Straight

4 turbojets

early 1944

Straight

Dec. 1944 Sept. 1945 Sept. 1945 Sept. 1945 Oct. 1945

Straight

4 turboprops 4 jets in fuselage

448-2-2

430 Sources:

Engines

Date

W.

4 jets in fuselage 6 jets in fuselage 6 jets in fuselage

Swept Swept Swept Swept

E. Beall, "Status of Jet

6 jets under wings

Bomber Program-Model

432," Memorandum to C. L. Egtvedt, April 19, 1945; Boeing for archives; Robert L. Perry, "Air Force Selection of the B-47

Development and Production," Memorandum, Nov. 1954; Files,

5,

HO/ASD.

concerns was Model 450, which was presented to the AAF in November 1945. Boeing moved the airbeneath the wings. This kept craft's six engines into four nacelles slung important aerodynamic the leading edges of the wings clean, still an and the consideration. The inboard nacelles housed two engines each,

the

AAF. Boeing's response

to these

outboard nacelles were virtually on the wing

tips.

41

Model 450 was aeronautically exotic in several respects. The swept Boeing wing was completely new to American aeronautical experience. aerowas still conducting basic wind tunnel research on swept-wing Model dynamics in late 1945 while it was finalizing the basic design for swept-wing aircraft built in the United 450. The first flight of the first eventually take place just a States, the North American XP-86, would weeks before the first flight of the B-47 itself in late *94 7- Not onl Y few

swept was the swept wing new to American aerodynamicists, but a large, not wing had never been used before by anyone on an aircraft this on the cutting even on an experimental model. The B-47 was definitely edge of

Although swept-wing military and commercial aircraft, this

this particular technological revolution.

designs later became

common

for

was a leap into the technological unknown in 1945Basic aeroThe B-47's wings were also extremely long and thin. not been dynamic problems such as stall, instability, and flutter had regime, and these exhaustively investigated for the high-speed flight wing. The selecproblems were especially pronounced with this kind of

41.

Ibid.

See also

SAC/HO,

B-47 Design, pp. 4-5-

Archives)

Company

(Boeing

designs

XB-47

Boeing

5.

Figure

Building a

Jet

Bomber

Boeing B-47 (Boeing

Company

Archives)

wing also ran up against an engineering trade-off: alwing, a though a thin wing was aerodynamically cleaner than a thick fatigue. Strength and thick wing would stand up better to stress and especially in the durability were important engineering considerations, especially probdesign of a weapon system; the B-47's novel wing was fuel tanks or lematic in this regard. Finally, it was impossible to house design of the landing gear in a wing this thin, which complicated the

tion of the thin

fuselage.

because the Putting the engines out on the wings was risky, especially outboard enoutboard engines were on the wing tips. The loss of one thin wing presented gine could create serious yaw problems. The long, and the B-47's engine control problems under the best of circumstances, questions placement only compounded this. In addition, there were

and fatigue about the thin wing s ability to stand up to the stresses engine placement once associated with carrying six engines; the B-47's engineering prob again compounded what was already a formidable

42 highly experimental. lem. In short, the B-47 was, in man Y respects, conference on The unanimous conclusion reached at a Wright Field November 1 was that Model 450 was far superior to Model 448-2-2. just a was concluded that the B-47 went a ste P beyond the

few days

later

B-45, B-46,

and

it

B-48. Estimates of the B-47's top

speed ranged from 622

and technological advances, see For a discussion of Model 45«'s design features (London: Putnam, 1966), pp. 323-33]- See also eter M. Bowers, Boeing Aircraft since 1916 Memorandum for the Asst. Chief of Air Staff Jen Alfred R. Maxwell, Chief, Reqs. Div., AAF, NARS; cited hereafter as Maxwell lay 21, 1946, XB-47 and B-47 Aircraft, RG 18, 42

/lemorandum.

Flying Blind

mph

to 635

mph, depending on

the ultimate weight of the aircraft. 43

The

engineering division's report to Washington observed:

The newly proposed XB-47 represents

advancement in the state of the art of airplane design by applying the use of a swept wing to bombardment aircraft. Swept wings are currently being proposed for fighter airplanes and it is believed that this new means of increasing performance should also be applied to a bombardment airplane in an effort to minimize the difference between fighter and bomber speed. The use of a swept wing on the XB-47 permits successful exploitation of a definite

.

.

increased engine ratings to attain a speed of 635 mph at 35,000 feet. Increased engine ratings cannot be utilized effectively by other jet bombers currently under contract due to serious compressibility problems encountered with a straight wing at high Mach numbers. 44

The

AAF was

fortunate that

procurement strategy involved competition in the development phase of the acquisition cycle. It did not have to ask North American, Convair, or Martin to redesign their bombers to take swept-wing aerodynamics into account, with all the costs and delays that would have entailed. It had another contractor trying to maximize the high-performance track, and Boeing had not yet begun building

have

felt

its

prototype. In the absence of this option, the AAF would obligated to redesign one of the bombers already under con-

its

struction, regardless of the cost.

And

if

production tooling had been

purchased concurrently with prototype construction, the cost of taking aerodynamic developments such as these into account would have been enormous.

Production Decisions

By the end of 1945, the AAF's gamble on the B-47 was about to pay off. The B-4/s performance capabilities would not be certified through flight testing for another two years, but all preliminary data indicated that the B-47 's performance would be far superior to that of the other bombers in the competition. Although the AAF had reasons for being optimistic about the B-47 at the end of 1945, it did not cancel the other programs outright. That would have been premature, especially given all the aerodynamic mysteries that had plagued and continued to haunt the jet bomber program. Successful flight testing could not be taken for granted in the case of an exotic aircraft such as the B-47, and, until all the aircraft Conference of Nov. 43 B-47 Aircraft, RG 18, AAF, -

44-

Bomb. Br„

1,

1945,"

NARS.

and "Conference of Nov.

6,

1945"; both in XB-47 47

Aire. Proj. Sect., Eng. Div., Letter to Asst. Chief of Air Staff, Aircraft, 18, AAF, NARS.

1945, XB-47 and B-47

RG

[

90

]

Nov

and 14

Building a

had taken

Jet

Bomber

to the air,

no

reliable data existed

about their relative ca-

pabilities.

Fortunately for the AAF, these four development programs were remarkably inexpensive. North American received approximately $13 mil45 Convair lion for developing the B-45 and building three prototypes. 46 The AAF paid Boeing got about $7.3 million for its work on the B-46. approximately $10 million through the end of 1947 f° r f w ° B 47 prototypes and receive a

its

extensive

maximum

wind tunnel

47 Martin testing program.

of $7.5 million for

its

was to two B-48 prototypes. 48 Even

allowing for the cost of government-furnished equipment, which might have been equal to the contract costs paid to the contractors, and allowing for the various contract change orders that affected contract costs was well slightly, the cost of the entire competitive prototype program

under $100

million. 49

This kind of parallel, competitive prototype program was affordable even though defense budgets were extremely lean in the early postwar demoyears. At the end of the war, the U.S. military establishment was Expenditures on bilized and military spending was drastically reduced. billion the armed services in fiscal years 1947-5° averaged only $12-13 per year. Once this was distributed to various accounts, the Air Force develreceived an annual average of about $155 million for research and opment in the early postwar years. This was "not enough to start new programs and barely enough to keep alive those started with wartime funds." 50 But even in the midst of this fiscal famine, it was possible to

fund a full-scale prototype competition. 47. Four jet bomber programs were affordable because a decision had been made not to integrate military subsystems in the aircraft until the include the cost of the 139-aircraft cost of the XB-45 prototype program did not Files, HO/ASD. procurement program; see "XB-45, North American, no date. because three of the four aircraft specified 46. The B-46's contract history is complicated canceled in 1946. The money saved on the B-46 in the original $10.4 million contract were The final cost of the one B-46 procontract was transferred to other Convair programs. in line, as one observer put it, with totype was therefore substantially less than $10 million, see Farry Outhne^See also Aire Pro). the $7.5 million cost of the B-48 prototype program; XB-48 Jet Bomber, RG 18, AAF, NARS. Sect Eng Div., Wire to CG, AAF, June 30, 1945. program are summaThe convoluted contract negotiations over the B-47 prototype Aircraft, HO/AMC, July 1950, pp. 3“7rized in Margaret C. Bagwell, The XB-47 and B-47 Bomber, RG 18, AAF NARS. 48. Letter Contract, June 30, 1945/ XB-48 Jet government-furnished equipment are found in T. A. Marschak, 49. Cost estimates for Rand Corporation Paper, P-2850, Jan. 1964, The Role of Project Histories in the Study of the cost of the entire prototype competition agrees p 103 The $100 million estimate for in Development, Rand Corporation with that in B. H. Klein et al., The Role of Prototypes

The

45.

'

R&D,

Report, R-333, Dec. 1958, p. 12. ,n the United States Air Force 1907-1957 (Pnncetom History A ed., of Goldberg, Alfred 50 Hearings before the Senate Armed Van Nostrand, 1957), p. 199. See also Study of Airpower, .

Services Committee, 84th Cong., 2d sess., April-June 1956,

p. 541.

.

Flying Blind

prototypes had been tested and the designs of the aircraft finalized. This is not to say that subsystem development was ignored in the late 1940s. To the contrary, a great deal of preliminary

work was done on bombing, navigation, radar, defensive armament, and other critical military subsystems. But, since only one or perhaps two bombers would eventually be produced, it would have been a waste of scarce financial resources to put together four complete weapon system packages. In addition, subsystem integration would have been counterproductive if it took money away from the prototypes themselves. These financial considerations were reinforced by powerful technical constraints. Until the prototypes flew and their actual capabilities became known, the performance parameters for bombing and navigation subsystems, for example, could not be defined. This was especially true in the case of the highly advanced B-47. Not only would it have been premature to rush in with fullscale subsystem development and integration at an early stage of development, it was impossible to do so in any meaningful way. The next few years, 1946-49, were devoted to building prototypes, evaluating their flight test programs, and making final production and cancellation decisions. In 1946/ the AAF concluded that it was necessary to begin producing a jet bomber immediately. Three considerations led to this decision.

was

growing conviction in the AAF that dramatic advances in fighter capabilities were making older, propeller-driven bombers obsolete. The AAF itself was already applying swept-wing aerodynamics to fighter design and developing the high-performance F-80, F-84, and F-86. It was therefore aware of the kinds of fighter capabilities that could be expected in the years ahead. The emergence of sweptwing jet fighters made the deployment of high-speed bombers more urgent than ever. Although the AAF had a massive inventory of literally thousands of bombers used during the war, propeller-driven bombers were not likely to go faster than 400-500 mph; there were limits to the improvements that could be made to the B-29 and its derivative, the First,

there

a

B-50. 51

Although the AAF's inventory of World War II aircraft was massive, most of it was put into storage immediately after the war. SAC's active forces were actually quite small in the early postwar years. Relatively few aircraft groups existed even on paper, and the vast majority of these groups were not ready for combat. Of the aircraft groups the AAF had 55 on paper in December 1946, only two were in a high state of readiness, and only one of these was a heavy bomber group. Moreover, the AAF had only B-29S modified' to carry 27 the atomic bomb in 1946. See Harry R. Borowski, A Hollow Threat: Strategic Air Power and 51.

Containment

before Korea (Westport, Conn.: Greenwood Press, 1982), pp. 48, 103; SAC/HO Dei’elopment of the Strategic Air Command, 1946-1986, Sept. 1986, pp. 1-20; Herman S. Wolk, Planning and Organizing the Postwar Air Force, 1943-1947 (Washington: Office of Air Force History, 1984), chap. 2.

Building a

Jet

Bomber

Second, the Soviet Union was beginning to make dramatic strides in air defense capabilities. Although the Soviet air defense system was

end of the war, the Soviet Union (like the rest of the Allies) received an enormous technological windfall when German aircraft development and production facilities were captured in 1945. In fact, much of the German aviation industry had been moved eastward in the last years of the war, to avoid British and American technologically primitive at the

bombing

attacks.

tion facilities

fell

As

a result, 80 percent of all

into Soviet

hands

at the

end

German

aircraft

produc-

of the war. In addition, the

Union acquired copies of complete German systems (including the Me-262, advanced turbojet engines, and advanced electronics) and captured over 300,000 German aircraft workers and technicians as well as entire aeronautical research staffs. The first flights of two Soviet jet fighters, the Yak-15 and MiG-9, took P lace in April 1946. Both aircraft were powered by copies of German engines; the service configuration of the MiG-9 had a top speed of 560 mph. Stalin is said to have personally ordered the designers of these two aircraft to have fifteen copies of each ready to fly over Red Square by that November. The requirements for a more advanced, high-altitude interceptor were also set in 1946. These requirements led to a three-way competition that ultimately produced the MiG-15, which first flew officially in December 1947. The MiG-15, which compared well with the American F-86, could fly at 675 mph with a service ceiling of 50,000 feet, making it a fairly formidable aircraft. Approximately 15,000 MiG-i5S were built in the late 1940s and early 1950s. German research on swept-wing aerodynamics and high-speed 52 flight was critical in the MiG-15 program. Although it is not clear exactly what the AAF knew about the status of the Yak-15, MiG-9, anc* MiG-15 programs in 1946, the AAF was aware of the technological windfall the Soviet Union received from Germany. After all, the AAF itself had studied Germany's aircraft programs carethat the fully in the months immediately after the war. The AAF knew Soviet Union had come into possession of highly advanced jet fighters, turbojet engines, and swept-wing technology. Even if the AAF was in the dark about specific Soviet fighters and flight tests, it knew much

Soviet

about the broad technological parameters of the Soviet

air

defense

effort.

The AAF surely knew in 1946 that jet fighters and swept-wing aerodynamics were in the Soviet air defense forecast. In any case, the AAF York: Day, 1962), pp. 109For more details, see Asher Lee, The Soviet Air Force (New Air Power (New York: Praeger, 1962), pp. 213119; Robert A. Kilmarx, A History of Soviet Force since 1918 (New York: Stein and Day, 1977), 233; Alexander Boyd, The Soviet Air Armitage and R. A. Mason, Air Power in the Nuclear Age (Urbana: j pp 205-217R. Whiting, Soviet A,r Power University of Illinois Press, 1985), PP- 144-150; Kenneth _ i 2 9(Boulder, Colo.: Westview Press, 1986), pp. 37~4°' i 2 7 52.

M

Flying Blind

had

performance requirements

set its

when

the situation

for the jet

bomber back

in 1944,

was

substantially less clear. These requirements significantly in the postwar years even though fighter

were not changed capabilities improved dramatically.

Third, as Soviet-American relations became increasingly strained, it seemed to be increasingly important for the AAF to have modern jet

bombers, not just under development, but would be more capable than older systems

bombers both conventional and

in the field. Jet

in

nuclear operations. Although the B-47 was expected to play both roles effectively in the long run, it would not be ready for deployment for

some

time. 53

With these considerations in mind, the AAF decided in 1946 to review its four jet bomber programs to see if one could be put into production at an early date. According to a memorandum for the secretary of the air force, the July 1946 decision in favor of the B-45

was

fairly straightfor-

ward. It was estimated that the B-47 an d B-48 flight test programs would not be completed for two to two and one-half years, which disqualified these aircraft as early deployment options. The B-45 and B-46 were expected to have virtually the same top speed (481 mph for the B-45; 4 78 mph for the B-46); either would be 100 mph faster than the B-29S then in service.

The

however, had distinct advantages over the B-46 in terms of payload and electronics-carrying capabilities. Moreover, since the B-45 involved fewer developmental risks than the B-46, it was less likely to be delayed by technological snags. This was an important consideration given that early deployment was the whole point of the exerB-45,

cise. 54

Since the B-45 production decision was made prior to actual flight testing, this was a step back from the sequential strategy that had guided the jet bomber program for years. The AAF did not,

however,

abandon the sequential

strategy at this point.

The B-46 was not canceled outright, although its development program was cut back; one B-46 prototype would still be built and a B-46 production program would still be available as a fallback option. 55 More important, the competitive flyoff was still the focal point of the overall program. 53. The B-4 /s payload for the atomic bomb; see

requirements did not change when it was designated as a carrier Gen. B. W. Chidlaw, Dep. CG, AMC, Letter to W. Stuart Symington, Secretary of the Air Force, March 23, 1949, XB-47 and B-47 Aircraft, RG 18 AAF NARS; cited hereafter as Chidlaw Letter. See also Heavy Bombardment Committee/ USAF Aircraft and Weapons Board, Report on Heavy Bombardment, Nov. 7, 1947, p. 6, Long-Ranee

Bomber Subcommittee Report, RG 54.

Chidlaw

341,

HQ/USAF, NARS.

Letter.

See also Farry Outline. When the B-45 was accelerated, four elements of concurrency were injected into the program: the design of the system was frozen; paper studies were relied on for important design decisions; a substantial financial commitment 55.

Ibid.

[94]

,

Building a

Jet

Bomber

programs began in 1947, less than three years after full-scale development got under way. The B-45 first flew on March 17; the B-46 took to the air on April 2; a B-48 prototype flew on June 22; and the B-47, which was completely redesigned in late 1945, began its flight test program on December 17, the 44th anniversary of the Wright brothers' flight. The flight test programs lasted for months, and a solid body of data about the relative strengths and weaknesses of the aircraft began to emerge in early 1948. At this time, the leadership of the newly independent U.S. Air Force evaluated the programs and determined that the B-47 best filled the requirement for a high-speed bomber. Once the evaluation of the fly-off All four flight test

was completed, the B-47

B-47A

placed into

fie

aircraft

Command (AMC) recommended that production quickly. 56 A contract to build ten

the Air Materiel

was signed

wo n

in

September

1948.

was much faster than the other contending aircraft, even the B-48. The B-47's swept wing neutralized the compressibility effect, as predicted, allowing the bomber to and at high speed. knife through the atmosphere relatively efficiently The B-47

the competition because

it

—

an era when

defenses consisted primarily of fighter aircraft, speed and defensive armament were the keys to survival for bombers. The B-47 was actually fast enough to outrun the fighters of its day, a useful 57 capability not enjoyed by the other three bombers in the competition. As a result, the only defensive armament it needed was a tail gun. The In

air

shortcoming was its range, which was expected to improve as its design matured; this issue became less critical once the Air Force 58 In any event, the Air instituted aerial refueling in the late 1940s. B-47's only

Force's

main

interest in the B-47

was speeds 4 As

Secretary of the Air

Force Stuart Symington put it, "speaking as the Secretary of this outfit, I wouldn't want to consider my son's life in an airplane from the stand60 The B-47's superior point of T know it is slower, but it uses less fuel'."

performance was enough

to

win the competition, even though the B-45

program even though it was still under development; and an early production decision was made. 56. Chidlaw Letter. The Air Technical Service Command became the Air Materiel Command in 1946. For more details, see the appendix. 57. Building a bomber that could outrun fighters had been a goal of bomber advocates since the days of the B-17 in the mid-i930s, as discussed in Chapter 2. 58. The Air Force began to emphasize development of aerial refueling capabilities in 1947, and its first aerial refueling exercise was conducted in 1948; see Heavy Bombardment Committee, Report on Heavy Bombardment pp. 1-7. 39. The Air Force believed that a high-speed penetrator was needed to complement a force of longer-range, but slower, bombers. See Chapter 4 for more details on the develop-

was made

ment 60.

RG

to the

of these longer-range aircraft.

"Transcript of Air Force Conference," Pentagon,

18,

AAF, NARS.

March

31, 1949,

XB-48

Jet

Bomber,

Flying Blind

was

a year or

two

farther along in the acquisition cycle.

The Air Force

opted for performance over faster delivery. Neither the B-48 nor the B-46 was put into production. Martin tried to resurrect the B-48 in 1949, but the Air Force declined to do so even though Martin needed the business. Indeed, Martin was on the verge of receivership when the B-48 program was canceled. 61 In the end, only one B-46 and two B-48 prototypes were built. The Air Force decided in July 1948 to halt B-45 production with the 139th production model. It did not want the B-45 1 ° interfere with the B-47. 62 In early 1949, the Air Force also canceled production of the B-54 medium-range bomber, an improved version of the B-50, which was in turn an improved version of the B-29. The B-45 an d B-54 cancellations helped to free more money for the B-47, which was essential given the severe budgetary constraints the Air Force was operating under at the time. The Air Force had to cut its planned force structure from 55 to 48 groups to stay within fiscal 1950 budgetary limits. Buying new bombers was obviously difficult under these conditions. 63

Development of the B-47 continued through 1948 and 1949 as its design was changed and refined. A more powerful engine was worked into the equation, for example, and engineering changes were made to the fuselage, wings, and landing gear. Still, the B-47 program was being conducted sequentially. The September 1948 contract authorized the procurement of only ten B-47A aircraft, along with the machine tools needed to build just these ten aircraft. The B-47A was to be used for further flight testing and, later, for training purposes. It was never intended to be the ultimate operational version of the aircraft. The next version of the aircraft, the B-47B, was to be the full-fledged weapon Since the B-48 was the only major Air Force project Martin was working on at the time, the cancellation of its contract meant that Martin would lose the B-48 contract money outright; there was no other Air Force project to which the money could be transferred. Even though Martin was aggressive in trying to resurrect the B-48, the Air Force was unwilling and unable to continue spending money on the program; budgetary constraints were especially severe in early 1949 as the fiscal 1950 budget shaped up. In both the medium-range and long-range bomber competitions in the late 1940s, the Air Force opted for the contractor that was able to provide the best aircraft performance (Boeing, in both cases), even though this resulted in an uneven distribution of contract dollars and an end to the Air Force's long-standing working relationship with Martin. For a discussion of Martin's desperate financial condition in the late 1940s, see John B. Rae, Climb to Greatness: The American Aircraft Industry, 1 920-1960 (Cambridge: MIT Press, 1968), p. 185. See also "Transcript of Air Force Conference," March 31, 1949; Report, Second Meeting of the Board of Senior Officers, USAF, Feb. 21-24, * 949 RG 341, HQ/USAF, NARS. 61.

/

62.

Vandenberg, Chief of Reqs.,

Procurement and Indus. Planning, Memorandum for Gen. July 22, 1949, and Gen. E. E. Partridge, Dir. of Training and

Col. R. A. Elliot, Dir. of

Memorandum

Staff,

for Chief,

Reqs. Div., July

1,

1948; both in Files, B-45 Airplane,

HO/AFLC. 63.

See Report, Second Meeting of the Board of Senior Officers.

[96]

Building a

Jet

Bomber

system that would be produced in quantity and delivered to SAC. Since the program was still proceeding sequentially, the design of the aircraft was still somewhat flexible, even at this late date, and changes could be made without discarding a lot of production tooling. By the end of 1949, the Air Force had placed an order for 87 B-47BS, along with the production tooling for these aircraft, and the program was moving smoothly in the direction of large-scale production. 64 As it turned out, the transition from development to production was anything but smooth.

Accelerating the B-47 into Production

The successes

development program were not automatiproduction program. Whereas the B-47's develop-

of the B-47's

cally transferred to its

ment program was noted for its technological accomplishments, brisk schedule, low cost, and performance demonstrations, its production program was long remembered for its wide range of problems. The basic problem with the B-47 production program was that subsystem development and production lagged while huge numbers of airframes were coming off the assembly line. In some cases, interim versions of these subsystems had to be installed while development continued. In other cases, even interim systems were not available, so

unfinished aircraft shells were simply set aside

many

when

they came

off

the

subsystems were plagued by problems of a more routine kind: equipment was deficient, defective, or unreliable. As a result, hundreds of B-47S needed expensive and timeconsuming modification and retrofitting before they could be used by SAC. The two biggest problems involved the defensive armament and bombing-navigation subsystems. Even though the key to the B-47's survivability was its speed, SAC felt strongly that the B-47 needed some defensive armament, at least a tail gun. On operational questions such as these, SAC's assessment usually carried the day. SAC's position on this issue was rooted in strategic bombardment doctrine, which had maintained since the 1930s that bombers should be heavily armed. The problem in the case of the B-47 was that no tail turret was available in 1951, and there was little prospect of perfecting even an interim system for at least another year after that. As far as SAC was concerned, this made the B-47 "unsuitable for combat operations." 65 Eventually an interim tail turret was installed in over 300 B-47S, and a more capable system was worked into the production picfactory floor. Finally, in

cases,

Bagwell, XB-47 and B-47 Aircraft, pp. 4-7, 10-13. 65. SAC/HO, B-47 Design, pp. 78-82. 64.

Flying Blind

ture starting with the 400th B-47. At that point, the Air Force retrofitted

more advanced system into the first several hundred B-47S. 66 The defensive armament subsystem, therefore, was not incorporated into the B-47 production program in a particularly efficient or timely manner. The B-47's bombing and navigation subsystem was also problematic. This system was vitally important; the bomber's mission could not be performed without it. The idea of conducting precise attacks on carefully selected targets had been a central feature of strategic bombardment doctrine since the 1930s, and it continued to influence Air Force doctrine in the atomic age. The B-47's impressive speed and altitude capabilities made accurate bombing extremely difficult, however. Exceptional demands were placed on the bombing system, and it consequently met with several development problems. These problems in turn disrupted the overall production program and degraded the operational effectiveness of the B-47 f° r several years. The reliability of the system was an especially serious problem. One SAC wing commander noted, "With the bombing system we have installed in the B-47, can say is the B-47 is a lovely flying airplane, but is not worth anything as a bomber the

I

because it is going to have about 60 percent electronic failures." 67 SAC predicted in 1931 that the system would have a 70 percent abort rate. Remarkably, these kinds of problems continued for years. A Boeing study reported in March 1954 that the B-47's bombing and navigation system was responsible for reducing combat efficiency 30 percent, and a

system continued

be the major limiting factor on the B-47's combat capability. In addition, the problematic interim version of the system was in short supply for many years. The more advanced version was in short supply as late as 1955. As a result, even the B-47 's retrofitting schedule slipped repeatedly. 68 The program's original deployment schedule, of course, quickly fell by the wayside. Problems also emerged with the B-qy's autopilot, ejection seats, aerial refueling equipment, fuel cells, jet-assist and rocket-assist equipment for take-offs, and drag chutes for landings. Some canopies cracked and blew off, and even the windshield wipers had problems. By September 1931, the Air Force had identified 93 major problems areas, 44 of which were considered to be critical items that required immediate action. 69 later report stated that this

to

Mason

Kipp, History of B-47 Procurement, Production, and Modification, 1 953-1956, Historical Division, AMC, Jan. 1957, p. 44. See also Margaret C. Bagwell, History of the B-47 Production Program, 1949-1953, Historical Division, AMC, Aug. 1954, pp. 41-46. 66.

R.

Quoted

SAC/HO,

B-47 Design, pp. 86-87. 68. See Kipp, History of B-47 Procurement, pp. 39-40. 69. USAF, Air Research and Development Command (ARDQ/HO, History of the Air Research and Development Command, July 1951 to Dec. 1952, vol. 2, p. 25. See also Bagwell, History of the B-47 Production Program, pp. 19-53; Kipp/ History of B-47 Procurement, pp. 3767.

60;

in

Marschak, Role

of Project Histories, pp. 103-105.

[98]

Building a

Jet

Bomber

The B-47's production problems were so extensive that the Air Force set up a series of modification centers apart from the production lines where the aircraft were actually being built. B-47S were sent to these centers as soon as they came off the assembly line in order to have latearriving equipment installed. It took a minimum of eight months and often a full year to turn the aircraft into operational weapon systems. Later (sometimes years later), the aircraft were sent back to the centers to have advanced versions of critical subsystems retrofitted. The modification effort added at least 10-20 percent to the cost of the program. 70 Why was the B-47 production program so unsuccessful? Many in the Air Force blamed the sequential procurement strategy that guided the program in the 1940s. The charge was that subsystem development had been neglected in the 1940s and that the Air Force was paying the price approach in the 1950s. According to this line of thinking, military subsystems and components should have been perfected along with the airframe and integrated into the overall effort at a much earlier date. 71 In other words, the Air Force argued that more concurrency in the 1940s would have prevented many production problems in the 1950s. In fact, the problem with the B-47 production program was that the sequential, successful procurement strategy employed in the 1940s was abandoned in the 1950s. The B-47's production problems were not caused by a lack of concurrency in the 1940s; the problem was that a great deal of concurrency was introduced into the program in the 1950s. Three main points need to be stressed in this regard. First, subsystem development for the B-47 was technologically challenging precisely because Boeing's development efforts were so successful. The B-47 's high-speed capabilities made the development of bombing subsystems, ejection seats, autopilots, deceleration chutes, windshield wipers, and the like problematic. Moreover, it was impossible to know precisely what kinds of capabilities would be needed in these subsystems until the aircraft's flight capabilities were known. As it turned out, the B-47's performance far exceeded earlier predictions. Boewe bad no idea ing's George Schairer noted: "In the case of our B-47 what would turn out when we built the airplane. At least we had no for this excessively sequential

•

•

•

Bagwell, History of the B-47 Production Program, pp. 53-58. Another study estimated that retrofitting added 25 percent to the cost of the aircraft; see Marschak, Role of Project Histories, p. 105. For more details on the B-47's modification efforts, such as Project IRAN (Inspect and Repair As Necessary), see Kipp, History of B-47 Procurement, pp. 61-70. 71. "An Analysis of Development, Procurement, and Production of Selected Weapon Systems," no date. Files, Office of Air Force History. Another Air Force critique of the B-47 program can be found in Col. J. F. McCarthy, Jr., Aire. Br., AMC, "Application of Slow 70.

Build-Up Concept,"

Memorandum, May

7,

1953, Files,

HO/ASD.

Flying Blind

would turn out as well as it did. There was no way to predict that it was going to come out that well. We thought we were terribly optimistic in our predictions. We guessed our drag would be about twothirds of what anybody in his right mind would have estimated, and it

idea

it

turned out to be another 25 percent lower than our optimistic estimate/' 72 It consequently would have been counterproductive to finalize subsystem performance requirements prior to 1948. Indeed, it was simply impossible to do so in any meaningful way prior to that time. Second, subsystem development was not ignored between 1944 and 1948. In fact, preliminary work on a bombing system began in 1944 and a contract for development of a defensive armament system was awarded in 1946. One problem, at least in the case of the defensive armament system, was that the Air Force decided in late 1948 to require a radaroperated remote fire control system for the B-47 instead of a simpler, manned tail turret. This electronically operated gun-aiming device proved to be enormously difficult to develop and the Air Force was later 73 A criticized for trying to take such a "long step forward" in this area. big part of the problem, therefore, was the Air Force's decision to push the state of the art in defensive armament at the same time that the B-47's high-performance capabilities were making subsystem development inherently challenging. This special requirement compounded the subsystem development problem. In addition, the Air Force relied on a single contractor for the development of this subsystem throughout the late 1940s, even though the competition built into the aircraft development program played a central role in the success of that program, as we have seen. Ironically, when some competition was injected into the de-

began to make some progress. 74 Therefore, the Air Force's failure to employ a competitive, sequential strategy in its defensive armament program was at least partially responsible for the problems that later emerged in this area. Third and most important, the B-47's production problems grew out of the sudden and dramatic acceleration of the program that took place starting in 1950. It is important to remember that the B-47A was intended to be a final developmental prototype and that the B-47B was fensive

armament program

in 1951,

it

intended to be the first production version of the system. In November 75 Al1949, only 10 B-47AS and 87 B-47BS were scheduled to be built. though it was expected that more B-47S would eventually be ordered,

Major Systems Acquisition Reform, Hearings before the Senate Government Operations Committee, 94th Cong., 1st sess., pt. 2, July 1975, p. 100. 73. Bagwell, History of the B-47 Production Program, p. 45. See also Bagwell, XB-47 and 72.

B-47 Aircraft, p. 20; Kipp, History of the B-47 Procurement, p. 37. 74. Bagwell, Histon/ of the B-47 Production Program, pp. 41-46. 75. Bagwell, XB-47 and B-47 Aircraft, p. 12.

[100]

Building a

Jet

Bomber

the short-term procurement plan called for a gradual expansion of B-47

production facilities and a smooth transition into mass production. The severe budgetary constraints imposed by the fiscal 1949 and 1950 defense budgets kept the B-47 program from moving too quickly into mass production. The Air Force's Board of Senior Officers, for example, recommended in 1949 that tooling for producing B-47S at a rate of thirty per month be procured as funds became available. 76 Funding for large-scale production of the B-47 was not available in 1949, however, and so the production plan was unambitious. The first test flight of the first B-47A took place on June 25, 1950, which coincided with the beginning of the Korean War. The B-47 P ro “ duction program subsequently began to expand dramatically. There seemed to be an urgent need to get large numbers of jet bombers deployed quickly. Modern jet aircraft would be needed in Korea itself, and many believed that the war could spread to Europe. 77 Moreover, the Soviet Union's successful test of its first atomic bomb in late 1949 demonstrated that

it

was

in the process of

making great technological

strides.

This worked to heighten American anxiety about the technological com-

between the superpowers in general and, more immediately, to make deployment of the B-47 seem like a short-term imperative. In addition, the budgetary resources for a large procurement program suddenly became available. These demand-side and supply-side considerations worked together to make large-scale B-47 production both (apparently) necessary and (definitely) possible. The production program subsequently expanded: by December 1950 the Air Force was making plans to open two additional production lines at Lockheed and Douglas 78 facilities to help build the 1,500 B-47S it had already ordered. The sudden, massive expansion of the B-47 production program would have been a bold move even if the aircraft and all its subsystems had been fully developed and were ready for mass production. Even in this best-case scenario, one would have expected production problems, such as shortages of parts and trained manpower, to emerge. But neither the B-47 nor its subsystems were ready for mass production in 1950. Basic developmental problems with essential subsystems persisted well petition

into the 1950s, as

we have

seen.

As

for the aircraft itself, the first test

June 1950, and the first test flight of the first B-47B (the first production model) did not take place until April 1951. By that time, the Air Force had placed production orders for 1,500 flight of the

B-47A took place only

in

Second Meeting of the Board of Senior Officers, p. 9. discussion of the immediate operational problems raised by the war in Korea, 77. For a see Armitage and Mason, Air Power in the Nuclear Age, pp. 20-45. 78. E. Clifford Snyder, History of Production Problems during the Air Force Build-Up, 19501954, Historical Division, AMC, Jan. 1956, pp. 186-189. 76.

Report,

[

101

]

Flying Blind

B-47S. Filling these orders

would have been problematic even

if

the

design of the aircraft had already been frozen, but the design of the B-47 was still being refined. By the end of 1950, when the Air Force was gearing up for quantity production, a total of 636 engineering changes had been made in the B-47A and 1,702 changes had been ordered for the B-47B. Boeing estimated in 1954 that 3,428 changes were made in the The B-47 was still in B-47 during the production phase of the program.

when

the process of being developed

the Air Force rushed

into pro-

it

Extensive retrofitting was inevitable under these circumto insist that destances. The decision to accelerate the B-47 program velopment and production take place concurrently was a prescription duction.

— —

for disaster.

The deputy inspector general of the Air Force described what happened as the B-47 production program unfolded: was superimposed upon a program of concurrent experimentation and service testing and suffered from the beginning. The operational effectiveness and capabilities of the great amount and variety of new and complex equipment needed for installation in the B-47 seems to have been taken for granted, although most of the engineering was still in the design or highly experimental stages. The result was a nearly complete breakdown in the coordination of the various phases of the overall program and when one phase slipped or surged, no effort was made to adjust any or all of the other phases to the new condition. The entire program was totally out of phase. The acceleration was accomplished at the expense of quality, workmanship, continuity of production, and interchangeability of parts. Added cost, confusion in the factory, and further slippages in the actual delivery schedule of completed units was inevitable. One of the chief causes of this condition was the insistence on the part of The production .

.

of B-47 airplanes

.

.

.

.

.

.

.

.

.

.

.

.

.

the Air Force that production of subject aircraft, unsuitable for tactical pur-

The attempt

produce the B-47 an d its components as a fully operational airplane concurrent with the development and service testing, the training of maintenance and operating personnel, the development and production of support equipment, the acquisition and development of necessary training and operating bases, all on an accelerated plan, was a program of such magnitude and complexity that it taxed the capabilities of the Air Force. Inasmuch as the B-47 was °f radically new design and configuration, extensive and careful planning should have been devoted to these matters. 80 poses, be continued at an accelerated rate.

.

.

.

to

Boeing, “Resume of the Development of the B-47 Production Program," Letter to W. E. Beall, Boeing, Dec. 18, 1950, Boeing archives. See also Kipp, History of B-47 Procurement, p. 20. 80. Gen. Thomas O. Hardin, Dep. Inspector General, USAF, “Special Survey of the B-47 Program," Sept. 9 to Oct. 6, 1951, Files, HO/ASD. 79.

J.

E. Schaefer,

(102]

Building a

Jet

Bomber

The deputy inspector general went on to note that the Air Force put Boeing workers on a six and one-half day work week to accelerate airframe production even though Boeing protested that the shortage of assembly parts would lead to costly out-of-sequence operations and overtime pay for work on airplanes that could not be delivered to SAC because they were not operational. Another Air Force review described how concurrency ultimately delayed the B-47 program: The decision to proceed with production before the required quality and performance had been attained undoubtedly delayed both the development and the production phases of the program. All estimates down to 1953 were overoptimistic. Total slippage in the development program over original estimates was four years. Instead of terminating in August 1955 as planned, the production phase lasted until February 1957. The overall production program from 1951 on was undoubtedly delayed by the necessity to conduct extensive modification of new planes and retrofitting service planes in order to have a qualitatively acceptable combat plane. Because of development and production delays, equipping SAC medium bomber wings with B-47S lagged two years behind original estimates 81 .

.

.

.

The B-47 production program

had a less than optimal outcome. It did not provide aircraft that could perform effectively as operational weapon systems. It did not proceed in a timely or efficient manner. It featured a vast number of schedule slippages and expensive aircraft modifications. Although the Air Force had hoped to build the B-47B for $1.2 million per copy, the average unit cost of this production model rose to $2.3 million by 1953. 82 Ultimately, 1,923 B-47S of various makes, models, and configurations were built. Production ended in 1957, and the last B-47 retired from service in 1967. clearly

Conclusions

One

of the

most

striking features of the jet

bomber program

is

the

by the Air Force and its predecessors in structuring the entire development and production effort. The AAC and AAF were responsible for initiating the jet bomber development program in the early 1940s. Time after time, these air force organizations took the lead in approaching engine and aircraft manufac-

dominant

81.

role played

Quoted

in

"Analysis of Development, Procurement, and Production of Selected

Systems." 82. Including the cost of government-furnished equipment; see Bagwell, History B-47 Production Program, p. 71.

of the

Flying Blind

turers about the project. in this process.

It is

The contractors played moreover, that the

clear,

a passive, reactive role

AAC

and

AAF

were not

motivated by the economic interests of the contractors. Their overriding concerns were strategic; they were anxious to anticipate emerging threats to It

is

bomber

operations.

also clear that the

AAC

and

AAF

were not simply reacting

to

emerging technological opportunities when they took these steps. To the contrary, they consistently called for the development of systems and capabilities that were far beyond the technological state of the art. The AAF's performance requirements for the jet bomber were set early in the development process, when little basic aerodynamic research had been done on the compressibility effect. At that stage, the program was confounded by monumental technological unknowns, not blessed by new and exciting opportunities. Moreover, the AAF compounded the problem by demanding impressive speed, range, altitude, and payload capabilities

nologically

same

The Air Force later ambitious defensive armament system as all

at the

time.

Doctrinal considerations played a

critical role in

called for a techwell.

requirements deliber-

was not a coincidence that speed, altitude, range, and defensive armament capabilities were the program's top priorities. But doctrinal imperatives also made it difficult for decision makers to face up to inherent performance trade-offs or to compromise along any of these dimensions. So, although AAC and AAF leaders were spurred into action by strategic developments, the content of their decisions was strongations.

ly

It

influenced by doctrinal predispositions.

The compressibility

effect

was

ultimately neutralized,

and the AAF's

speed requirements ultimately met, but only because of an unforeseen technological breakthrough: swept-wing aircraft designs. Technological progress, therefore, ultimately enabled Boeing to meet the program's performance requirements. The key point, though, is that the AAF demanded developments such as these, and the critical technological breakthroughs that did take place came long after the program's performance requirements were set. The process was driven by AAF demand, not technological supply. Because its development objectives were extremely ambitious, the jet bomber program was confronted not just by major engineering problems but by fundamental scientific questions. The attainment of these development objectives ultimately required major technological breakthroughs in aerodynamics as well as propulsion

and

electronics. In short, the B-47 required

system

level to

quently belongs

meet

its

new

technologies at the performance requirements. The B-47 conse-

second category on the nine-point scale of technological ambitiousness outlined in Chapter 1 (see Table 1). in the

Building a

The

jet

Jet

Bomber

bomber program was

a

model

development in and many designs were

of sequential

the 1940s. Aircraft designs were highly flexible,

considered before prototypes were built and flown; this was especially true of the B-47. Development schedules were flexible and, again, this was especially true of the B-47. Bs program featured extensive hardware testing and a prototype fly-off. Competition was extended into flight testing, which gave the Air Force several production options even in the late 1940s. This, in turn, created important incentives for the contractors and useful levers of influence for the Air Force. The designs of the prototypes themselves were flexible, so flight-testing results led to design changes as needed; the B-47 program took advantage of this in 1948 and 1949. The prototypes, moreover, were no-frills development models that did not carry a full array of integrated military subsystems. They were true development prototypes; that is, they were not built on production tooling. Because major financial commitments were not made prior to production, sunk costs were at a minimum when the time came to cancel some of the projects. The critical production decisions followed the flight testing program; the first steps toward a B-47 production program were taken only after its flight testing program concluded. In short, the B-47

program

of the 1940s exhibited

of concurrency outlined in

Chapter

1

none

(see Table

of the characteristics

2).

approach, the development phase of the jet bomber program was quite successful, even though it was extremely ambitious technologically. Major technological advances were made in remarkably little time. The four prototypes were engaged in their flight test programs only three years after the AAF issued its performance requirements; the B-47 began its flight test program only two years after it was completely redesigned. The B-47 ultimately surpassed the pro-

Because of

this sequential

performance requirements and the AAF's most optimistic expectations. In addition, the four prototype programs were inexpensive and did not incur significant cost overruns. They were affordable even in the midst of the severe budgetary constraints of the late 1940s. This picture changed in 1950, when the B-47 program was suddenly reoriented toward concurrency. Five major changes were made in the program. First, production was begun even though the aircraft and its critical military subsystems were still in development; development work began to overlap with production activities. Second, the system's design was assumed to be frozen and producible when it was still evolving. Third, early integration of subsystems was emphasized even though the subsystems themselves were not yet fully developed. Fourth, massive funding commitments were made to the program while this development work was being completed. And fifth, the program's

gram's

critical

Flying Blind

production schedule was planned in elaborate detail at this point. Therefore, five of the features of concurrency outlined in Chapter 1 (see Table 2) were injected into the program in the 1950s. Given that the program was technologically ambitious from the beginning and given that some technological questions still had to be resolved

not surprising that the B-47 production program experienced a wide range of problems in the 1950s. Extensive retrofitting and modification were caused by the decision to push an orderly developin 1950,

it

is

ment program prematurely into high-rate production. The acceleration of the program in the 1950s was a big step in the direction of concurrenproved to be a mistake. It proved that even the most successful program can be derailed by the introduction of an inappropriate procurement strategy. cy. It

premium on system performance than it did on availability and cost. It was willing to stretch schedules and pay for extensive retrofitting, but it was adamant about having its performance requirements met sooner or later. Given a choice on the mix of program outcomes, the Air Force was not willing to trade away performance for time or money. It is

interesting that, in the end, the Air Force put a higher

Whereas one of the main accomplishments of the B-47 development program in the 1940s was its careful resolution of major technological uncertainties, one of the most striking features of the B-47 production program in the 1950s was its arbitrary disregard of these kinds of uncertainties. In 1950, the design of the aircraft itself had not yet been fully refined and critical military subsystems were still under development. As the Air Force deputy inspector general correctly pointed out, it was literally taken for granted that the design of the aircraft and the development of its subsystems would work out. The program's complex production schedule and predictions about the availability of the B-47 as a weapon system were also taken for granted. All these uncertainties were literally assumed away. Although assuming technological uncertainties away in this fashion eliminate them from the program, it did objectively did not actually eliminate them from the Air Force's subjective operating environment.

—

—

Consequently, the Air Force could make elaborate plans for testing, production, training, maintenance, basing, deployment, and operaall important activities from the Air Force's perspective. tions Concurrency also served the Air Force's bureaucratic interests. It, along with the other services, had suffered through the budgetary famine of the early postwar years, and it was anxious to make an irreversible commitment to B-47 production. The leadership in the Air Force was well aware that U.S. defense budgets had historically gone through a "feast-or-famine" cycle; budgets went up during wartime and they dried

—

[106]

Building a

Jet

Bomber

up during peacetime. Since it was not at all clear how long the budgetary feast of the early 1950s would last, the Air Force moved quickly to ensure that it would emerge with large numbers of B-47S in the force structure regardless of what happened on the budgetary front later in the decade. From a bureaucratic perspective, it did not matter if large numbers of unfinished aircraft were coming off the production lines, as long as large numbers of aircraft were materializing in one form or another. The Korean War provided the Air Force with the budgetary resources it needed for a massive production program, while the deteriorating world situation provided it with a strategic rationale for doing what it was inclined to do for bureaucratic reasons anyway. As it turned out, defense budgets did not dry up with the end of the Korean War in 1953, although they were trimmed by the Eisenhower administration.

Nor

did the strategic rationale for major

weapon

acquisi-

programs disappear as the 1950s unfolded. In fact, the strategic arms competition with the Soviet Union heated up. These supply-side and demand-side considerations enabled the Air Force to institutionalize concurrency as the dominant procurement strategy in the 1950s. The Air Force introduced what it called the weapon system concept, which assumed that military subsystems needed to be integrated into the design effort at the earliest possible moment. The weapon system concept was said to be the solution to the problems experienced in the B-47 production program. The deputy inspector general's warnings about how concurrency backfired in the B-47 program were ultimately ignored; the analytic case for the sequential approach was overwhelmed by concurrency's bureaucratic and organizational appeal. As a result, the Air Force ended up with the worst of both worlds: it failed to duplicate the successes of the B-47's development program, and it institution

tionalized the failures of

The Air Force's Chapters 5 and 9. 83.

its

production program. 83

institutionalization of concurrency

is

analyzed

in

more

detail in

[

The

4]

First Intercontinental Bombers:

The B-35, B-36, 6-49, B-52,

and B-60

made

concerted effort to build a bomber with intercontinental capabilities in the 1940s and early 1950s, an effort that included the 6-35, 8-36, 8-49, 8-52, and B-60 programs. The competition between these programs was complicated because they began at different times over a ten-year period. U.S. air force leaders

The development

a

began

when

AAC

became concerned that Britain was about to fall to Germany. If this happened and if the United States entered the war, the AAC would not have any forward bases for its relatively short-legged bombers. It would need, literally, an intercontinental bomber. The propeller-powered B-35 and B-36 were built with this requirement in mind. The second phase of the development effort began in 1944, when intelligence reports about highspeed German

effort

in early 1941,

AAF

the

requirements for a much faster long-range bomber. Boeing's B-52, geared toward meeting this second set of requirements, eventually became involved in an informal competition with the B-36, as the B-35 fell by the wayside. This competition pushed the contractors to maximize aircraft performance, and it ultimately led Boeing to completely redesign the B-52, switching from turboprop to turbojet engines and from straight to swept wings in the process. The redesigned B-52 subsequently won out over the B-36. A jet fighters

led the

turbojet version of the B-35, the B-49,

to issue

was thrown

into the competition

budget crunch of the late 1940s. An all-jet version of the B-36, the B-60 was proposed in 1950, but the B-52's superior performance carried the day in this competition as well. as well, but

it

failed to survive the

[108]

The

First Intercontinental

Bombers

impetus for these programs came from the strategic arena, the AAC and AAF were responsible for determining performance requirements. Again and again, these requirements were set far beyond the state of the art, and contractors were hard pressed to meet the service's demands. As a result, these programs were all technologically ambitious. They were, however, guided by different and shifting procurement strategies. The B-35 and B-36 started out as sequential programs, but wartime pressures led to production contracts in 1942 and 1943 even though prototypes of the aircraft had not yet flown and, indeed, would not fly until after the war. The B-49 program involved some concurrency because of the heavy investment the AAC and AAF had already made in the B-35. The B-52 program was complicated because, like the B-47, it was highly sequential in the 1940s and accelerated after the Korean War began. Later, it was accelerated again in response to concerns that the Soviet Union was racing ahead of the United States and creating a potentially dangerous "bomber gap." The B-60 is a

Although the

original

truncated case;

it

was canceled

after losing a fly-off

with the B-52.

analyzing the outcomes of these programs is gauging the relative impact that sequential and concurrent strategies had on these technologically ambitious ventures. As it turned out, the major technological problems that confounded the B-35 anc b _ 49 programs were

The key

to

*

badly

compounded by

a fairly

mark

program in the late 1940s and early 1950s, a hallconcurrency. The B-52 development program was highly suc-

major

of

concurrency. The B-36 ultimately had to undergo

cessful in

retrofit

many

respects in the late 1940s but, not surprisingly,

it

was

adversely affected by the introduction of concurrency in the early 1950s.

Origins of the B-35 AND b-36 Programs

The AAC had a strong and continuing interest in building long-range bombers in the 1930s an interest based on strategic beliefs, doctrinal inclinations, and bureaucratic motivations (see Chapter 2). Given this outlook, it is not surprising that the AAC built several long range bombers in the 1940s. But AAC predispositions do not provide a sufficient explanation of why the B-35 and B-36 programs were initiated in 1941 and why intercontinental range requirements were established for

—

them.

The

catalyst

was

the changing strategic situation.

As

the

war in Eudependent

rope progressed in late 1940, the implications of being totally on overseas bases became clearer in the United States. If Germany succeeded in defeating Britain, the AAC might need a long-range bomber

Flying Blind

capable of reaching the continent from American bases. The most capable American bomber in service, the B-17C, had a maximum range of only 2,400 miles, which meant that it had a useful radius of action of 1

only 900 miles or so, once allowances were made for taxiing, taking off, assembling bomber formations, and a round-trip mission. 2 The most advanced bombers under development the B-15, B-19, and B-29 also

—

—

fell far

A new bomber was needed. was spurred the AAC to do what

short of intercontinental capabilities.

So, the changing strategic situation

it

do anyway, and it provided compelling strategic rationale for building a long-range bomber which effectively neutralized bureaucratic opposition to such programs. Doctrinal predispositions and parochial bureaucratic interests were galvanized by legitimate strategic needs. In April 1941, the AAC's Materiel Division asked Consolidated and Boeing to make design studies based on a preliminary set of military requirements. At a minimum, the new bomber was to fly an 8,000-mile mission with a 4,000-pound payload at a top speed of 350 mph and an altitude of 35,000 feet. If possible, the bomber was to fly a 12,000-mile the AAC mission at 450 mph and 45,000 feet. The Materiel Division division responsible for weapon development and the in-house source was unsure about whether of technical expertise on these matters inclined to

—

—

consequently asked the contractors to explore "the aeronautical art and to determine what may be technically possible ... in the field of large airplanes within the next three or four years." 3 The AAC's plan was to delay finalizing the performance requirements until the results of these preliminary investigations came back from Consolidated and Boeing. The Materiel Division strongly supported this approach, since it would allow the AAC to identify the program's major technical challenges and determine how much time it would need to overcome them before setting the parameters for actual hardware development. 4 Both Consolidated and Boeing agreed to conduct design studies. By the end of May, two other companies, Northrop and Douglas, were also involved in the program. All these studies, it must be emphasized, were highly exploratory. The contractors were simply trying to determine the these requirements could be met.

1.

2.

It

HO/AMC, Case History of B-36 Airplane, May 1949, p. 1. A bomber's effective radius of action was about three-eighths of its total

range. This

standard changed over time, however, as longer-range aircraft were developed; a longrange bomber spent proportionately less of its total mission time taxiing and taking off. For details on the B-17, see Gordon Swanborough and Peter M. Bowers, United States Military Aircraft since 1908 (London: Putnam, 1963), pp. 87-96. 3. Lt. Col. A. E. Jones, Chief, Contrast Sect., Mat. Div., Letter to Consolidated Aire. Corp., April 11, 1941, XB-36 Airplane, RG 18, AAF, NARS. 4. Lt. Col. F. O. Carroll, Chief, Exper. Eng. Sect., Mat. Div., Memorandum to Chief, Eng. Sect., Mat. Div., Office of the Chief of the Air Corps, April 11, 1941, XB-36 Airplane,

RG

18,

AAF, NARS.

[no]

The

First Intercontinental

Bombers

general configuration of a long-range bomber and, given some preliminary wind tunnel work on various models, the rough performance capabilities of various designs. Consolidated estimated that six months of study would be needed before fully documented proposals could be

developed. 5 Only then could the technical feasibility of the project be meaningfully assessed. Events in Europe accelerated this timetable, however. Germany invaded the Soviet Union and put together a string of military successes in the summer of 1941. It was not inconceivable that Hitler was on the verge of conquering all of Europe. These events added to the sense of urgency already surrounding the long-range bomber project. A high-level meeting on the project was held at the Washington headquarters of the recently reorganized AAF on August 19. The chief of the Materiel Division repeatedly attested to the magnitude of the technological problems the project faced, and he estimated that it would take at least

two and one-half years

just to build a preliminary prototype.

The

Division argued that the deteriorating international situation nonetheless made the development of a long-range bomber essenfor such tial. As far as it was concerned, the one mandatory requirement

Air

a

War Plans

bomber was an

overall range of 10,000 miles,

which would provide

a

6 4,000-mile radius of operation. This argument carried the day. Although the Materiel Division had serious reservations about meeting the April requirements in the near term, several of the bomber's

performance requirements were raised in August. The minimum range requirement was raised to 10,000 miles, as the Air War Plans Division insisted, and the payload requirement was raised to 10.000 pounds. In addition, the altitude requirement was increased to 40.000 feet, and the defensive armament requirements were specified in able to detail. According to the AAF, the long-range bomber had to be

most

critical

defend

would

against three fighters attacking at the same time, which enable it to perform daylight missions; at a minimum, it was to be itself

equipped with six 37-mm cannon and eight 50-caliber machine guns. Although demanding defensive armament requirements were consistent with the strong emphasis strategic bombardment doctrine placed on defensive firepower, they added to the weight of the aircraft and 7 made it more difficult for range and payload requirements to be met. surrounding In fact, the magnitude of the technological unknowns

5.

Consolidated Aire. Corp., Letter to Asst. Chief, Mat. Div.,

AFLC. 6.

Mai.

B.

3,

1941, Files,

HO/

Mat. Div., Office of the Chief of the Air Air Corps, Aug. 19, to Chief, Mat. Div., Office of the Chief of the

W. Chidlaw, Chief, Exper. Eng.

Corps, Memorandum 1941, XB-36 Airplane,

May

RG

18,

AAF, NARS;

Br.,

cited hereafter as

Chidlaw Memorandum.

$

Flying Blind

the program were not appreciated even by the Materiel Division, because the final performance requirements were set before Boeing and

Consolidated completed their exploratory investigations. Boeing did not get formal approval for its studies until after the August 19 requirements meeting took place. Consolidated was able to provide only some partial preliminary data in advance of this meeting. 8 The August requirements decisions, therefore,

say the

were not based on complete and systematic analy-

Moreover, the available technical information was not entirely encouraging, as the Materiel Division recognized. The AAF leadership nonetheless wanted to begin constructing pro totypes as soon as possible. The long-range bomber program was seen ses, to

least.

as a top national priority,

and the Materiel Division was

told that design

and development "should be expedited to the greatest possible degree." The Air War Plans Division was thinking in even more grandiose terms: its assessment was that, in order to "realize the quantities laid down by a recent staff study in this Division, it would be necessary to proceed with an immense production plan immediately." 9 Pressures were already building for a large-scale production program even though exploratory design studies had not yet been completed and hardware development had not even begun. The four contractors were pressed to submit proposals at once, and all proposals were in the AAF hands by mid-September, just one month after the performance requirements were radically changed. The Materiel Division acknowledged that the wide variation in the designs was attributable

to

the inadequate

amount

of time

allowed for careful study. It concluded that the most promising design was Northrop's exotic flying wing, an aircraft that had no conventional fuselage or tail and was, literally, all wing. Although Northrop's design was projected to have a range of only 8,000 miles, the flying wing was seen as having great long-term potential. 11 The most promising of the conventional 10

O. Carroll, Chief, Exper. Eng. Sect., Mat. Div., Teletype to Consolidated Aug. 7, 1941, and H. A. Sutton, Chief Engineer, Consolidated Aire. Corp., Letter to Asst. Chief, Mat. Div., Aug. 16, 1941; both in XB-36 Airplane, RG 18, AAF, NARS. Also, Contract with Boeing Aire. Co., Aug. 23, 1941, Files, HO/AFLC. 8.

Lt.

Col.

F.

Aire. Corp.,

9.

Chidlaw Memorandum.

10. Carl Arnold, Exper. Eng. Sect., Mat. Div., "Comments on Preliminary Design Studies: Long-Range (Heavy) Bombardment Airplanes," Memorandum Report, Oct. 3, 194 1 XB-36 Airplanes, RG 18, AAF, NARS; cited hereafter as Arnold Report. Also, Col. H. Z. Bogert, Chief, Tech. Staff, Memorandum to Col. F. O. Carroll, April 29, 1942, Files HO/ r ,

AFLC.

11. Lt. Col. F. O. Carroll, Chief, Exper. Eng. Sect., Letter to Chief, Mat. Div., Office of the Chief of the Air Corps, Sept. 13, 1941, cited hereafter as Carroll B-35 Letter, and Lt. Col. A. E. Jones, Chief, Contract Sect., Mat. Div., Letter to Northrop Aire., Inc., May

1941; both in Files,

Mat. Div.,

May

27,

HO/AFLC.

Also, John K. Northrop, Letter to Col. Howard Z. Bogert, XB-35 21, 1941, Airplane, RG 18, AAF, NARS; cited hereafter as Northrop

Letter.

[

112 ]

The

First Intercontinental

Bombers

designs was Consolidated's six-engine bomber with the engines in a pusher configuration. The Boeing and Douglas designs were rejected on 12 the grounds that they were overly conservative Clearly, the AAF made important decisions in an informational vacuum. Its evaluation of the proposals was based entirely on tentative paper calculations and some preliminary wind tunnel testing. None of the proposed aircraft had ever flown and, moreover, no aircraft of this general size and weight had ever been built. Northrop's flying wing was simply experimental, not that much more was known about the conven.

The contractors, moreover, did not provide all the technical information needed for a thorough assessment of their proposals. The Materiel Division reported that "in no case was sufficient aerodynamic data submitted to permit a complete check" of predicted aircraft performance, "only one of the design studies contained data giving the amount of fuel carried, and only one study was accompanied by a 13 How could one estimate range cageneral arrangement drawing ." pabilities without knowing how much fuel the aircraft carried? Range tional designs.

was, after all, the critical performance issue in this program. Although it was impossible to make a meaningful assessment of the technical or operational capabilities of these designs under these conditions, the Materiel Division nonetheless conducted its review and concluded that Consolidated's design could meet the AAF's performance requirements. Douglas believed, to the contrary, that the performance requirements for 14

program were unattainable, given the state of the art The Materiel Division's final benediction was that Northrop's and Consolidated's designs were "within the practical bounds of the current development of the aeronautical art ." 13 Only six months earlier, this division had been forced to ask its contractors to explore the "aerothis

.

nautical art" because

it

knew

so

little

about the technological pos-

of building an intercontinental bomber had not been seriously considered prior to that time. Yet by October, the Matewing was riel Division was claiming that even Northrop's exotic flying sibilities.

The problem

within the state of the art. It is not at all clear how the Materiel Division came so far, so quickly, on so little hard data. The reservations it had expressed in August gave way to a "can do" endorsement in October even though virtually nothing had been learned in the interim. The Materiel Division's recommendation to proceed with prototype construction

was acted on

12.

Arnold Report.

13.

Ibid.

quickly. Informal approval

Maj. Gen. L. C. Craigie, Chief, Eng. Div., XB-52 Airplane, RG 18, AAF, NARS. 15. Arnold Report. 14.

AMC,

Letter to

came from Wash-

CG, AAF,

July 11, 1947,

Flying Blind

ington within two weeks, and formal contracts were awarded by the end of the year. Northrop committed itself to build two B-35 bombers for $4.45 million. The first of these aircraft was to be delivered by November *943 Consolidated agreed to build two B-36 bombers for $15.8 million, with the first of its prototypes to be delivered by May 1944. 16 Several points should be emphasized. First, although the long-range •

bomber program was launched in response to strategic developments, the AAC was the driving force behind it. The idea of building an intercontinental bomber certainly was not an industrial initiative. Contractors

were not even contacted by the

AAC

about the project until after

preliminary performance requirements were established for it in April 1941. All evidence suggests that the AAC approached these contractors,

and not the other way around. Moreover, the AAC was not reacting to a technological breakthrough, nor was it anticipating such a breakthrough when it launched its effort. Its assessment of existing development projects suggested that major technological advances were needed if intercontinental capabilities were to be developed. And the reports from the contractors were not encouraging. Even Northrop, the most technologically aggressive and financially desperate of the lot, did not claim to be able to meet the AAC's performance requirements. 17 Douglas was highly skeptical about the project. Finally, the Materiel Division expressed strong reservations about the technological feasibility of the project throughout most of the year. Air force leaders, therefore, were not being technologically oppor-

when

they launched the long-range bomber program in 1941. To the contrary, they set the major performance requirements for their new tunistic

bomber far beyond the state of the art. The 10,000-mile range requirement was approved in spite of serious doubts about its feasibility. Air force

compounded the program's technological challenges by setting demanding requirements in several areas at once, even though many of planners

these capabilities existed in a trade-off relationship. A comparison of the August 1941 requirements and the capabilities of the most advanced bombers then in service or even under development shows that the former were several times more demanding in several critical respects: 18 16. "Contract," Northrop Aire., Inc., Nov. 22, 1941; "Change Order," Dec. 12, 1941, XB-35 Airplane, RG 18, AAF, NARS. "Contract," Consolidated Aire. Corp., Nov. is uui XB-36 Airplane, RG 18, AAF, NARS. 17. Northrop Letter. 18. I he B-17C was in service. The capabilities of the B-15 and B-19 with operationally useful payloads were limited, mainly because they were slow. The B-29 was in development. In the long run, it did not meet all its requirements, although the AAF might have assumed in 1941 that its requirements would be met. The B-29S that went into service later in the war had a range of about 3,300 miles and a top speed of 358 mph. See Swanborough and Bowers, United States Military Aircraft, pp. 87-103, 556.

540,

[

114

]

The

First Intercontinental

Bombers

B-29

August 1941 requirements

Top speed (mph) Range (miles) Payload (pounds)

B-15

B-17C

350

145

291

10,000 10,000

3,500 2,000

2,400

4,000

B-19

224 5,200 2,500

requirements 400 5/333 2,000

designed to meet these requirements were, inevitably, technologically adventurous. This was obviously true in the case of the B-35, which was, as one AAF observer accurately put it, "revolutionary in design." 19 Little experimental work had been done on the aerodynamic characteristics of flying wings prior to 1941, when Northrop

The

aircraft

began to work on a bomber-sized aircraft. And, needless to say, neither Northrop nor anyone else had any experience whatsoever in building

numbers of such aircraft. The B-36 program was deceptively ambitious. If the flying wing had been out of the picture, the B-36 would have been seen as the highly ambitious venture that it was. To begin with, it was much larger and heavier than any aircraft the AAF had ever attempted to build. The gross weight of the B-36 was almost twice that of the experimental B-19 and 20 This size represubstantially more than that of the B-35, for example. sented a major engineering challenge in itself, but it also meant that new, advanced subsystems would have to be developed. The hydraulic, electrical, and flight control systems, for example, all involved major

large

technological advances. 21 Moreover, to fly a 10,000-mile mission the B-36 needed low aerodynamic drag, which was the main reason for installing the engines in a pusher configuration behind the wing, keep-

ing the leading edge of the wing clean. There were, however, aerodynamic unknowns associated with this wing-engine configuration. In addition, new manufacturing processes would be needed to produce the B-36's smooth wing and fuselage surfaces. In all, building the B-36

was

a formidable technological challenge.

Acceleration of the B-35 Program ‘ the contracts for the B-35 and B 3 6 prototypes were signed in stratlate 1941, the AAF planned to employ sequential procurement

When

Aircraft," Col. B. S. Kelsey, Chief, Aire. Proj. Br., "Information on Bombardment Memorandum, May 29, 1945, Experimental Bombers, RG 34 L HQ/USAF, NARS. pounds, al20. The gross weight of the B-36 was projected to be 265,000—273/000 have to a gross projected though it ultimately grew to over 400,000 pounds. The B-35 was 19.

weight of 162,000 pounds, and the B-i9 had a gross weight of 160,000 pounds. American Aviation 21. Meyers Jacobsen, "Design Development of the XB-36," Society Journal 15 (Winter 1970), 224-235.

Historical

Flying Blind

The Materiel Division estimated in 1941 that it would take from two to two and one-half years simply to build the first prototypes. It strongly recommended that limited production be initiated only after a twelve- to eighteen-month prototype testing program had been completed. 22 This meant that production would not begin until 1944 or 1945, but the technological uncertainties that permeated both programs would be resolved prior to production. The sequential plan would also enable the contractors to design the new tooling, manufacturing, and construction processes needed for high-rate production. The prototype programs themselves, moreover, were inexpensive. Given the technological uncertainties in the two programs, this was a reasonable way to proegies.

ceed.

Pressure for an immediate production program was, however, already building in 1941. It was certainly strong in the summer of 1941,

when

the military situation in Europe seemed to be deteriorating, but it actually intensified in 1942 and 1943 even though the situation in Eu-

rope had begun stalled, fore,

and

it

to stabilize.

was becoming

Germany's invasion of the Soviet Union less likely that Britain

would

fall.

There-

bases near the continent would probably be available for American

bombers. Even collapse.

was still possible that the Allied position could The long-range bomber program served as a hedge against this

catastrophe.

so,

it

hedge against the collapse of the campaign in the Pacific, where the United States had not yet acquired the forward bases it would need to bomb Japan. Finally, it served as a hedge against serious problems in the B-29 and B-32 programs, which were geared toward the Pacific campaign. It was certainly prudent to continue the B-35 and B-36 development programs as insurance against these kinds of contingencies, but there was no compelling strategic need to accelerate these programs into production in 1942 or 1943. The AAF, however, was anxious for bureaucratic reasons to make an irreversible commitment to a long-range bomber. Its case for organizational independence from the Army would be strengthened significantly if it could deploy a bomber that was dependent on neither the Army nor

Navy

It

also served as a

forward bases. In fact, if the AAF could field a bomber with intercontinental range, it could argue that large ground and naval forces were unnecessary in peacetime. The AAC's efforts to build long-range bombers of this kind had been stymied repeatedly in the 1930s, and the AAF was quick to push ahead with the B-35 and B-36 in the early 1940s when the opportunity to do so presented itself. The AAF consequently took several steps in 1942 and 1943 to move, first, the B-35 and, then, the B-36 into production, even though enthe

for

[116]

The

First Intercontinental

Bombers

gineering development and prototype construction were far from complete in both cases. The AAF pushed the B-35 into production first 22 because it seemed to have more performance potential than the B-36,

but the B-35 seemed especially promising precisely because it was especially adventurous technologically. It was, in fact, even less suitable for concurrent development and production than the B-36. If there really

was an urgent strategic need for a long-range bomber, it would have made more sense to push the B-36 into production. The fact that the kept the B-36 in development suggests that it was responding more to its internal doctrinal preferences for high-performance aircraft than to external strategic circumstances. The AAF began to lay the groundwork for the B-35 production pro-

AAF accelerated

the B-35 while

it

September 1942, when it began to expand Northrop s limited production facilities. 24 Even so, Northrop, which had been founded only in August, 1939, did not have the engineering or industrial resources for large-scale production. The AAF anticipated this problem

gram

in

and integrated the Martin aircraft company into its production plan. It canceled Martin's B-33 program in November 1942 in order to free up 25 With this engineering personnel and production facilities for the B-35 production team in place, the AAF forged ahead in December and signed contracts with Northrop and Martin for thirteen B-35 prototypes and 200 production models. Although Northrop was the prime contractor,

it

was responsible

for building just the first prototype, in addition to

designing the aircraft. Martin was to provide critical developmental en26 gineering assistance and handle the bulk of the production effort. Meanwhile, Northrop was having serious problems perfecting the wings, the flying wing. It had been counting on one of its smaller flying provide hard aerodynamic data based on actual flight testing. The N-9-M was especially critical once the B-35 program was accelerated because no B-35 prototype would fly prior to production. But the N-9-M N-9-M crashed flight test program was plagued with problems. The first

N-9-M,

to

Craven and James Alfred Goldberg, "The Quest for Better Weapons," in Wesley F. vols. (Chicago: University of Chicago L. Cate, eds., The Army Air Forces in World War II, 7 Press, 1948-1955). vol. 6, p. 245. Dep. Chief of Air Staff, Sept. 30, 1942, 24. Maj. Gen. O. P. Echols, Memorandum to XB-35 Airplane, RG 18, AAF, NARS. „ "Classified Technical Instructions to Technical Executive, W. Sessums, 23.

25.

Col.

Jr.,

J.

Wright Field, Nov. 25, 1942, XB-35 Airplane, RG 18, AAF, NARS. would cost $22.8 million and that 26. It was estimated that the first batch of 13 B-35S million; see Brig. Gen. B. E. the production run of 200 B-35S plus 40 spares would cost $263 to Technical ExecuMeyers, Chief of Staff, Mat. Comm., "Classified Technical Instructions Dec. 17, 1942; Col. J. W. Sessums Jr., tive," Wright Field, Dec. 11, 1942; "Contract," Wright Field, Jan. 8, 1943; Glenn "Classified Technical Instructions to Technical Executive," Field, May 6 1943; all in XB-35 Airplane, RG 18, L. Martin Co., Letter to CG, AMC, Wright ,

AAF, NARS. [

117

]

Flying Blind

May

in

1943 an d most of

mechanical

failures. 2'

45 flights had been terminated because of AAF was taken aback by this chronic mechani-

The

its

he second N-9-M made its first flight in June 1943, a nd the first reliable drag data on the flying wing were obtained the following September. The news was not encouraging: the B-35's drag would be 7-12 percent greater than wind tunnel estimates had indicated; Northcal difficulty.

I

rop's early predictions about the B-35's drag

and performance were, as the AAF came to recognize, questionable. As N-9-M flight testing continued, it became clear that the flying wing also had serious directional and longitudinal stability control problems, dangerous spin characteristics, control problems at low speeds, stall problems, and problems with its elevator and rudder control surfaces. 28 These were, of course, fundamentally important design issues.

The

program in 1942, therefore, led to the initiation of production roughly one year before basic, critical information about the flying wing became available. Given that the flying wing was truly experimental, it is not surprising that the N-9-M flight test program identified a variety of major design problems. As it turned out, the concurrent production program interfered with the timely resolution acceleration of the B-35

of these problems.

Northrop would have struggled

under the best of circumstances, simply because it did not have vast numbers of engineers to draw on. It had only 30-35 engineers working on the project in late 1942. The AAF consequently asked Martin to provide some of the engineering resources that Northrop needed. 29 Martin's main interest in the project, though, was its production contract. Martin needed Northrop's blueprints to begin its production run, but it had no vested interest in providing extraordinary assistance to one of its competitors. The AAF recognized in mid-1943 that the program was in trouble, and it conducted an investigation into the problem. It discovered that Martin's production activities were interfering with Northrop's development work. Instead of helping Northrop draw up the preliminary blueprints needed to build the prototypes, Martin was focusing its effort on the detailed drawings it would need for its production run and on various to perfect the B-35

production tooling problems. According to the Materiel 27.

Gen.

Airplane, 28.

ect,"

F.

RG

Ibid.

O. Carroll, Chief, Eng. Div.,

18,

AAF, NARS;

See also Tech.

Memorandum

cited hereafter as Carroll

Staff,

Eng. Div., Mat.

Comm.,

to

Command,

CG, AAF,

jan.

1,

the

1944, XB-35

Memorandum. "Plan for Expediting XB-35 Proj-

Memorandum

Report, June 11, 1943, cited hereafter as Technical Staff Report, and Brig. Gen. F. O. Carroll, Chief, Eng. Div., Memorandum to Gen. Echols, Jan. 10, 1944, cited hereafter as Carroll Status Report; both in XB-35 Airplane, RG 18, AAF, NARS.

Northrop had fewer than one hundred engineers working on the project in 1944; see Carroll Status Report. See also Brig. Gen. Melvin E. Gross, Chief, Reqs. Div., Memo29.

randum

for the

Chief of the Air Staff, March 22, 1944, XB-35 Airplane, Gross Memorandum.

cited hereafter as

RG

18,

AAF, NARS;

The

First Intercontinental

Bombers

Northrop B-35 (USAF Photographic Collection, National Air and Space Museum, Smithsonian Institution)

main reason the program was behind schedule was "the pressure placed on Northrop by Martin to furnish production engineering data." Production planning was emphasized so much that "development engineering was definitely pushed into the background and hindered to a most dangerous degree." As a result, "the B-35 has not been developed with asto a point where drawings for production can be released surance." 30 The Materiel

Command, which had

development strategy

for the B-35

originally favored a sequential

program, gave these warnings

in

June

is

not

1943;

The present experimental and production program for sound and will lead to serious trouble, if not corrected. .

.

the B-35 .

W. Chidlaw, Chief, Mat. Div., Memorandum to Brig. Gen. E. S. NARS; cited hereafter as Chidlaw Perrin, Dec. 28, 1943, XB-35 Airplane, RG 18, AAF, Division became the Materiel Status Report. See also Technical Status Report. The Materiel 30

Gen

Brie

Command

in

B.

March

1942. For

more

details, see the

appendix.

Flying Blind

view of past experience with conventional airplanes, the unconventional and radical design of the 6-35, and the aerodynamic and control problems which have not yet been definitely solved, it is considered illogical and highly unsound to attempt to bring the "X" [prototype] and "B" In

[production] airplanes out within such a short interval of time. In order to expedite the first flight of the "X" airplane and lengthen the time interval between "X" and "B" airplanes, it is necessary for Martin to concentrate at this time on the minimum engineering and drawings that are .

.

.

required to get Northrop started on the "X" airplane and pick up the detail design for production as soon as experimental engineering requirements

have been met. Martin [should] resume production engineering only when mutually agreed by Martin and Northrop that no interference with the experimental engineering will result. 31 .

.

.

In other words, the Materiel least

some

of the

Command

believed that

it

could undo

damage caused by concurrency by emphasizing

at

pro-

totype construction, creating a distinct gap between development and production, and shifting back to a more sequential strategy. It estimated that, by stopping all engineering activities not directly related to the

development of the XB-35 prototypes, approximately two months could be saved in the development phase of the program. Although Martin's production run would consequently be delayed by two months, the program would move more quickly in the long run because fewer design changes would be needed after the bomber went into production. 32 The B-35 program was clearly floundering by late 1943. The first flight of the first prototype had originally been scheduled for November 1943, but the program was many months behind schedule at that point. In fact, the development schedule slipped at an alarming rate throughout *943- In J u ly 3 943 it was estimated that the prototype's first flight would take place in August 1944. The AAF's December 1943 estimate was that the first flight would take place in April 1945; the schedule had slipped eight months in just the last five months of 1943. Given this rate of slippage, the prototype would, literally, never get completed. By May the date for the prototype's first flight had slipped again, to August 1944 3 94 5 c*rid it now seemed to the AAF that the B-35 could not be produced >

until after the war. 33 This

Normandy

invasion had

was

a pretty grim assessment given that the not yet taken place and the duration of the war

Technical Staff Report. 32. Ibid. Also, Chidlaw Status Report. 33. Maj. Gen. B. E. Meyers, Dep. Asst. Chief of Air Staff, Memorandum to Chief of Air Staff, May 6, 1944, XB-35 Airplane, RG 18, AAF, NARS; cited hereafter as Meyers Memorandum. See also Carroll Memorandum; Chidlaw Status Report; Carroll Status Report. 31.

[

120

]

I

The

First Intercontinental

was unknown.

Bombers

meant, moreover, that the strategic rationale for an accelerated production program had evaporated. There were also concerns about the B-35's skyrocketing costs and dubious performance capabilities. According to the 1941 contract, the prototype program was supposed to cost $ 4-45 million. By 1944 Northitself

It

—

program cost including the two prototypes, the four N-9-M flying models, and Martin's development and 34 The original production engineering would be almost $25 million. estimates of the B-35's performance capabilities were also beginning to look optimistic. As Brigadier General Chidlaw of the Materiel Command put it, there were two basic questions about the B-35's performance at this point: "One, where did the range go? Two, where did the speed go?" 35 By 1944 it was estimated that the B-35 would be able to carry a

rop was estimating that the

total

—

10,000-pound payload only 5,200 miles. 36 The AAF consequently canceled the B-35 production contract in May 1944. Work would continue on the two prototypes and the batch of thirteen B-35S ordered in 1942. These aircraft would all be used as developmental prototypes and flight test vehicles. Martin was to continue to provide Northrop with developmental engineering assistance. Despite its problems, the B-35 was still seen as promising in the long run, although a production program was neither warranted nor possible in mid-1944. 37 The attempt to accelerate the B-35 program was doomed for two reasons. First, the B-35 itself was fundamentally experimental. Its aerodynamic puzzles and formidable engineering problems could not be assumed away, and there was little that could be done to resolve them quickly. 38 Second, these development problems were exacerbated by the introduction of concurrency and the imposition of a large-scale production

program

in late 1942.

The production

effort itself

was doomed from But, more impor-

the start, given the immaturity of the B-35's design. tant, it was counterproductive because it interfered with development

be completed before production could begin. It is no coincidence that the B-35's development schedule slipped repeatedly in 1943 when production activities expanded.

activities that

34.

Brig.

Gen.

RG

F.

to

O. Carroll, Chief, Eng. Div., Letter to CG, AAF, Feb.

AAF, NARS. Chidlaw quoted in Mary

Airplane, 35.

had

18,

Heavy Bombardment

Aircraft,

1917-1949,

HO/AMC,

Carroll Status Report.

37.

Meyers, Carroll, and Gross memoranda.

Development and Production of USA Dec. 1950, Study No. 195, p. 94.

Vanik, Project Engineer, Letter to Chief, Bomb. XB-35 Airplane, RG 18, AAF, NARS.

M.

F.

|l2l]

XB-35

R. Self, History of the

36.

38.

17, 1944-

Br.,

Eng. Div., Nov. 27, 1944,

Flying Blind

Acceleration of the B-36 Program

The B-36 program was

also accelerated during

World War

although not until after problems emerged with the B-35. Consolidated recommended in August 1942 that the AAF begin construction of 100 B-36S concurrent with the development of the prototypes, arguing that this would expedite the program by at least two years. Otherwise, production would not begin until 1947. 39 The AAF decided in 1942, though, to put the B-35 into production first. By mid-1943 the B-35 program was experiencing serious problems. At

same

II,

had not improved dramatically over the year before. The stepping-stone campaign to acquire forward bases in the Pacific was just under way, and there was no guarantee when or if it would succeed. Even if it was eventually successful, there was no guarantee that the B-29 and B-32 programs would produce the long-range bombers that would be needed in the Pacific. 40 The AAF was able to argue, therefore, that an intercontinental bomber was still needed, and it placed an order for 100 B-36S in June 1943. Each of these bombers was expected to cost $2.6 million. The program was to have a the

time, the strategic situation

top priority. 41

A more

AAF's interest in committing to needed to do so before the Navy

cynical explanation for the

B-36 production in 1943 is that it acquired a string of forward bases for the B-29 in the Pacific. Once these bases were in American hands, the strategic justification for an intercontinental

bomber would be undermined.

In

any event, the

AAF wanted

be irreversibly committed to building such a bomber before the war ended, the country demobilized, and defense budgets were slashed. It would undoubtedly be extremely difficult, if not impossible, to get a production commitment for a new and expensive bomber after the war. The assistant chief of the air staff. Major General Echols, implied that these kinds of considerations pushed the AAF to accelerate the B-36 to

program when he said, "The ordinary procedure would result in a delay of one year in the production of these airplanes, and it was very possible M. Laddon, Exec. Vice Echols, CG, Mat. Comm., Aug. after as Laddon Letter. 39.

40.

I.

AMC/HO,

Consolidated Aire. Corp., Letter to Maj. Gen. O. P. 1942, XB-36 Airplane, RG 18, AAF, NARS; cited here-

Pres., 1,

Case Historx/ of B -}6 Airplane,

War (Bloomington: Indiana University Press,

p. 3;

Russell

1973), P- 37 2

F.

Weigley, The American

Way

of

-

Verbal authorization for the production program was given in June 1943. A letter of in July 1943, although a formal contract was not drawn up until August 1944. See "Letter Contract," Aug. 7, 1943; "Contract," Consolidated-Vultee Aire. Corp., Aug. 19, 1944; CG, ATSC, Teletype to Asst. Chief of Air Staff, May 23, 1945; all in XB-36 Airplane, RG 18, AAF, NARS. 41.

intent

was sent

The

First Intercontinental

Bombers

Convair B-36 (U.

S.

Air Force)

war would be over before production could commence." 42 The AAF, in other words, did not need a long-range bomber to fight the war; it needed the war to build the bomber. When the B-35 production program began to fail in 1943, the AAF wasted no time accelerating the B-36 that the

into production.

The B-36 program began to show signs of trouble soon after this decision was made. Consolidated's schedule for the first flight of the B-36 prototype began to slip from the original May 1944 target date. By 1944 the AAF's best estimate was that flight testing would begin in June 43 Some of the delay was caused by technical problems that were 1945. glossed over until late 1943. For example, Consolidated had to change the B-36 's airfoil section, which in turn necessitated redesigning the entire wing. This kind of major design change should not have been made in the production phase of a program. There was also a wing flutter problem that could not be resolved until a prototype became 42.

Memorandum

to Brig.

Gen.

B.

W. Chidlaw, Chief, Mat.

Div.,

June

19, 1943,

XB-36

RG 18, AAF, NARS. 43. AAF Representative, Consolidated Aire. Corp., Wire to Eng. Div., Mat. Comm., June 10, 1943; Lt. Col. R. E. Ludick, AAF Rep., Consolidated-Vultee Aire. Corp., Letter to Airplane,

1944; Brig. Gen. F. O. Carroll, Chief, Eng. Div., Memorandum to Comptroller, Mat. Comm., July 9, 1943, cited hereafter as Carroll XB-36 Status Report; all Dir.,

in

ATSC, Nov.

XB-36 Airplane,

6,

RG

18,

AAF, NARS.

Flying Blind

ground flutter tests. 44 Production would be well under way before this problem could be solved. One persistent problem was the B-36's weight, which was directly related to its projected range; if everything else was equal, range decreased as aircraft weight increased. The AAF estimated that, if the B-36 's gross weight exceeded 265,000-275,000 pounds, its range would drop to 7,000-8,000 miles. If this happened, the AAF believed the design should be scrapped. 45 But, while remaining adamant about range, the AAF compounded the weight problem by increasing its defensive armament requirements several times. It added 6,300 pounds of defensive armament to the B-36 in late 1942, and it added a 20,000-pound nose turret in 1944 even though this drove the weight of the aircraft over 285,000 pounds. 46 As these design problems accumulated in 1943 and 1944, the cost of the program more than doubled. The contract for the two B-36 proavailable for

totypes was eventually rewritten to reflect the fact that estimated costs had increased from $15.8 million to $32 million. Consolidated admitted that the concurrent production program it had lobbied for was responsible for

many

of

its

problems. 47

By 1944, then, the B-35's days as a top-priority production program were over. It was simply a development project. The B-36 program had begun to have cost, schedule, and performance problems, and, although its production contract remained in effect, it lost its priority status in the summer of 1944. 48 These events coincided with the AAF's decision to build a long-range bomber more advanced than either the B-35 or the B-36.

Origins of the B-52 Program Several strategic developments led the

mance requirements

for long-range

AAF

bombers

to reassess its perfor-

in

1944.

Heavy

aircraft

Consolidated Aire. Corp., Letter to CG, AMC, Oct. 5, 1943, and Col. P. H. Kemmer. Chief, Aire. Lab., Eng. Div., Memorandum to Chief, Tech. Staff, Aug. 26, 1944; both in XB-36 Airplane, RG 18, AAF, NARS. See also Carroll XB-36 Status Report. 45. AMC/HO, Case Historx/ of XB-36, YB-36, and B-36 Airplanes, May 1948, p. 3, 7. See 44.

also

Laddon

46.

Lt.

Letter.

Col.

M.

F.

Bombardment, Nov. 1944; both

AMC/HO

in

Summerfelt, Office of 23, 1942,

XB-36 Airplane,

Dir. of

Bombardment, Memorandum

to Dir. of

and Bomb. Br., Eng. Div., Memorandum Report, Oct. 25, 18, AAF, NARS. See also Carroll XB-36 Status Report;

RG

Case History of XB-36, YB-36, and B-16 Airplanes, p. 7. 47. Consolidated-Vultee Aire. Corp., Letter to Dir., ATSC, Jan. 25, 1945; Maj. E. A. Levinson, AAF Contract Off., Consolidated-Vultee Aire. Corp., Letter to Dir., ATSC, Feb. 22, 1945; “Change Order," March 14, 1945; all in XB-36 Airplane, RG 18, AAF, NARS. 48. AMC/HO, Case History of B-36 Airplane, p. 4.

The

First Intercontinental

Bombers

Regensburg and Schweinfurt in late 1943 forced the AAF to face the fact that the B-17 needed a long-range fighter escort (see Chapter 2). But even then, the AAF persisted in thinking that a self-defending bomber could be built; it would simply have to be much faster than the B-17. The AAF was also jolted in early 1944 by intelligence reports indicating that German jet aircraft might be capable of speeds of 630 mph. These reports led the AAF to step up its efforts to build a high-speed jet bomber and take a closer look at the speed capabilities of its long-range bombers. Since the B-36 was expected to be only marginally faster than the B-17, an d the B-35 program was in developmental limbo, 49 the AAF formulated requirements for an advanced bomber to succeed the B-35 an d b-36. Preliminary requirements were drawn up in April 1944 and circulated to the major aircraft manufacturers in May. Boeing's assessment was that advanced turboprop engines would have to be built before these requirements could be met. Engines this advanced, however, were not even in development at the time. The AAF's requirements were so demanding that none of the contractors who were approached about the project were interested in submitting losses in the attacks over

proposals. 50

AAF

forged ahead anyway, formally issuing a set of requirements 51 Once again, the for a high-speed, long-range bomber in August 1944. aircraft industry's response was less than enthusiastic: the AAF did not

The

receive a single proposal.

As

the

ments "were so completely out

AAF

later

acknowledged,

its

require-

of line with the state of the art that

no

52 Clearly, the contractor could be interested in attempting a design."

aircraft

industry

was not

the driving force behind this program.

The AAF was nonetheless determined to build a long-range bomber that was substantially faster than the B-35 or b-36/ and within a few months of the war's end it issued another set of requirements for such a bomber. Its November 1945 requirements called for a bomber with a top speed of 450 mph and a 5,000-mile radius of action, roughly equal to a 12,000-mile range. In addition, the new bomber was to carry a 10,000pound payload on the 12,000-mile mission and payloads of up to 80,000 pounds on shorter missions. To fly daylight missions, it was to be 1946 that the B-36 would have a top speed of 325-343 mph. The B-17S then in service had a top speed of 317 mph. See Res. and Eng. Div., Aire. Br., Bomb. Sect., "Bomber Aircraft," Feb. 18, 1946; Aircraft Projects (Bombers), RG 341, HQ/USAF, NARS. Also, AMC/HO, Case History of XB-36, YB-36, and B-}6 Airplanes, p. 13; Goldberg, 49.

"Quest

It

was estimated

for Better

in

Weapons,"

p.

251.

Memorandum

to C. L. Egtvedt, April 19, 1945, Boeing archives. 51. "Military Characteristics of Aircraft," Memorandum to Asst. Chief of Air Staff for briefing, "B-52 Presentation," no date; both in Files, Materiel, Nov. 17, 1944, and HO/ASD. Also, "B-52 Presentation," Dec. 17, 1948, B-36 Investigation, RG 341, HQ/USAF, 50.

W.

E. Beall,

AMC

NARS. 52.

AMC

Briefing, "B-52 Presentation."

€

Flying Blind

equipped with defensive armament that would leave a minimum

number

of blind spots. Finally,

its

service ceiling

was

to

be 40,000

feet. 53

These requirements were extraordinarily demanding. The most capable bombers then in service, the B-17E and B-29A, had top speeds of

mph and

358 mph, respectively, and they could not come close to meeting these range or payload requirements. 54 The AAF's require-

only 317

ments

medium-range jet bomber called for a top speed of 500-550 mph, not much more than what was required of the long-range bomber. But the former did not have the latter's intercontinental range or heavy payload requirements. Even so, an important technological breakthrough had to be made before the medium-range bomber's requirements could be met (see Chapter 3). In short, the new bomber had to fly farther than the B-35 and B-36, which were originally optimized for long-range flight, but it also had to fly almost as fast as the B-47, which was built for speed. The AAF, moreover, "recognized when these characteristics were forwarded to the industry that the requirements, as stated, were beyond the state of the art at that time, in that no existing power plants capable of the desired cruising speed and range were available." 55 The AAF estimated that the necessary engines would not be available for five to ten years, so it is clear that the AAF was not trying to capitalize on some new techfor a

nological breakthrough

when

it

issued

its

requirements in

late 1945. 56

These requirements were not poised on the cutting edge of technology; they were set far beyond it.

The

AAF was

nonetheless anxious to begin developing a new bomber at once. It initiated a design competition in February 1946, and it gave contractors only two months to put together proposals. Three contractors submitted proposals this time around: Boeing, Martin, and Convair (as Consolidated was generally known after 1943). 57 The industry's sud-

den interest in the project can be explained by the fact that, although it was technologically more intimidating than ever, contracts were scarce immediately after the war. These contractors recognized, moreover, that Gen. Alfred Maxwell, Chief, Reqs. Div., Memo to Asst. Chief of Air Staff, Nov. 23, 1945, XB-52 Airplane, RG 18, AAF, NARS. According to the three-eighths rule used for shorter-range bombers, a 3,000-mile radius would have meant a 13,333-mile range. But, the AAF probably did not intend to apply this ratio to the long-range B-52, and indeed it later set a 12,000-mile range requirement for the program. Therefore, the AAF probably had a 12,000-mile range requirement in mind at this stage of the program. 34. Swanborough and Bowers, United States Military Aircraft, pp. 87-103. 53.

Brig.

55.

AMC

36.

ATSC, Wire

Briefing,

March

to Asst.

29, 1949, Files,

Chief of Air

Staff,

HO/ASD. Feb.

2,

1946, Experimental

Bombers,

RG

341,

HQ/USAF, NARS. The Vultee Aircraft Corporation bought a controlling interest in Consolidated in December 1941. The two companies officially merged in March 1943 and were generally 57.

known

as Convair thereafter.

[126]

The

First Intercontinental

Bombers

new bomber was

intended to be the mainstay of the postwar air force. The AMC concluded at the end of May that Boeing's proposal came closest to meeting the AAF's requirements, and a contract for preliminary development work on the B-52 was signed in June. 58 the

The B-52 and the Long-Range Bomber Competition The B-35 and B-36 programs continued to struggle while the B-52 program got under way. Since the B-35 itself had no future as a production bomber, the AAF and Northrop agreed to build an all-jet version that would address the AAF's growing interest in high-speed capabilities. The AAF, which had always hoped that the flying wing would produce a performance breakthrough, authorized the development of the jet-powered flying wing, the B-49, in 1945. Northrop predicted that the B-49 would make its first flight in June 1946. As it turned out, the B-35 finally made its first flight in June 1946, and the B-49 did October 1947. 59 Progress on the B-36 was also slow in the 1944-46 period. There were several technical problems with the prototype, some requiring major design changes. Its wing, tail, and landing gear, for example, all had to be redesigned late in the game. The AAF was forced to acknowledge in December 1945 that engineering on the prototype had not jelled and that the B-36 was still in a developmental stage. 60 The magnitude of the program's residual technological problems, therefore, was badly undernot take to the

air until

George E. Price, Chief, Aire. Proj. Sect., Eng. Div., Letter to Boeing Aire. Co., Feb. 13, 1946, and Brig. Gen. L. C. Craigie, Chief, Eng. Div., AMC, Memorandum to CG, AAF, May 26, 1946; both in XB-52 Airplane, RG 18, AAF, NARS. See also Warren E. 58.

Col.

1945-1953, Historical Branch, ARDC, May 1956, p. 4; L. E. Preston, Contract Negotiations and Results in Aircraft Procurement, Rand Corporation Research Memorandum, RM-3254-PR, Sept. 1962, p. 40. The Air Technical Service Command became the Air Materiel Command in 1946. For an overview of these

Greene, The Development

of the B-52 Aircraft,

organizational changes, see the appendix. 59. Lt. Col. H. E. Warden, Aire. Proj. Sect., Memorandum to Lt. Col. L. V. Cook, Eng. Sect., March 2, 1945, and Maj. F. E. Loudy, Aire. Proj. Sect., Memorandum to Brig. Gen. L. C. Craigie, Eng. Div., May 28, 1945; both in XB-35 Airplane, RG 18, AFF, NARS. See also

ARDC,

Air Force Developmental Aircraft, April 1957, pp. 139-151, 193-199. 60. Convair changed the airfoil section to improve the performance of the aircraft; this, in turn, forced a redesign of the entire wing. A single tail was substituted for the twin tail in the original design to keep the weight of the aircraft under control. The landing gear had to be redesigned toward the end of the prototype program because the AAF belatedly discovered that its runways could not handle the B-36's weight if it was concentrated at

couple of points. See Col. G. A. Hatcher, Chief, Aire. Sect., Procurement Div., ATSC, Memorandum to Brig. Gen. E. W. Rawlings, Chief, Procurement Div., Dec. 29, 1945, and Consolidated-Vultee Aire. Corp., Report to CG, AAF, March 21, 1946, cited hereafter as Consolidated Report; both in XB-36 Airplane, RG 18, AAF, NARS. Also, AMC/ HO, Case History of XB-36, YB-36, and B-36 Airplanes, p. 17.

just a

Flying Blind

when

was made to begin production, and the concurrent production program compounded the problem. Manpower and materials were in short supply during this period, and the effort that went into gearing up for production inevitably diverted scarce resources from engineering development. The AAF worried that the lack of visible progress in the program would give Congress an excuse to cancel it once the war was over, so it decided in early 1946 that the only changes that would be made in the B-36 would be those required "to make the thing fly." 61 The first flight of the B-36 finally took place in August 1946, twenty-seven months behind schedule and just as the B-52 program was estimated

getting

the decision

under way.

Unfortunately for the AAF, development and production budgets were extremely tight in the early postwar years, which meant that it might not be able to deploy both the B-36 and B-52. As a result, some competitive pressures were brought to bear on Convair and Boeing. Convair had a five-year jump on Boeing in sorting out its aircraft's basic

development problems, and the B-36 was already moving, however fitfully, toward production. Convair realized that, if the B-36's performance could be improved to a point where it was roughly comparable to the B-52's, the rationale for building the B-52 as a follow-on system would be undermined, the B-52 would probably be canceled, and huge production orders would come Convair's way. Boeing, for its part, had to make sure that the B-52's performance was substantially better than the B-36's in order to justify a

new production program.

Since the B-36

already pushed the state of the art in several respects, this was an inherently difficult proposition. The B-36 consequently had some advantages over the B-52 as the

two programs began

to

compete

for the

AAF's

affections in the late i94o's.

B-52 Designs

Boeing proposed over one dozen different designs for the B-52

in the

meet the AAF's increasingly demanding performance requirements and stay one step ahead of Convair. Model 462, which won the 1946 design competition, was a conventional aircraft with straight wings and six turboprop engines. Boeing estimated its top speed at 440 mph, quite close to the 450-mph requirement, and its cruising speed at 410 mph, substantially better than the 300-mph capability for which the AAF had asked. Model 462 essentially met the AAF's late

1940s as

it

tried to

E. W. Rawlings, Chief, Procurement Div., to Maj. Gen. E. M. Powers, Gen. Phillips, Chief, Mat. Div., to Brig. Gen. O. R. Cook, May 23, 1945; Consolidated-Vultee Aire. Corp., Letter to Dir., ATSC, Jan. 30, 1945; CG, ATSC, Wire to Asst. Chief of Air Staff, May 23, 1945; in XB-36 Airplane, RG 18, AAF, NARS.

61.

Brig.

Gen.

Jan. 2, 1946; also,

The

First Intercontinental

Bombers

speed requirements, but it fell far short in terms of range. Although the AAF had stressed the speed requirement throughout the design competition, it focused more on Model 462's range deficiencies as 1946 progressed. Boeing responded by proposing Model 464, a smaller, fourengine bomber with better range. 62 The new bomber's range was the main focus of a November meeting attended by Boeing representatives and several senior members of the air staff, including Gen. Curtis LeMay, the deputy chief of staff for research and development. LeMay stressed the importance of deploying some bombers that were not totally dependent on forward bases. Since Model 462 did not have intercontinental capabilities, LeMay suggested that two bombers be built. The first would be a versatile, general-purpose bomber capable of carrying up to 100,000 pounds of conventional bombs on short-range missions. It would need all-around defensive armament because it would be involved in sustained military operations. LeMay also proposed building a squadron of special-mission

bombers to complement the general-purpose force. The special-mission bomber would be designed specifically for the atomic bomb, which would ease its payload requirement. In addition, it would need only minimal defensive armament because its special operations would have, "the advantage of surprise." 63 The reduced payload and it was said, defensive armament requirements would enable the bomber to fly intercontinental distances. LeMay's recommendations were formally approved by the AAF in December. Boeing was directed to conduct design studies of both bombers under its original contract. The AAF went on to specify that the special-mission bomber should have a 400-mph cruising speed

64 in addition to a 12,000-mile range.

Once

again, the

and range

mph

AAF

capabilities.

ran roughshod over the trade-off between speed It

raised

its

cruising speed requirement from 300

400 mph, turning Model 462's performance bonus into a permanent requirement, and it simultaneously reemphasized long-range capabilities. No allowance was made for the fact that Model 462's speed

was

to

originally attained at the

expense of range.

Boeing presented its designs for the special-mission bomber, Model 464-16, and the general-purpose bomber, Model 464-17, to the AAF in January 1947. Boeing estimated that a stripped-down version of the Model 462 had an estimated radius of only 3,560 miles, short of the 5,000-mile requirement; see L. C. Craigie, CG, AMC, Memorandum to CG, AAF, Nov. 26, 1946, and 62.

Gen. Alfred R. Maxwell, Memorandum to Gen. Partridge, Nov. 27, 1946, cited hereafter as Maxwell XB-52 Memorandum; both in XB-52 Airplane, RG 18, AAF, NARS. Brig.

See also Greene, Development

of the B-52, pp. 4-5.

Maxwell Memorandum. 64. Maj. Gen. E. M. Powers, Asst. Chief of Air XB-52 Airplane, RG 18, AAF, NARS. 63.

Staff,

Memo to CG, AMC,

Dec.

7,

1946,

>

Flying Blind

general-purpose bomber could carry a io,ooo-pound payload 12,400 miles, provided that it used droppable external fuel tanks. It could also carry a 90,000-pound payload 7,500 miles. The AAF consequently decided to forego the special-mission bomber altogether and proceed with the

development of Model 464-17. 65 The B-36 underwent one of its periodic reviews around this time. Gen. George Kenney, SAC's commanding general, wrote to Gen. Carl Spaatz, commanding general of the AAF, in December 1946 complaining about the B-36 's range. As Kenney later recounted, "I could find no indications from our own engineers or those in the industry that anyone really expected that the B-36 would make its originally predicted 10,000 mile range. Even the Consolidated people talked about 8,000 mile ranges as about what we could expect." 66 Other reports suggested that the B-36's useful range might only be 6,500 miles and that its top speed would be 50 mph lower than originally guaranteed. In addition, the program was accumulating significant cost overruns, an important consideration given the state of the defense budget. Kenney recommended canceling the B-36 production contract and building only a few prototypes. 67 Gen. Nathan Twining, the commanding general of the AMC, defended the B-36 by arguing that, since the B-52 would not be available until 1953 or 1954, the B-36 was needed to bridge the gap. Although Spaatz sided with Twining in deciding to continue B-36 production, it was clear that the B-36 's performance would have to improve if the program was to survive. It was not a coincidence, therefore, that a more powerful engine was proposed for the B-36 in March 1947. Convair estimated that, with the new variable discharge turbine (VDT) engine, the B-36 would have a top speed of 410 mph and a range of 10,000 miles. This was a substantial improvement over the projected top speed of the B-36B, 325 mph, and it compared favorably with the projected top speed of the B-52, 440

mph. 68 The B-52 program

keep up with the AAF's requirements. LeMay felt that, if the B-52's range fell below 12,000 miles, the program would have to be reevaluated. 69 The June 1947 performance estimates for Model 464-17 predicted a range of only 11,600 miles. The 65. 18,

Maj. W. C. Brady,

also struggled to

Memorandum

for the Record, Jan. 7, 1947,

AAF, NARS. Also, Greene, Development of the B-32, p. 66. Kenney quoted in SAC/HO, The B-36, Aug. 1951,

XB-52 Airplane,

RG

5.

p. 5. See also AMC/HO, Case History of B-36 Airplane, pp. 7-9. 67. See AMC/HO, Case History of B-36 Airplane, pp. 7-9. Also, AMC/HO, Case History of XB-36, YB-36, and B-36 Airplanes, p. 15; Maj. Gen. E. M. Powers, Asst. Chief of Air Staff, Letter to CG, AMC, March 25, 1946, XB-36 Airplane, RG 18, AAF, NARS.

Based on the account in AMC/HO, Case History of B-36 Airplane, pp. 7-10. 69. Maj. Gen. Curtis E. LeMay, Dep. Chief of Staff, Letter to Lt. Gen. Nathan F. Twining, CG, AMC, May 15, 194 7 XB-52 Airplane, RG 18, AAF, NARS; cited hereafter as 68.

LeMay

Letter.

The

First Intercontinental

Bombers

AAF was ice,

consequently afraid that the B-52 program was "on very thin considering the ultimate cost of the project." 70 It concluded that a

stringent weight-control

was going

to

program would have

have the range

At the same time, the

it

to

be instituted

if

the B-52

needed.

AAF was

concerned about the B-52's vulnerability to Soviet air defenses. The basic problem was that, given its projected speed, the B-52 would spend nine hours within range of air defense interceptors during a typical mission. There were three basic ways of addressing this problem. First, the bomber's cruising speed could be increased, which would cut the amount of time it would spend in the danger zone. Second, its top speed could be increased, which would help it to escape from enemy interceptors. Third, the bomber's defensive armament could be increased. Increasing the B-52's require-

ments

in

any

of these areas

would

of course cut into

its

range.

The

AAF

nonetheless decided in June 1947 to increase the B-52's cruising speed requirement from 400 to 420 mph, although it softened the blow by lowering the top speed requirement from 450 to 420 mph. 71 This was not the AAF's final answer to

questions about B-52 survivability, however, and the internal debate over range, speed, and defensive armament continued to simmer throughout the rest of the year. its

Boeing continued with its design effort while these deliberations were going on. Models 464-23, 464-25, and 464-27 culminated in Model 464-29, which appeared on the scene late that summer. Boeing estimated that Model 464-29 would have a top speed of 455 mph, which exceeded the AAF's June requirements, and an operating radius of 5,000 miles, equivalent to a 12,000-mile range. 72

The Aircraft and Weapons Board

The National Security Act of 1947, passed in July, led to the establishment of the independent U.S. Air Force in September of that year. In one of his first acts as deputy chief of staff of the new Air Force, Gen. IToyt Vandenberg created the Aircraft and Weapons Board, a senior body that was to oversee the Air Force's many procurement programs. The board, Maxwell, Chiefs, Reqs. Div., Memorandum to Res. and Eng. Div., Asst. Chief of Air Staff, April 21, 1947; Maj. Gen. E. M. Powers, Asst. Chief of Air 70.

Brig.

Gen. Alfred

R.

Memorandum to CG, AMC, April 25, 1947; A. Boykin, Jr., Aire. Proj. Sect., Memorandum Report, June 23, 1947, cited hereafter as Boykin Memorandum Report; Maj. Gen. L. C. Craigie, Chief, Eng. Div., AMC, Memorandum to CG, AMC July 11, 1947, cited hereafter as Craigie XB-52 Memorandum; all in XB-52 Airplane, RG 18, AAF, NARS. 71. Brig. Gen. Alfred R. Maxwell, Chief, Reqs. Div., Memorandum to Asst. Chief of Staff,

J.

Air Staff, June 23, 1947, and Carl E. Reichart, "Conference on XB-52," Report, Dec. both in XB-52 Airplane, RG 18, AAF, NARS. 72. Greene, Development of the B-52, p. 6.

1,

1947;

.

Flying Blind

Heavy Bombardment Committee, which was directed to take a fresh look at the strategic bombardment mission and the various bomber programs the Air Force had under way. This committee met throughout the autumn before issuing its report in November. The committee's report started by assuming that the Soviet Union might be able to overrun American forward air bases around the perimeter of Europe and Asia with its large ground forces and that Soviet air defense interceptors would be technically equivalent to those of the United States but superior in numbers. Strategic bombers would therefore need both intercontinental and high-speed capabilities if they were to be able to attack the Soviet Union with some degree of reliability. A faster, medium-range bomber would not be able to fly a round-trip mission from the continental United States; one-way missions were considered but rejected. The Heavy Bombardment Committee concluded that the B-52, as presently configured, would be much too slow to pene73 trate and survive Soviet air defenses. Once again, the Air Force was caught between a rock and a hard place. in turn, established the

How

could it raise the B-52's speed requirements without sacrificing range? Fortunately for the Air Force, the emergence of aerial refueling enabled it to finesse this trade-off. According to the committee's calculations, a bomber that could fly 8,000 miles without refueling could fly a round-trip mission to the Soviet Union with only one in-flight refueling. The committee therefore recommended dropping the B-52's unrefueled

range requirement to 8,000 miles and raising its top speed requirement to 550 mph, provided that provisions for aerial refueling were made. The committee also decided, over some protest, that the only defensive armament the B-52 needed was a tail gun. 74 Aerial refueling was not, however, seen as a panacea. SAC was especially reluctant to rely on tankers because they would be the weak link in any operation and because they complicated the planning process. SAC therefore continued to press the case for bombers with true intercontinental capabilities. Given the technologies of the day, however, Although the AAF had been briefed in 1946 about the threat surface-to-air missiles would pose to strategic bombers in the years ahead, it continued to define the air defense threat in terms of fighter capabilities; see U.S. Aircraft and Weapons Board, Summary Minutes of the First Meeting, Aug. 19-22, 1947; Minutes of U.S. Air Force Aircraft and Weapons Board Meeting of Sept. 19, 1947, RG 341, HQ/USAF, NARS. Also, Heavy Bombardment Committee, USAF Aircraft and Weapons Board, Report on Heavy Bombardment Nov. 7, 1947, Long Range Bomber Subcommittee Report, RG 341, HQ/USAF, NARS. See also Edmund Beard, Developing the ICBM (New York: Columbia University Press, 1976), pp. 73.

,

86-87.

See Heavy Bombardment Committee, Report on Heavy Bombardment See also Col. George F. Smith, Chief, Aire. Proj. Sect., Eng. Div., AMC, "Fleavy Bomber Design Study," Memorandum for the Record, Oct. 9, 1947, XB-52 Airplane, RG 18, AAF, NARS. 74.

The

First Intercontinental

Bombers

was simply no other way around the Air Force's dilemma. If a top speed of 550 mph was needed, then aerial refueling was a necessary there

evil.

and Weapons Board decided to continue B-36 production despite the fact that it had little enthusiasm for the B-36 itself. 75 Two considerations led to this decision. First, the B-52 program was up in the air due to the latest round of requirements changes, and the board felt that it was important to have another long-range bomber in the pipeline.

The

Aircraft

from the continental United States, and it could be deployed in a relatively short period of time should the international situation require it. Finally, roughly $129 million had been spent on the B-36 by late 1947, and the Air Force was anxious to get something for its investment. After going back and forth on the issue of the VDT engine, the board decided to install the new, more expensive engine on just 34 B-36S. General Kenney, however, believed that the B-36 should be canceled because he doubted that significant improvements would materialize. 76 The number of B-36S to be built was cut from 100 to 93 to keep total program costs the same. Decisions had to be made about the B-52 program as well. The Air Force had serious reservations about the B-52 even before the Heavy Bombardment Committee met in late 1947, because of the new bomber's In addition, the B-36 could operate

some Air Force estimates predicted that it would take 8-10 years to develop, produce, and deploy a bomber this advanced. Finally, there was an ongoing debate over the B-52's performance capabilities. 77 The committee's recommendation to size,

complexity, and expense. In addition,

change the requirements unsettled the situation even more. In June 1947, the B-52 was supposed to meet 12,000-mile range and 420-mph speed requirements. But the committee called for 8,000-mile range and 550-mph speed requirements. It was not at all clear that an aircraft designed to meet the first set of requirements could be adapted to meet the second. At a minimum, the B-52 would have to be completely redesigned.

Work on the B-52 was stopped in November 1947 while the Air Force decided what to do. Many felt that a new design competition was in order, arguing that Boeing was not entitled to build the new high-speed bomber simply because it had once won a contract for a different bomber. Moreover, it was argued, Boeing was not necessarily the most 75.

According

AMC/HO, Case History of B-36 Airplane, See also AMC/HO, Case History of XB-36,

to

pp. 10-13.

YB-36, and B-36 Airplanes, p. reach an important funding 10-11. B-36 was about to The 31; SAC/HO, The B-36, pp. deadline: the wartime appropriations that had supported the program for many years were scheduled to run out at the end of June 1948. The B-36 would have to compete with the B-52 for new appropriations starting with fiscal year 1949. 76.

Ibid., p. 19.

77.

Heavy Bombardment Committee,

Report on Heavy Bombardment,

p. 3.

Flying Blind

new

qualified contractor to build the

new requirements

aircraft.

Some

believed that the

favored a flying wing design, which Boeing

interested in pursuing.

The

was not

secretary of the Air Force, Stuart Symington,

decided in December to cancel the Boeing contract and proceed with a new design competition. 78 Word of this decision leaked out before it was actually implemented, and Boeing protested in the strongest possible terms. The president of Boeing, William Allen, complained to Symington on December 26 that a new design competition would constitute a serious injustice because it would allow Boeing's competitors to take advantage of his company's work on the B-52. In addition, he argued that Boeing should be allowed to state its case for adapting the B-52 to the new requirements before the Air Force

a final decision. 79

made

While this debate was going on, the Air Force was trying to set the bomber's performance requirements. The Fieavy Bombardment Committee had called for a top speed of 550 mph, a recommendation the requirements division supported. Others believed that it would not be possible to build a bomber capable of flying 8,000 miles with so much speed. Rand Corporation studies indicated that a range of 8,000 miles was theoretically possible, but only if the speed requirement was set at 500 mph. AMC studies, even more discouraging, concluded that a range of only 7,500 miles was possible with a 500-mph speed requirement. It was not at all clear, moreover, how anyone was going to get a turboprop

bomber to fly that fast, given the constraints of propeller technology. The Aircraft and Weapons Board was told that propeller performance at speeds over 450 mph was questionable and that considerable development in engines and propellers would be needed to meet even a 500-

mph

requirement. 80 Nonetheless, in late January 1948 the board settled on 8,000 miles and 500 mph for range and cruising speed requirements.

The top speed was

be whatever was possible above the cruising speed. In setting these requirements, the board overrode the objections 78.

Gen. Hoyt

the Air Force, Dec. Staff for Materiel,

Chief,

Bomb.

S. 8,

to

Vandenberg, Vice Chief of 1947; Col.

J.

Memorandum

S.

Staff,

Memorandum

Holtoner, Chief, Aire.

Br.,

for the Secretary of

Office of the Dep. Chief of

Gen. Craigie, Nov. 28, 1947; Maj. William D. Brady, Dep. Chief of Staff for Materiel, Memorandum 1947, cited hereafter as Brady Memorandum; Maj. Gen. E. E. to

Sect., Aire. Br., Office of the

for the Record, Dec. 2,

Partridge, Dir. of Reqs.,

Memorandum

for the Record, Jan. 8,

1948, cited hereafter as Col. C. S. Irvine, Asst, to the Chief of Staff, Letter to Maj. Gen. L. C. Craigie, Dir. of Res. and Devel., Nov. 28, 1947, cited hereafter as Irvine Letter; all in

Partridge

Memorandum;

XB-52 Airplane,

RG

18,

AAF, NARS.

79. William M. Allen, Pres., Boeing Aire. Co., Letter to Stuart Symington, Secretary of the Air Force, Dec. 26, 1947; also, Maj. Gen. E. E. Partridge, Dir. of Reqs., Memorandum to Dir. of Res. and Devel., June 15, 1948, cited hereafter as Partridge B-52 Program Memoranin XB-52 Airplane, RG 18, AAF, NARS. See also Partridge Memorandum. Irvine Letter. See also Conf. Report, Conf. with the Bombardment Subcommittee,

dum; both 80.

Nov.

18, 1947,

XB-52 Airplane,

RG

18,

AAF, NARS.

The

of

own

its

First Intercontinental

Bombers

AMC

and ignored evidence about raised its speed requirements and set

technical experts in the

what was technically possible. It them beyond the state of the art. 81 By this time, Boeing was ready to

offer a

new bomber

design in order case to the Air

second design competition. Presenting its Force hierarchy in February 1948, Boeing claimed that Model 464-35, a small, four-engine bomber, would be able to meet the new speed and range requirements. The Air Force subsequently reversed its position and decided against holding another design competition. 82 The implicit competition between the B-36 and B-52 put pressure on Convair and Boeing to improve the performance capabilities of their aircraft throughout the 1946-49 period. Convair, under intense and conto forestall a

tinuing pressure in late 1946 and early 1947 to justify building the B-36 even as an interim bomber, proposed equipping it with the advanced

VDT engine,

which would increase the projected speed from 325 to 410 mph. Boeing was also under considerable pressure in late 1946 and 1947. Influential people in the Air Force

argued that

it

was

essential for the

B-52 to have a real intercontinental capability in addition to impressive speed. Boeing consequently scrambled to develop a whole series of

designs in 1947, leading up to Model 464-29. At this point, the B-52 appeared to be clearly superior to the B-36 in terms of range (12,000 to

somewhat better in terms of speed (455 to 410 mph). and Weapons Board introduced a new factor into the

8,000 miles) and

The

Aircraft

competition in late 1947. Its acceptance of aerial refueling made unrefueled range less critical, because a bomber with an unrefueled range of 8,000 miles could fly an intercontinental mission with one in-flight refueling. With a stroke, then, the B-36's 8,000-mile range became adequate and the B-52's main competitive advantage was neutralized. In addition, the B-36's range had already been proven in flight tests, while the B-52's had not. For the B-52 to be worth building, it had to offer a substantial improvement over the B-36 in some critical area of performance; otherwise, a follow-on program could not be justified. The advent of aerial refueling, therefore, shifted the focus of the competition

from range to speed. Unfortunately for Boeing, the estimated speed of the B-36 equipped with the VDT engine was not unimpressive. This situation, in turn, influenced the Air Force's decision about the B-52 's performance requirements. Once the Air Force accepted aerial

81.

Brig.

Gen.

F.

H. Smith,

Jr.,

Secretary,

USAF

and Weapons Board, Memo to XB-52 Airplane, RG 18, AAF, NARS.

Aircraft

the Dep. Chief of Staff for Materiel, Dec. 8, 1947, Also, Brady Memorandum. 82. Lt. Gen. H. A. Craig, Dep. Chief of Staff for Materiel, Memorandum for Vice Chief of Staff, Feb. 17, 1948, XB-52 Airplane, RG 18, AAF, NARS. See also Partridge B-52 Program Memorandum; Greene, Development of the B-52, p. 7.

Flying Blind

B-36C (the VDT version of the B-36) it could afford to raise the B-52's speed requirements. In fact, once it took these steps, the Air Force had to raise the B-52. s speed requirements, it could no longer justify building a follow-on bomber with speed cahad to pabilities of only 420-455 mph. The Aircraft and Weapons Board refueling

and decided

to build the

transform the B-52 into an even more ambitious program in order to distinguish

it

from the B-36.

was also affected by the beginning of the B-47 flight testing program. It was hard for the Air Force to justify building a longrange bomber with a top speed of 455 mph while it was arguing that substantially better speed capabilities were required in the B-47. The advent of aerial refueling, moreover, made the medium-range B-47 ess dependent on forward bases and more capable of covering the Soviet target set. Aerial refueling, therefore, brought the medium-range bombanother er into more direct competition with the long-range bomber This decision

^

—

reason for the B-52's speed to be competitive. Otherwise, the Air Force ran the risk of having the program canceled by a fiscally conservative president and Congress. Naturally, the Air Force argued that there were legitimate strategic reasons for changing the B-52's requirements. It maintained that the

B-52 had to

fly

defenses.

was

It

500

mph

to penetrate the Soviet

certainly true that the Soviet

Union's improving

air

Union had embarked on

a

defenses in the early postwar period, and that the Air Force was aware of at least the broad parameters of this effort. The Air Force was also aware that the Soviet Union had acquired advanced jet fighters from Germany at the end of the war, and

vigorous program to modernize

its air

conclude that the Soviet Union would eventually deploy high-speed jet interceptors of its own. It was certainly prudent for the Air Force to begin developing high-speed bombers in anticipation of this development. But it is hard to argue, as the Heavy Bombardment

it

was reasonable

Committee

to

did, that the strategic situation forced

it

to revise the B-52's

speed requirements in January 1948. After all, the B-52's top speed requirement had been lowered in June 1947 us f a few months before the committee began its deliberations. There is no indication in the records of the Aircraft and Weapons Board or its Heavy Bombardment Committee that the Air Force was reacting to new intelligence about a sudden development on the Soviet side when it raised the B-52's speed require'

ments

j

in early 1948.

Modifying the B-36

Competitive and bureaucratic pressures continued to shape the B-36 and B-52 programs throughout 1948. Convair's prospects in the com[136]

The

petition cult to

First Intercontinental

were

largely

Bombers

dependent on the

adapt to the B-36. As a

VDT engine,

result, the

which proved

performance estimates

diffi-

for the

B-36C deteriorated badly. The Air Force estimated in April that the B-36CS top speed would be 385 mph instead 410 mph, and that its cruising speed would be 262 mph instead of 322 mph. Its range, furthermore, would be 1,800 miles less than earlier predictions had indicated. The B-36CS performance would actually be inferior to that of the B-36B (without the

VDT

engine) in

some

respects. Cancellation of the

VDT

experiment became a foregone conclusion. It was, as the Air Force came 83 to recognize, a complete failure. The cancellation of the B-36C reopened the issue of what to do about the B-36 program as a whole. One option was to cancel the entire production run except for the 22 aircraft that had already been built and then wait for the B-52. Fortunately for Convair, long-range flight tests in April and May demonstrated that the B-36 had an unrefueled range of over 8,300 miles even without the improved engines that were to be

on the B-36B. It was also fortuitous that a final decision on the fate of the B-36 was put off until June 24, the day the Soviet Union blockaded West Berlin. At that point, even Kenney agreed that the first 84 95 B-36S had to be built. Although the B-36 earned a reprieve from the Air Force, the fate of the program was still far from secure. Once the VDT experiment collapsed, the B-36 was vulnerable to criticism that it was too slow to be militarily effective. It seemed to be barely adequate as an interim bomber, and it installed

certainly did not look promising as a long-term alternative to the B-52. Convair responded to these concerns on October 5 by proposing a ten-

engine, jet-assisted version of the B-36. It suggested adding two of the twin-engine pods that were being developed for the B-47 to the B-36. Convair predicated that this would boost the B~36's top speed to 435 mph and that a prototype could be ready to fly in four months. 83 The Air Force approved Convair's proposal, and development of the B-36D began.

Redesigning the B-52

Boeing was under considerable pressure from the Air Force in mid-1948 because Model 464-35 began to evolve in disturbing ways. By June, the bombers projected cruising speed had dropped to 445 mph

AMC/HO,

Case History of B-36 Airplane, pp. 17-20, 23-29. for the Record, June 25, 84. Gen. Muir S. Fairchild, Vice Chief of Staff, Memorandum AMC/HO, Case History of B-36 1948, B-36 Investigation, RG 341, HQ/USAF, NARS. See also Airplane, pp. 29-36. 83. AMC/HO, Case History of B-36 Airplane, p. 40. 83.

Flying Blind

range had crept back up to 11,600 miles. In other words, it began to resemble the bomber the Air Force had explicitly rejected just a few months before. Some accused (although

its

top speed

was 500 mph) and

its

Boeing of using bait-and-switch tactics to dupe the Air Force into canceling the proposed design competition. As the Air Force's director of requirements put it, "the Boeing Company is giving us the old B-52 with a

new

coat of paint." 86

At the same time, the Air Force was thinking about turning away from turboprop bombers such as Model 464-35 altogether. Although it had been interested in turbojets for years, jet engines had been unsuitable for long-range aircraft prior to this time because they consumed too much fuel. This problem was beginning to appear surmountable in 1948, however, with the development of more efficient turbojet engines and the emergence of aerial refueling. The Air Force consequently asked Boeing in May to make a preliminary study of a long-range turbojet bomber based on the design of Model 464-35. The result was Model 464-40, which, like all of Boeing's designs to date, had a conventional fuselage and fairly straight wings. 87 Turbojet engines were simply substituted for turboprops. The effect of this switch on the aircraft's performance estimates was fairly predictable. Model 464-40's top speed was higher than its counterpart's, but its range was only 6,750 miles. Boeing submitted its design study for the jet bomber in July and continued to work on the turboprop bomber already under contract. 88 Convair's B-36D proposal turned the tables on Boeing, however. The new, improved B-36 neutralized some of the B-52's speed advantages, and the erosion of Model 464-35's cruising speed compounded Boeing's problems. As a result, the B-52's competitive position was suddenly tenuous. The Air Force rejected both Boeing models in October. Once again, Boeing responded to these competitive and bureaucratic pressures, this time in dramatic fashion. It dispatched a team of its best aircraft designers and engineers to Wright Field to find out how the B-52 needed to be changed. Boeing's engineers arrived at Wright Field on Thursday, October 21, less than three weeks after Convair made its B-36D proposal. They discovered that the Air Force wanted a radically improved, high-performance jet bomber. They subsequently retired to

—

—

and spent the weekend redesigning the B-52. Drawing on their experience with the B-47 program, they abandoned the straight wing and conventional fuselage of their earlier B-52 models and adopted their hotel

Program Memorandum. wings were swept 20 degrees

86.

Partridge B-52

87.

The

B-52's

at this time;

Greene, Development

of the

B-52, pp. 8-9. 88. A. G. Carlson, Chief Proj. Eng., Airplane, RG 18, AAF, NARS.

Memorandum

to

CG, AMC,

July 28, 1948, XB-52

[138]

The

First Intercontinental

Bombers

Boeing B-52 (Boeing

Company

Archives)

wing and streamlined fuselage of the B-47. The wings of the B-52 were swept back 35 degrees, and eight jet engines were put in four nacelles under the wings. Boeing's engineers submitted their new design, complete with a thirty-three page report and hand-carved model, to the evaluation committee on Monday morning, October 25. They estimated that the redesigned B-52, Model 464-49, would have a range of 8,000 miles, a cruising speed of 520 mph, and a top speed of 572 mph. Some estimates suggested a top speed of 610 mph. 89 Of all the changes made in the design of the B-52 over the years, this was far and away the most radical. Although Boeing's engineers were able to draw on their experience with the B-47 when they redesigned the B-52, the idea of building a long-range turbojet bomber was still bold, to say the least. The B-47 program itself was technologically ambitious, and it was not confronted by the B-52's range and payload requirements. These were precisely the sorts of requirements for which turbojet aircraft were not well suited. It was not a coincidence that Boeing's new design for the B-52 materialized soon after Convair's B-36D proposal and the Air Force's rejection the swept

89.

Based on the accounts

in

Greene, Development

of the B-52, pp. 10-11;

Harold Man-

The Story of Boeing (New York: Duell, Sloan, and Pearce, 1966), pp. 175-195; Margaret C. Bagwell, The XB-52 Airplane, HO/AMC, Aug. 1949, pp. 2-3. See also, Lt. Gen. H. A. Craig, Memorandum for Gen. Norstad, Oct. 16, 1948, XB-52 Airplane, RG 18, AAF, sfield, Vision:

NARS.

,

Flying Blind

of Boeing's other B-52 designs. Boeing's chief aerodynamicist, Schairer, claimed that Air Force

procurement

officers

were

George

explicit in

Boeing that its recycled B-52 designs did not promise enough improvement over the B-36D to warrant production. As Schairer noted, Boeing would not have come up with a radically redesigned B-52 in October 1948 without the competitive pressure of the B-36. 90 Boeing's gamble paid off, and the Model 464-49 was well received by the Air telling

Force.

The B-49 Option

There was one other contender

procurement dollars as took to the air in October

for Air Force

drew to a close: the B-49. The B-49 finally 1947, and one encouraging piece of information came out of its flight testing program. The B-49's drag was lower than expected, so its speed would probably be in the neighborhood of 500 mph. The Air Force, deep in the deliberations of the Aircraft and Weapons Board and the Heavy Bombardment Committee at the time, was looking for a high-speed bomber, so interest in the flying wing was rekindled. In March 1948 Northrop proposed a version of the flying wing suitable for both reconnaissance and bombing, the RB-49. The Air Force accepted this proposal in April, and by June it was in the process of buying 30 1948

RB-49S. 91

begin producing the RB-49 even though the vast majority of the reports from its flight test program were discouraging. In particular, the B-49 had serious stability and flight control problems. It tended to pitch and yaw violently at high speed because it did

The Air Force decided

to

Indeed, one of the B-49 P ro_ totypes crashed in June 1948, and one report claimed that the aircraft was somersaulting out of control just before it hit the ground. 92 The Air Force nonetheless continued with its production plans for the RB-49. I n not have a conventional fuselage or

October

its

own

test pilots

tail.

found the B-49 extremely unstable and

diffi-

because of these control problems. The pilots constantly had to fight to control the aircraft and even then could not hold a steady

cult to fly

90.

Competition

Defense Procurement Hearings before the Subcommittee on Antitrust 91st Cong., 1st sess., July 1969, pp. 128-129. Case History of 8-36 Airplane, p. 34; ARDC, Air Force Developmental Airin

and Monopoly, Senate Judiciary Committee, 91.

AMC/HO,

Ardath M. Morrow, Case History of the YB-35, YB-49 Airplanes, HO/AMC, Feb. 1950, pp. 7-13. See also John K. Northrop, Letter to Gen. Carl A. Spaatz, Chief, USAF, Dec. 19, 1947, YB-35, YB-49 Airplanes, RG 18, AAF, NARS; Maj. Gen. F. O. Carroll, Dir. of Res. and Devel., Eng. Div., AMC, Memorandum to Chief of Staff, USAF, Feb. 12, 1948, XB-32 Airplane, RG 18, AAF, NARS. 92. ARDC, Air Force Developmental Aircraft, pp. 144, 193-198; Morrow, Case History of YB-35, YB-49 Airplanes, pp. 7-13.

craft,

pp. 193-198;

The

First Intercontinental

Bombers

Northrop B-49 (USAF Photographic Collection, National Air and Space Museum, Smithsonian Institution)

It eventually became clear that course or constant speed and altitude. the RB-49 was not stable enough to be used for either reconnaissance or bombing. The flying wing's lack of aerodynamic stability was so serious that it interfered with bombing accuracy, which was important to the Air

Force even in the atomic age. Clearly, the decision to begin

producing the RB-49 before

its

flight

program was completed was a mistake. Why did the Air Force do this? First, it was anxious to deploy a high-speed bomber, and until late 1948 the B-36 was not particularly fast. The B-52, for its part, would not be available for many years. The RB-49 seemed to combine impressive performance with early delivery. As the international situation deterio-

testing

rated over the course of 1948, the Air Force's anxiety about deploying some kind of high-speed, long-range bomber only increased. Second,

program would be relatively straightforward because the B-35 had been flying for some time. But the assumption that the B-49 was aerodynamically identical to the B-35 was simplistic. As Northrop and the Air Force eventually found out, the B-35's propeller shafts and huge propellers helped to stabilize it. the Air Force

assumed

that the RB-49's flight testing

Air Force Developmental Aircraft, pp. 193-198; Morrow, Case History ofYB-35, Y B-49 Airplanes, pp. 7-13. Also, Aire. Br., Dep. Chief of Staff for Materiel, Memorandum to Brig. Gen. D. L. Putt, Dir. of Res. and Devel., Dep. Chief of Staff for Materiel, Nov. 8, 1948, 93.

ARDC,

YB-35, YB-49 Airplanes,

RG

18,

AAF, NARS.

Flying Blind

be added to the B-49 to compensate for the lack of these stabilizers, and even then it was not as stable as the B-35. Third and last, the Air Force wanted to get something out of its $67 million investment in the flying wing program. 94 Its investment in the flying wing was high because the AAF had committed itself to developing and Vertical fins

had

to

producing the B-35 concurrently in the early 1940s. Although the B-35 production program was eventually canceled, the money spent on it gave the Air Force an added incentive to move toward production of a flying

wing

in 1948.

B-36 Production and B-49 Cancellation

Competition between the B-36, B-49, and B-52 converged in late 1948. The Air Force could not build large numbers of all three bombers because it was under budgetary pressure from the Truman administration. The administration felt that excessive military spending would weaken the economy and ultimately undermine national security. Defense budgets were consequently kept to an average of about $13 billion per year throughout the early postwar period. Although these budgets were lavish by prewar standards, they would not allow the Air Force to deploy 70 aircraft groups, its long-standing goal. 95 This goal was endorsed by the president's Air Policy Commission (also known as the Finletter commission) and the Joint Congressional Aviation Policy Board (the Brewster-Hinshaw committee), both of which argued in early 1948 that the deployment of a 70-group air force was vital to national security. The Air Force felt that these endorsements gave it a mandate to proceed towards the 70-group goal, and it structured its procurement plans accordingly. Congress even appropriated a supplemental Air Force budget of $822 million in the spring of 1948 as a down payment on the 70-group program, but the Truman administration refused to release these funds on the grounds that the Air Force's program

would cost too much. Although the Air Force had hoped to proceed toward the 70-group goal, however slowly, budgetary ceilings set by the administration in ARDC,

Air Force Developmental Aircraft, p. 144. 95. This section is based on Samuel P. Huntington, The Common Defense (New York: Columbia University Press, 1961), pp. 25-47; Warner R. Schilling, “The Politics of National Defense: Fiscal 1950," in Warner R. Schilling, Paul Y. Hammond, and Glenn H. Snyder, Strategy, Politics, and Defense Budgets (New York: Columbia University Press, 1962), pp. 2853, 71-79, 135-213; Walter S. Millis, ed., The Forrestal Diaries (New York: Viking, 1951), pp. 94.

492-530;

Herman

S.

Wolk, Planning and Organizing

the Postwar Air Force,

1943-1947 (Wash-

ington: Office of Air Force History, 1984), pp. 45-79; Perry McCoy Smith, The Air Force Plans for Peace, 1943-1943 (Baltimore: John Hopkins Press, 1970), pp. 54-74; Weigley, American

Way

of War, pp.

363-381.

[142]

The

Bombers

First Intercontinental

1948 threw its plans into reverse. The administration announced in the spring of 1948 that the ceiling for the fiscal year 1950 defense budget

would be $15 billion. Since some of this money was earmarked for stockpiling and unexpected costs on earlier contracts, available funds came to $14.2 billion, about 60 percent of what the services had originally requested. When it became clear late in 1948 that Truman was not going

to raise this ceiling, the Air Force

operational inventory from 59

had

to consider cutting its

current level) to 48 groups. Since the new budget had to be presented to Congress in January 1949, the Air Force had to decide on its force structure plans in December 1948. This

was

(its

the setting for the Air Force's decisions about the B-36, B-49,

and

B-52 programs. Force's Board of Senior Officers held a series of meetings

The Air starting in

December 1948

new commanding

to

address these issues. Curtis LeMay, SAC's

recommended that four of the Air Force's 48 groups be heavy (or long-range) bomber groups and that one group of strategic reconnaissance aircraft be deployed. It was generally agreed that the RB-49 was not suitable for either of these missions, so its production program was canceled. 96 The board decided that all heavy bomber groups would be equipped with B-36DS and that the strategic reconnaissance group would be equipped with RB-36DS, even though the jet-assisted version of the B-36 had not yet flown. The board also strongly supported the development of what it saw as the heavy bomber general,

swept-wing B-52. 97 At a second round of meetings in February 1949, the board canceled the medium-range B-54 and replaced the group of RB-54S in the plan with a second group of RB-36DS. It also decided to increase the number of aircraft in all six of the B-36 groups from 18 to 30. Since some aircraft would be used for training and as replacements, the board decided to buy 138 B-36S beyond the 95 already on order. This was a turning point in the history of the B-36 program. Although it had sputtered along for years, surviving mainly on leftover wartime procurement funds, the of the future, the turbojet,

B-36 its

was

finally established in the Air Force's plans as, at the bomber. 98

very

least,

interim long-range

by this time that full-fledged RB-49S would not be available until flying wings 1953, when they would have to match up against the B-52. Ultimately, fifteen were built (the two original B-35 prototypes and thirteen other B-35 prototypes, three of which were converted into B-49S) at a total cost of $83.9 million; see ARDC, Air Force 96.

It

was

also clear

Developmental Aircraft, pp. 151, 197. and Jan. 3-6, 1949, 97. Board of Senior Officers, Summary Minutes, Dec. 29-31, 1948, Program, RG 341, Group Final Report of Board of Officers on the Composition of the 48

HQ/USAF, NARS. 98. This commitment was critical because the funding that had supported the program for many years had run out in June 1948. The B-54 was a modified version of the B-50, which in turn was a modified version of the B-29. The extensive deliberations over the

Flying Blind

would have been canceled had it not been redesigned by Boeing in October 1948. The turboprop B-52 was only marginally more capable than the B-36D, and the Air Force would not have been able to support a marginal program in fiscal year 1950. The turbojet B-52, on the other hand, appeared to represent a major breakthrough in long-range bomber design. The board, though, felt that it was important to keep some competitive pressure on Boeing, and it It

seems

clear that the B-52

decided in February that B-36 production should be extended until the B-52 itself was ready for production."

The B-36/ Supercarrier Controversy

The decisions

were ratified by the civilian leadership in the Department of Defense. The decisions about the B-36 proved to be highly controversial, however, as a longsimmering conflict between the Navy and the Air Force over roles and missions came to a boil in the spring of 1949. The Navy had been wary of long-range bombers since the 1920s, but its concerns heightened after World War II. It felt threatened by claims that strategic bombardment would undoubtedly be decisive in the next war, given the advent of the atomic bomb, and that the strategic bomber had therefore replaced the capital ship as the country's first line of defense. The Navy's worst-case scenario was that air forces would come to be seen as the key to national security. If this came to pass, the Navy's military missions would be deemphasized, its budgets would suffer, and the Navy itself might be subsumed organizationally. Unease about these possibilities was aggravated by the strong support that existed in many quarters for stronger air forces and by the establishment of the U.S. Air Force in 1947. 100 The centerpiece of the Navy's postwar defense plans was a new generation of large, flush-deck "supercarriers" that would be capable of handling fairly long-range aircraft. The Navy, which originally received approval for the new carriers in 1946, saw them as the key to preserving its of the Air Force Board of Senior Officers

long-term bureaucratic position and, although this was rarely made explicit, the key to its participation in the strategic bombardment mission.

military effectiveness and cost-effectiveness of the various options are summarized in Report of a Board of Officers Convened to Consider the Production Program and the Research and Development Program for the USAF, Feb. 21-24, 1949/ Report of Board of Officers, RG 341,

HQ/USAF, NARS. See

also Gen. Curtis LeMay, CG, SAC, Letter to Gen. Hoyt S. Vandenberg. Chief of Staff, USAF, Feb. 2, 1949, B-50 and B-54, RG 341, HQ/USAF, NARS. 99. Report of a Board of Officers. 100. For a

during the

more

late 1940s, see Paul Y.

priations, Strategy,

between the Air Force and the Navy Hammond, "Super Carriers and B-36 Bombers: Appro-

detailed analysis of the conflict

and

Politics," in

Harold Stein, ed., American Civil-Military Decisions

(Birmingham: University of Alabama Press,

1963), pp. 465-567.

The

First Intercontinental

Bombers

Secretary of Defense Louis Johnson's decision in April 1949 to cancel the supercarrier therefore represented a direct threat to the Navy's bureau-

removed sweepstakes. The fact

cratic standing,

strategic

and

it

the

Navy from

that

Johnson had

Force's plans to proceed with the B-36 only agitation.

What became known

and the B-36 was caught

Anonymous

the increasingly lucrative just

approved the Air

compounded

the Navy's

as "the revolt of the admirals" ensued,

in the interservice crossfire.

charges against the B-36 program became public in April

by the House Armed Services Committee in August and October. The main charge was that the B-36 had serious performance deficiencies. It was said to be obsolete, given that the Soviet Union had just unveiled a new generation of MiG-13 jet fighters. The decision to build the B-36, therefore, was a "billion dollar blunder," as one Navy spokesman argued. The Air Force pointed out that the B-36 had recently demonstrated impressive speed and range capabilities. The jet-assisted B-36D made its first flight in March 1949, and it went on to fly at 435 mph with an altitude of 45,000 feet. In addition, a B-36 had flown a simulated 10,000-mile mission with a 10,000-pound payload, without refueling, in late July. The B-36 had finally met its range requirements, and with the Air Force's recently activated aerial refueling squadrons in place its range was virtually unlimited. The Air Force argued that these capabilities would enable the B-36 to carry out long-range missions even against advanced air de-

and

led to a full-scale investigation

fenses. 101

The Air Force was ultimately vindicated by the congressional investigation. Although some argued that it should nonetheless hold off buying the B-36 and wait for the B-52, the Air Force insisted that it was essential to deploy some long-range bombers as soon as possible. The Berlin airlift, which had dramatically demonstrated the importance of air forces in general, had just ended in June 1949. The Soviet Union tested its first atomic bomb in August, and China fell to the Communists in October, which suggested that the international situation was highly volatile and, if anything, becoming even more dangerous. Strategic arguments notwithstanding, the Air Force had a vested bureaucratic interest in deploying a long-range bomber, and it did not want to run the risk of waiting for the B-52. The B-36 program consequently continued uninterrupted. Production continued until 1954, and 366 B-36S of various models were eventually built.

B -}6 Bomber Program, Hearings before the House Armed Services Committee, 81st Cong., 1st sess., Aug. and Oct. 1949; “Why B-36 Was Made USAF Top Bomber," Aviation Week, August 15, 1949; "New B-36 Performance Revealed at Probe," ioi.

Investigation of the

Aviation Week,

August

22, 1949.

Flying Blind

The B-6o Option

The B-52 program was still not out of the woods in 1949. The B-36's long-range flights drew attention to the B-52's dependence on aerial refueling. General LeMay, whose opposition to aerial refueling went back to his support for the special-mission bomber in 1946, complained emphatically in October 1949 that range was not being emphasized enough in the B-52 program. He stated at a conference in November that he would accept the B-52 if it had an unrefueled radius of 4,250 miles, equivalent to a range requirement of over 11,000 miles. But, at the same time, he would not consider sacrificing any of the bomber's speed, even though these speed capabilities were possible only because the Aircraft and Weapons Board had been willing to settle for an 8,000-mile range. The AMC's director of procurement suggested that another design competition might be in order, and some speculated in late 1949 that the B-52 program would be canceled. With these concerns in mind, Boeing's engineers redesigned the B-52 yet again in December. They increased the size of the aircraft and its fuel load, which gave Model 464-67 an unrefueled radius of 3,500 miles, equivalent to a range of around 9,300 miles.

LeMay

ultimately accepted this version of the aircraft even

though it meant that the B-52 would be dependent on aerial refueling for most missions. 102 The final round of competition in the long-range bomber program began in April 1950, when Convair proposed a swept-wing, all-jet version of the B-36, the B-60. The Air Force decided to build two B-60 prototypes in order to maintain competition and ensure against delays in B-52 production. 103 Although the B-60 was seen as a fallback option, the competition between the two bombers was fairly direct. The first flight of a B-52 prototype took place on April 15, 1952, and the first flight of a B-60 prototype took place just a few days later, on April 18. The B-52 demonstrated considerably better performance and growth potential in the prototype competition that followed. 104 The B-60's top speed was only 520 mph, and its estimated range was only 9,000 miles. Even though the B-60 was relatively inexpensive, the Air Force decided to build only the two prototypes. Given a choice between economy and performance, the Air Force opted for the latter. It was also unwilling to sacrifice performance to spread production contracts evenly among its top contractors. Convair needed a B-60 production contract because the 102.

Based on the account

103.

Report of the

Greene, Development of the B-52, pp. 13-15. Tenth Meeting of the Board of Senior Officers, Dec. 20, 1950, in

RG

341,

HQ/

USAF, NARS. 104.

ARDC,

Air Force Developmental Aircraft, pp. 205-211. See also Chief of Staff, HQ/ to CG, ARDC, Dec. 13, 1951, YB-60 Correspondence: 1952, RG 341,

USAF, Memorandum

HQ/USAF, NARS.

The

First Intercontinental

Bombers

Convair B-6o (U.

B-36 would be phased out

when

S.

Air Force)

B-52 production began; Boeing, on the

other hand, had already been awarded contracts to build over one thou-

sand B-47S.

Assessing the B-52 Development Program

Several points about the B-52 development program should be

em-

was driven by demanding performance requirements. These requirements were set beyond the state of the art at the outset of the program, and they became substantially more challenging as time went by. As a result, Boeing was constantly striving to keep up with the shifting demands for more speed, more range, and more technological phasized.

First,

it

breakthroughs (see Table 6). Second, the B-52 development program was sequential in the late 1940s; developmental prototypes were to be flight tested before a commitment was made to production, and military subsystems were not fully integrated into these prototypes to maximize design flexibility and minimize program costs. Clearly, the design of the B-52 was highly flexible throughout the late 1940s. Boeing worked on a dozen different designs between 1946 and 1949, as summarized in Table 7. The program's development schedule was also flexible, and the costs of the program were modest; a total of $30 million was invested in the program through late 1949. 105 The Air Force's low level of investment in the B-52 105.

Greene, Development

of the B-52, p. 184; Preston, Contract Negotiations, p. 40.

Flying Blind

Table 6

.

B-52 performance requirements

Date

Nov. 1945 Dec. 1946 Dec. 1947

Mar. 1948 Nov. 1949

it

Cruising speed

(miles)

(mph)

(mph)

450 450 420 500 + 500 + 500 +

300 400 420 500 500 500

11,000+

Range requirements were often

mum

allowed

Top speed

12,000 12,000 12,000 8,000 8,000

June 1947

a

Range a

listed in

required radius of action; see note

to threaten

terms of a mini-

2.

Boeing with cancellation on three separate occa-

sions: in late 1947, late 1948,

and

late 1949. Finally, the flexibility of the

program was strongly reinforced by the competition from the 6-36, 6-49, and B-60 competition that forced the contractors to maximize aircraft performance. The competition itself was affordable even though the Air Force's budget was extremely tight throughout this period. 106 Third, the B-52 program was successful in the 1940s. It made major

—

gains in aircraft performance in remarkably

little

time.

On

the whole,

Boeing kept up with the Air Force's escalating performance requirements. And because of the program's sequential nature, Boeing was able to switch designs without having to write off many sunk costs. As a result, the B-52 program seems to have come close to meetings its original cost estimates. It was estimated in July 1947 that the prototype pro-

between $40 million and $50 million. The actual cost of the entire development program, through 1954, was between $60 million and $65 million and the complete development program involved

gram would

many

cost

activities

not part of the simple prototype package.

107

program was subsequently influenced the B-47 program. The initia-

In spite of this efficient start, the B-52

shaped by the same forces

that

Korea and the general deterioration of the international situation led to growing pressures to begin large-scale production. The sequential approach of the late 1940s was subsequently abandoned, and a more concurrent strategy guided the program in the tion of hostilities in

early 1950s.

106. craft,

The B-60 program

cost only $14.4 million; see

ARDC,

Air Force Developmental Air-

pp. 205-211.

Maj. Gen. L. C. Craigie, Chief, Eng. Div., AMC, Letter to CG, AAF, July 11, 1947, XB-52 Airplane, RG 18, AAF, NARS. See also Greene, Development of the B-52, p. 184; Preston, Contract Negotiations, p. 40. 107.

[148]

The

Table

7.

First Intercontinental

XB-52 designs

Model

Date

462 464 464-16 464-17 464-23 464-23 464-27 464-29

Apr. 1946 Oct. 1946 Jan. 1947 Jan. 1947 1947 1947 1947 Sept. 1947 Feb. 1948 July 1948 Oct. 1948 Dec. 1949

46405 464-40 464-49 464-67

Bombers

Estimated

Estimated range

Estimated top speed

cruising speed

(miles)

(mph)

(mph)

Main

8,500

440

410

Original proposal

NA

NA

NA

440 440

420 420

Smaller bomber Special-mission bomber

NA NA NA

NA NA NA

12,000 8,000

455 500

NA NA NA NA

a

12,800 12,400

6,750 8,000 9,300

500

536

NA

572-610 572-610

520 520

feature

General-purpose bomber Led to 464-29 Led to 464-29

Led to 464-29 Met 6/47 requirements Met 12/47 requirements First turbojet model Swept-wing turbojet Forced to meet new range requirements

a

Not

available.

Accelerating the B-52 into Production Although the Air Force started to think about B-52 production in 1949, it took no concrete steps in this direction until after the war in Korea began in June 1950. The onset of hostilities in the Far East led many in the Air Force and the American foreign policy community at large to conclude that a confrontation between the United States and the Soviet Union was imminent. General LeMay argued in September 1950 that "the high probability of war occurring between the United States and the Soviet Union prior to 1955" made the immediate deployment of the B-52 an absolute necessity. A bomber with "substantially higher performance capabilities than the B-36" was needed, he believed, and only the B-52 could be deployed in the near term. LeMay recommended that production of the B-52 and its critical military subsystems be accelerated in

order to have the

1953

first

operational unit ready by the

autumn

of

i° 8

The Board

of Senior Officers

came

to a similar conclusion in

Sep-

tember 1950 after reviewing the intelligence forecast for Soviet offensive and defensive forces. It noted that "Soviet atomic capabilities will be so increased by 1954 as to greatly increase the priority of the counter-atomic mission of the Strategic Air Command" and it expected that, "in the Gen. Curtis E. LeMay, CG, SAC, Letter to Dep. Chief of Staff for Ops., Sept. 12, HQ/USAF, NARS. Also, 1950, Seventh Meeting of the Board of Senior Officers, RG 34U Greene, Development of the B-52, p. 179. 108.

Lt.

Flying Blind

near future, the primary mission of the Strategic Air Command will probably be shifted from a Russian industry concept to a counter-air .

force concept,

which

.

.

will require daylight operations in addition to night

operations." 109 Daylight operations, however, would be challenging because of rapidly improving Soviet air defenses. Intelligence projections indicated that, because of

would

its

speed and altitude limitations, the B-36

become obsolete in 1954. Ironically, SAC was just beginning to deploy the B-36, and indeed it would not become fully operational until 1951. SAC argued that it needed a bomber with a top speed of 573 mph, which happened to coincide with the current performance estimate for the B-52. The board concluded: start to

The B-52 can be made immediate

available before 1954 if procurement is initiated in the future. In this connection, the Board has fully considered the

fact that the

B-52 has not flown and, therefore, data based on actual performance is not available. The Board is of the opinion that it would not be in the national interest to delay initiation of production of the B-52 pending .

completion of service

.

.

tests. 110

The situation in Korea deteriorated in late 1950 when China entered the war and United Nations troops were thrown into retreat. By the end of the year, the board was arguing that the B-52 program should be pushed even harder. It recommended that production be accelerated and that Convair be put to work building B-52S instead of B-36S. Boeing and the AMC estimated that the first production models of the B-52 could be completed by April 1953. 111 With these considerations in mind, the Air Force signed a contract for 20 B-52A production models in February 1951. Since the B-52's flight testing program did not begin until April 1952, this was a distinct shift away from the sequential strategy that had guided the program in the late 1940s. Although one of the concurrency's main assumptions is that the design of the system in question is fundamentally stable, LeMay insisted on a major change in the B-52's design after this production contract was signed. In March 1951 he ordered Boeing to redesign the aircraft to allow the pilot and copilot to sit side by side, instead of one Fourth Partial Report of the Seventh Meeting of the Board of Senior Officers, Sept F 18 ' iqso ' 341, HQ/USAF, NARS. 110. Ibid. See also Qualitative Requirements Forecast for Strategic Operations: 1950-1960 and Summary Minutes of the Seventh Meeting of the Board of Senior Officers, April 18, 1950 through Sept. 18, 1950, Seventh Meeting of the Board of Senior Officers, RG HO/ 109.

^

RG

USAF, NARS. 111.

'

Supplemental Report to the Fourth Partial Report of the Seventh Meeting of the Board of Senior Officers, Dec. 20, 1950, RG 341, HQ/USAF, NARS.

The

First Intercontinental

Bombers

behind the other as they did in the prototypes. He also wanted the forward fuselage extended by 21 inches for additional electronic equipment. 112 Widening and extending the forward fuselage were not minor issues, however, especially in a program that was moving into production. Even after several years of development, the B-52's design was still evolving. Many more design changes were yet to come, but the program

was becoming highly concurrent. By January 1952, $19 million had been committed to B-52 production. The initial production run was fully under way in August when the decision was made to cancel the B-60. The final production specifications for the B-52, however, were not approved until October 1952. Given that the design was still evolving, it was inevitable that these specifications were different from those outlined in the original contract. Naturally, this disrupted the production plans that were already under way.

A

second, larger production contract followed in 1953, but the program continued to undergo major alterations. In fact, four distinct

models (B-52A, B-52B, B-52C, and B-52D) were developed between October 1952, when the production specifications for the aircraft were supposedly set, and August 1954, when Boeing signed a contract to build the B-52D. The design changes that took place during this 22-month period were not insignificant: advanced engines were worked into the design; droppable wing-tip fuel tanks were added, which required modification of the wing structure; and the gross weight of the aircraft increased by 30,000 pounds. Over two hundred engineering changes were made to the airframe alone during this period. 113 The B-52's capabilities naturally improved with each model change, and the B-52D was ultimately built in large numbers. Unfortunately, the 89th aircraft produced was the first B-52D. The 88 earlier, less capable aircraft were ultimately relegated to training and reconnaissance functions. The acceleration of the program in 1951 and 1952, therefore, prevented Boeing from consolidating all design improvements in the development phase of the program, which lasted until 1954. There were other problems associated with the introduction of concurrency into the B-52 program. The advanced engines needed for the aircraft were still in development in the early 1950s, and they were consequently in short supply when the program moved into production. In fact, there were barely enough engines available for the B-52 and

112.

Greene, Development

of the B-52, pp. 179-80.

A. Marschak, The Role of Project Histories in the Study of D, Rand Corporation Paper, P-2850, Jan. 1964, pp. 92-102. See also Inspector General, AMC, B-52 Weapon System Survey, Feb. 6-18, 1955, pp. 1-2, Files, HO/ASD. 113.

R&

Ibid., pp. 127, 156, 180; T.

Flying Blind

B-6o prototypes in 1952, a grand total of four aircraft. The scarcity of engines was subsequently a constraint on the accelerated production

program. 114

bombing and navigation systems were still in development long after airframe production began. The basic problem was that the B-52 flew high and fast, which made it difficult for the bombing system to track and lock onto its target before the aircraft overflew it. A new bombing system had to be developed, but its performance parameters could not be defined until the capabilities of the B-52 itself became clear. These capabilities, however, continued to improve throughout the early 1950s as new models were developed. The development of the bombing and navigation system consequently lagged and indeed was bound to lag behind the development of the aircraft itself. The goal of concurrently developing, producing, and deploying a fully operational aircraft with a fully capable bombing and navigation system was thus inherently difficult, if not impossible. The development of this subsystem could not be accelerated, however fast airframes were being built. Early procurement of production tooling and mass production of airframes could not solve the problems faced by technologically immature subsystems. An advanced bombing and navigation system was ultimately developed, but it was not possible to phase it into the production program until 1956. 115 There were also problems with the defensive armament subsystem, a tail gun. This was another case of subsystem development lagging because of the B-52 's high speed; tracking and targeting became inIn addition, the B-52's

creasingly difficult as speed improved. Concurrency

compounded

the

problem because the system had to be rushed into production. SAC eventually agreed to deploy an interim tail gun until a more advanced system could be retrofitted, but there were delays in making even the interim system available in quantities sufficient for the expanding production program. The performance of the tail gun, moreover, continued to be inadequate. The general assessment of the B-52's defensive armament system, even in the late 1950s, was that it was operationally unsuitable. 116

The B-52 's problems in the 1950s were similar to those experienced by the B-47 program when it was accelerated. The B-47's problems were somewhat more severe because production orders for 1,500 B-47S descended on the program between June and December 1950. Develop114.

Greene, Development

of the B-52, chap. 2.

67-76. Also, Kenneth

Patchin and James N. Eastman, B-52 Deficiencies, Historical Monograph No. 9, HQ, Oklahoma City Air Materiel Area, July 1961, pp. 3-4, 56. 116. Patchin and Eastman, B-52 Deficiencies, p. 8. Also, B-52 Weapon System Survey; Greene, Development of the B-52, pp. 91-115. 115.

Ibid., pp.

L.

The

First Intercontinental

Bombers

merit of B-47 airframes and subsystems

became hopelessly out of sequence, and the retrofitting effort became truly gargantuan. Although the B-52 eased into production more gradually, it also had major problems as the design of the aircraft evolved, the development of critical subsystems lagged, and airframes kept coming off the assembly line. As a result, the B-52 program also involved extensive retrofitting, which proved

to

be extremely

costly. 117

program did not provide SAC with an early operational capability, which was the whole point of the exercise. The Air Force had argued in 1950 that it needed operational B-52S by 1953 or 1954 to offset improving Soviet air defenses. When the decision was made to accelerate the program, it was estimated that the first production model would be ready in April 1953, with an operational capability to follow in 1954. The first flight of the B-52A, however, did not take place until August 1954. By then, the Air Force had decided that the B-52D would go into high-rate production. SAC did not have a wing (or group) fully equipped with B-52S until March 1956, about eighteen months after the Air Force's target date, and this wing was composed of Finally, acceleration of the B-52

B-52BS, not B-52DS. 118 Concurrency, therefore, did not provide

SAC

with an early operational capability.

The B-52 and the Bomber Gap The B-52 program was

and expanded again in the mid-1950s, in the midst of growing concern that the Soviet Union was building more strategic bombers than the United States and creating a "bomber gap." The development of Soviet strategic forces had been viewed with great anxiety ever since 1949, when the Soviet Union detonated its first atomic bomb. It was generally conceded that it was only a matter of time until the Soviet Union developed long-range bombers that could attack the continental United States. The most capable bomber in the Soviet air force, however, was the Tu-4 Bull, which was literally a copy of a B-29 the Soviet Union acquired near the end of World War II. The Bull consequently did not have enough range to pose a serious accelerated

threat to the United States.

The appearance of a new long-range bomber at the Soviet Union's 1954 May Day celebration was therefore viewed as a particularly ominous development. This new bomber, the turbojet Mya-4 Bison, was clearly much more capable than the Bull. But was the solitary aircraft “Analysis of Development, Procurement, and Production of Selected Weapon Systems," no date; Files, Office of Air Force History. 118. SAC/HO, Development of the Strategic Air Command 1946-1986, Sept. 1986, p. 59. 117.

Flying Blind

by the parade ground a developmental prototype or a production model? The ambiguous status of the Bison program made it impossible for American intelligence analysts to make firm predictions about the future of the strategic balance, but it did contribute to a growing sense of anxiety about the magnitude of the Soviet nuclear threat to the United States. 119 that flew

The picture seemed

when Western

May Day celebration, new Soviet bomber, the

to clarify after the 1955

observers spotted yet another Tu-95 Bear. Although the Bear was a turboprop bomber, and therefore slower than the Bison, it was more likely to have intercontinental ca-

Even more ominous, at least twelve and perhaps as many as twenty Bisons were observed at this time. American intelligence analysts revised their estimates of when the Bison program started, where its production program stood, and the kind of bomber force the Soviet Union was capable of building by the end of the decade. The national intelligence estimate approved on May 16 predicted that the Soviet Union would have 400 Bisons and 300 Bears deployed by 1959. These increasingly pessimistic projections were reinforced by the Sopabilities.

Union's Aviation Day show in July 1935. A total of 28 Bisons flew past the reviewing stand, far more than had appeared in May and four times the number of B-52S in SAC. This demonstration turned out to be

viet

however: the same squadron of ten bombers flew past the reviewing stand three times; one aircraft failed to make it around a second time. This spectacle nonetheless led the American intelligence community to revise its projections of Soviet bomber strength upward. At the end of 1933 it estimated that the Soviet Union would have 600800 Bisons by the end of the decade and that it would be capable of producing Bears at a rate of twenty-five per month. Some analysts concluded that the Soviet Union would have a substantial numerical advantage over the United States by the end of the decade, provided that it built as many bombers as it seemed to be capable of producing and that the United States continued on its present course. This led many to argue that the United States faced a potentially dangerous bomber a deception,

g a P-

Some were

of the

more

pessimistic projections about the strategic balance leaked to the press, generating enormous public and congressional

pressure on the Eisenhower administration to expand the B-32 program. The Senate Armed Services Committee began a long series of hearings

My

bomber gap is based on the accounts in Lawrence Freedman, U.S. Intelligence and the Soinet Strategic Threat (Princeton: Princeton University Press, 1986), pp. 65-67; John Prados, The Soviet Estimate (Princeton: Princeton University Press, 1986), pp. 38-50; Colin S. Gray, "Gap Predictions and America's Defense: Arms Race Behavior in the Eisenhower Years," Orbis 16 (Spring 1972), 257-274. 119.

discussion of the

The

First Intercontinental

Bombers

on the state of the U.S. strategic bomber program in April 1956, during which the administration was repeatedly criticized for not devoting enough resources to bomber production. Ironically, the Eisenhower administration had actually favored the strategic bomber as the country's most cost-effective deterrent; it provided "more bang for the buck." Even so, the administration was reluctant to dramatically increase spending on the B-52 program; it believed that there were limits to what the nation could afford and how many bombers SAC really needed. 120 The B-52 program, moreover, had already been accelerated and expanded by the time the Senate hearings began. The Air Force began to think about accelerating B-52 production from twelve to seventeen per month in early 1955. A decision to do so was made in May 1955 after the sighting of the first Bison in May 1954 but before the revised intelligence

estimates about Bison production were issued in 1955. The B-52 program was accelerated again in April 1956, increasing the production rate

per month. The administration also decided in April to expand the overall production program from 400 to 500 aircraft, a total to

twenty

aircraft

subsequently raised

Most

to

600

aircraft. 121

development problems had been solved by the time the program was reaccelerated in 1955-56. The B-52 would have been moving into production at this point in any event. The main problems the second acceleration had to contend with were relatively routine production bottlenecks and shortages in materials and skilled manpower. There was, then, a great deal of difference between the second acceleration of the program and the first. In the mid-1950s, the Air Force accelerated a production program that was already under way. In the early 1950s, it accelerated an immature development program into production, which proved to be highly problematic. The lesson to be learned is not that acceleration is always foolhardy and counterproductive; rather, it is that acceleration of a development program into production the establishment of a concurrent development and production program is risky if major developmental uncertainties still exist. The bomber gap scare ended soon after the B-52 program was accelerated in early 1956. The United States began to conduct reconnaissance flights over the Soviet Union with U-2 aircraft in June 1956. These flights provided the first hard evidence about the magnitude of the Soviet strategic bomber program and it quickly became apparent that the Soviet

—

of the B-52 's nagging

—

See Glenn H. Snyder, "The 'New Look' of 1953," in Schilling, Hammond, and Snyder, Strategy, Politics, and Defense Budgets, pp. 379-524; Huntington, Common Defense, pp. 65-113; Jerome H. Kahan, Security in the Nuclear Age (Washington: The Brookings 120.

institution, 1975), pp. 9-73. 121. Study of Airpower, Hearings before the Senate Cong., 2d sess., April-June 1956, pp. 435-440, 1588.

[155]

Armed

Services Committee, 84th

Flying Blind

Union was not producing Bisons and Bears at the rates earlier ascribed to them. Intelligence estimates of Soviet bomber production were consequently revised downward in the spring of 1957 and regularly thereafter. According to one authoritative study, by late 1959 the projection for mid-1961 was only 19 percent of that estimated in 1956. By early 1961, the Soviet Union had built only 190 long-range bombers, the ma122 jority of which were Bears rather than Bisons. The bomber gap failed to materialize as predicted for several reasons. First, the Soviet Union apparently encountered technical problems with bombers, particularly with the high-speed Bison. Second, the Soviet Union did not devote all its aerospace resources to Bison and Bear production, as some assumed they would. It eventually became clear that its

Union was also putting substantial effort into ICBM development, which led to speculation in the late 1950s about the emergence of a "missile gap." Third, the most pessimistic predictions about Soviet bomber deployments came from the Air Force intelligence office, which had a vested organizational interest in inflating the Soviet threat. A more benign explanation is that the Air Force's strong organizational identification with strategic bombers influenced its intelligence assessments; it assumed that any rational group of leaders would want to build large numbers of bombers. Fourth, a great deal of ambiguity surrounded the Soviet bomber program prior to 1956. Worst-case assumptions about it, therefore, were inevitably pretty grim. Finally, the bomber gap failed to materialize because the United States simply outproduced the Soviet Union. By early 1957 it was clear that the United States was building more long-range bombers than its rival. The rate of B-52 production was consequently cut back to fifteen aircraft per month in April 123 However, the size of the B-52 program was 1 957 not cut back to pre-1954 levels. In fact, the program actually expanded in the late 1950s, and 744 B-52S were ultimately built. There were three main reasons for the Soviet

-

Anxiety over the strategic balance remained high throughout the late 1950s, and the B-52 program was the country's only near-term deployment option at the time. The B-70 bomber, which the Air Force hoped would succeed the B-52, encountered several development problems. Extended B-52 production provided what was supposed to be an interim capability until the B-70 could be deployed. 124 Finally, the performance of the B-52 improved substantially with the development of the B-52G and B-52H models in the late 1950s and early 1960s. The B-52 this.

Freedman, U.S. Intelligence, p. 67. Gray, “Gap Predictions," p. 262. 123. 124. In addition, the B-36 was being phased out of the strategic inventory in the late 1950s. The last B-36 was retired from service in 1959. For more details on the B-70 program, see Chapter 6. 122.

(156]

The

First Intercontinental

Bombers

ultimately had a top speed of 660

mph, and

12,500-mile mission without aerial refueling. made the B-52 an attractive strategic system.

B-52H flew a simulated These kinds of capabilities a

The last B-52 was delivered to the Air Force on October 26, 1962, in the midst of the Cuban missile crisis. At the time, the Soviet Union had built

around 190 Bisons and Bears, compared to over 740 B-52S and nearly 2,000 B-47S on the American side. The United States, moreover, had extensive aerial refueling capabilities and forward basing options. A sizable bomber gap existed in the late 1950s and early 1960s, but it did not favor the Soviet Union.

Conclusions

The AAC and AAF were clearly responsible for initiating the bomber development programs that began in 1941 and 1944. The intelligence reports that triggered these efforts could not, however, provide explicit

guidance about performance requirements.

was not

When

these programs be-

what operational capabilities would be needed many years in the future. This was certainly true as far as speed was concerned and it was even true with respect to range. Both the B-36 and B-52 were ultimately required to perform missions very different from those for which they were originally conceived. In each case, the AAC and AAF set performance requirements far beyond the state of the art, ignoring warnings from the aircraft industry and their own technical experts in the process. As a result, these programs involved major technological advances. The B-35 was the most revolutionary of the lot, because it featured a new and radically different system design. The B-36 and B-52 involved several major advances at the subsystem level. The B-35, therefore, belongs in the first position on gan,

it

at all clear

the nine-point scale of technological ambitiousness outlined in Chapter

The B-36 and B-52 belong in category four. The B-36 and B-52 programs ultimately met their demanding requirements, but not until technological developments that were totally unforeseen at the outset of the programs materialized. It was not until 1

(see Table

1).

1949, when the turbojet-assisted B-36D began its flight test program, that the B-36 met its 1941 speed requirements. The B-52 was turned into a turbojet,

swept-wing bomber

in 1948,

which

clearly

was not

antici-

pated when the AAF established its requirements for a high-speed, long-range bomber in 1944 and 1945. The Air Force placed a premium on performance, and at no time was it willing to compromise operational capabilities to save time or money. The deliberations over the B-36 program were tortuous in the late 1940s 1157]

Flying Blind

precisely because the Air Force

was not anxious

to build a

bomber with

mediocre performance, even though it could be deployed relatively quickly. The Air Force avoided choosing between performance and availability by repeatedly pressuring Convair to improve the B-36's speed. The Air Force was also reluctant to choose between different performance capabilities. Decision makers are always reluctant to make value trade-offs, but strategic

problem especially

bombardment

difficult for

doctrine

made

the value trade-off

the Air Force because

range, speed, altitude, and defensive

armament

it

specified that

capabilities

were

important. The Air Force finessed this trade-off problem by focusing

all

first

on one requirement and then another, ratcheting them up over time. This took place repeatedly in the B-52 program, where a couple of tactical compromises were made along the way, but in the end the bomber's range and speed requirements were both set at very high levels. The process of ratcheting performance requirements up over time does not seem to have been part of a grand design; rather, the Air Force focused on the B-32 's most prominent performance deficiency at any given time. The B-35, B-36, and B-52 programs were all guided by a sequential procurement strategy at first, but each program was eventually accelerated into production. The B-35 program became highly concurrent in that it relied heavily on paper studies for design and production decisions; the aircraft design was assumed to be frozen when production efforts began; an elaborate production schedule was mapped out even though development activities were still under way; substantial financial commitments were made to the program; competition was eschewed; and development and production overlapped to a great degree. The B-36 program was concurrent in all these respects as well, except that its production schedule was less clearly defined than the B-35's; it compensated for this by featuring a great deal of subsystem integration, mainly in terms of defensive armament. The B-35 and B-36 programs therefore exhibited six of the attributes of concurrency outlined in Chapter 1 (see Table 2). The B-52 program started out as a model of sequential procurement. But when it was accelerated in the early 1950s, the design of the system was assumed to be relatively stable; its production schedule was mapped out in great detail, even though development activities were still taking place; subsystem integration was pursued aggressively; a substantial financial commitment was made; and development and production began to overlap. The B-52 program therefore exhibited none of the attributes of concurrency in the late 1940s and five in the early 1950s. Wartime budgets provided the resources for these production programs, and wartime needs provided a strategic rationale for doing what the AAF and the Air Force were strongly inclined to do anyway: build and deploy long-range bombers. But the feasibility of accelerating these

The

programs

First Intercontinental

into production

Bombers

was not

carefully assessed.

The

AAF

and the

Air Force did not have a clear sense of the technological unknowns that permeated these programs when they decided to employ more concur-

rency in them. As a result, the production phases of these programs were plagued by a wide variety of serious cost, schedule, and technical problems.

The outcomes

and B-49 programs are clear even though large-scale production programs were never completed. The introduction of concurrency into the B-35 program certainly did not accomplish what it was intended to do: provide large numbers of bombers for the

AAF

of the B-35

during the war. In

fact,

concurrent production

activities interfered

with the extensive development work the flying wing needed. The B-35's costs consequently skyrocketed and its flight testing program was repeatedly delayed. The B-35 prototype program was supposed to cost $4.45 million, but by 1944 it was estimated that it would cost $25 million. By 1948 the total investment in the flying wing program had reached $67 million, one of the main reasons the Air Force decided to build the RB-49. The schedule of the B-35 program slipped repeatedly over the years, but most horribly during 1943 and 1944, when the concurrent production program was in place. The B-35 flight test program did not begin until 1946, approximately two and one-half years after it was scheduled to begin. Finally, the performance of the flying wing never

measured up to the expectations the AAF and the Air Force had for it. The decision to begin producing the RB-49 before its flight test program was completed proved to be a mistake, as the turbojet version of the flying wing had serious stability problems that prevented it from being a useful military system. The enormous technical problems the flying wing faced were not resolved in a timely manner. The diversion of engineering resources away from these development problems, which ultimately determined whether the flying wing succeeded or failed, helped to undermine the program as a whole. The B-36 program also met with serious difficulty. The cost of the prototype program was originally estimated at $15.8 million, but it ultimately cost $39 million. The first hundred production units were supposed to cost $2.6 million each, but they actually cost $6.25 million apiece. 125 The program's development schedule slipped repeatedly, and the first prototype flight took place twenty-seven months behind schedule. Although the point of accelerating the program had been to deploy the B-36 before the end of the war, it did not achieve an initial operating capability with SAC until 1951. The Air Force itself acknowledged in 1948 that beginning B-36 production "prior to completion of engineering de-

125.

Weigley, American Wax/ of War,

p.

372.

Flying Blind

velopment on the experimental models

.

.

resulted in extensive

.

126 costly modifications to the production airplanes."

and

A

1949 review of instrumental in increasing

program concluded that concurrency was costs far beyond those estimated by the contractor and also in forcing readjustments of the delivery schedule. 127 The modifications eventually required in the B-36 were so extensive that SAC was caught up in the the

retrofitting effort, to

its

considerable displeasure.

128

The two main phases of the B-52 program had distinctly different outcomes. The development phase was sequential and very successful. The cost of the prototype program was originally estimated at $40-50 million, and it ultimately cost less than $60 million. Although the program did have a cost overrun, it was not nearly as severe as those incurred by the B-35 and B-36 programs, for which the actual costs of development were two and one-half to five times the estimated costs. In addition, the B-52 program moved expeditiously into flight testing. The first prototype flight took place roughly three and one-half years after the design of the aircraft was set, whereas the "accelerated" B-35 and B-36 programs took nearly five years to get to the same point. The B-52 also went on to demonstrate its critical performance capabilities quickly. The B-52 's sequential strategy was critical to the program's overall success in all these respects. The cost of the development program was kept to a minimum, which made it relatively easy for Boeing to redesign the aircraft. This design flexibility was critical to the ultimate success of the program. The competition with the B-36 gave Boeing an added incentive to maximize aircraft performance, affordability, and availability. The introduction of concurrency failed to accomplish what it was intended to accomplish in the second phase of the B-52 program. The acceleration did not provide the early operational capability

the

first

B-52 units were activated eighteen

Concurrency was actually counterproductive to the

production of 88

it

promised;

months behind schedule. in several respects.

It

led

aircraft before the final configuration of the

and it compounded the already formidable problems associated with subsystem development. Other problems, such as the

bomber was

set,

extensive retrofitting eventually required to tionally useful,

can also be traced

make

to the injection of

the aircraft opera-

concurrency into the

program. 126.

Col. Frederick A. Bacher,

p. 13, Files,

127.

Brig.

Jr.,

Memorandum

to

Chief of

Staff,

USAF,

July 13, 1948,

HO/AFLC. Gen. C.

P.

Kane, Chief, Procurement

Div., Office of the Inspector General,

Report to the Inspector General, USAF, July 25, 1949, 128. SAC/HO, The B-36, pp. 16-17, 60.

p. 2, Files,

HO/AFLC.

The Push

to

Develop

Supersonic Capabilities: The B-58

The B-58 supersonic bomber program was similar to the B-47 and B-52 programs in some respects but radically different in others. One similarity was that it was initially triggered by strategic developments, in this case intelligence reports

about German progress

in building jet

Another was that the AAF was aggressive in pushing the state of the art: it began to explore the idea of building a supersonic bomber more than three years before the sound barrier itself was broken. When the Air Force forged ahead and issued formal requirements for a supersonic bomber in 1951, its system specifications were extremely demanding. At this point, the B-58 program took a different course from the programs of the late 1940s, because the Air Force structured it around a highly concurrent procurement strategy. Prototypes proposed by the main contractors in the program, Convair and Boeing, were not built, flown, and compared. Instead, source selection was based entirely on paper studies. When the Air Force awarded the B-58 contract to Convair in 1952, it decided that developmental prototypes would not be built and that flight testing would not take place prior to production. It assumed that major design changes would not be needed and that production aircraft could be used for flight testing. In addition, the Air Force insisted that military subsystems be fully integrated into the program at an early stage. The B-58 program was a classic example of concurrency at work. fighters in 1944.

As

turned out, the B-58 was rushed into production before its engineering development was complete and, indeed, before some basic it

[161]

Flying Blind

questions about supersonic flight had been answered. quently experienced serious problems. scientific

It

conse-

Origins of the Program

program can be traced back to early 1944, when intelligence reports indicated that Germany was making progress in building fighters capable of flying 650 mph. This information led the AAF to begin thinking about building a strategic bomber that would be even faster than the 8-47, a bomber with supersonic capabilities. 1 The AAF included a design study for a supersonic bomber in its preliminary postwar research and development program, which it outlined in the

The

origins of the B-58

summer

of 1944. 2

That the AAF was even thinking about supersonic bombers in 1944 is simply remarkable. This was a time when the most advanced bomber in the world was the B-29, a straight-wing, propeller-powered aircraft with a time when a top speed of 364 mph, roughly half the speed of sound even the subsonic speed requirements for the B-47 were beyond the state of the art (see Chapter 3). The B-47 faced technological problems that, at the time, seemed to be insurmountable. It was not until late 1945 that Boeing discovered the breakthrough that ultimately enabled the B-47 to fly at high subsonic speeds. The breakthroughs needed to make supersonic flight possible were even more distant. The experimental Bell X-i program was just getting under way in 1944, and the X-i did not break the sound barrier until October 14, 1947. There were some doubts up until that time about the feasibility of breaking the sound barrier. It

—

seems

clear that the

AAF was

nological breakthrough

bomber

sonic

when

it

in 1944,

when

nor was

it it

not trying to exploit

began

some

recent tech-

about building a superanticipating an imminent breakthrough to think

did so.

The AAF's

interest in a supersonic

bomber

intensified

when

it

began

study the advanced German aircraft captured at the end of the war. The Me-i63B, for example, was put through an extensive flight testing program in the United States beginning in October 1945. The AAF knew to

that the Soviet

Union had come

By

late

same German

development and production facili1945, reports indicated that the Soviet Union was building

aircraft as well as ties.

German

into possession of the

aircraft

of sound, or Mach 1, is around 740 mph at sea level. Wright Field, "Scope and Procedure Plans: Project B-7, Post-War Res. and 2. Eng. Div., Devel. Prog., Five Year Period, FY 1946 to FY 1950," Memorandum, Aug. 28, 1944; cited in Alfred Goldberg, "The Quest for Better Weapons," in Wesley F. Craven and James L. Cate, eds.. The Army Air Forces in World War II, 7 vols. (Chicago: University of Chicago Press, 1.

The speed

1948-1955), vol. 6, p. 243.

The Push

the Me-163

to

Develop Supersonic Capabilities

Me-262 and studying the Me-263, an advanced version of the Me-163. 3 The AAF consequently stepped up its supersonic bomber development effort, initiating two sets of exploratory design studies, one conducted by Convair and the other by Boeing. ar, d

Design Studies

Convair's project study.

Its first

1949.

GEBO

wing

types,

stage,

was known

GEBO

I,

bomber (GEBO) October 1946 and continued into

as the generalized

began

in

a highly exploratory study of several aircraft designs,

I,

and engine combinations, analyzed the effects of various design parameters such as wing shape and degree of sweep on aircraft speed and range. No one yet knew how even these most basic design elements affected transonic and supersonic performance. 4 The second stage of the study, GEBO II, began in June 1949 after the Air Force told Convair that a ''new and possibly unconventional approach to the intercontinental bomber problem" would be needed to generate the supersonic capabilities the B-47 and B-52 lacked. 5 The Air Force's stated goal was to develop a medium-range bomber with a 1,380-2,875-mile radius of action, a 520-mph cruising speed, and the best possible top

—

—

speed.

Because supersonic aircraft would burn fuel quickly, the trade-off between speed and range was especially pronounced. The Air Force commented on this dilemma in one review of the GEBO II project: "It has long been recognized that high-speed and long-range performance are incompatible, and that considerable design problems arise when either great range or high speed is required. In this project, however, we want both simultaneously." 6 In April 1950, the Air Force reoriented the

GEBO

study toward an especially unconventional approach to the speed-range problem. It decided that a "parasite" bomber might be able to fill the bill. The idea behind parasiting was that a small, speedy bomber would be carried most of the way to the target by a larger aircraft, such as the B-36. Once released, the parasite bomber would dash to the target at high speed and complete the attack. At that point, depending on the design of the bomber, it would either rendezvous with the carrier and reattach itself, 3.

ASD, 4.

Mach 5.

Files,

Richard D. Thomas, History of

the

II

Development of the 8-5S Bomber, Historical Div.,

1965, pp. 16-17. Ibid., p. 33.

0.8 to Brig.

Mach

Unusual aerodynamic conditions

roughly

1.2.

Gen. D.

HO/ASD;

exist in the transonic regime,

L. Putt, Dir. of Res.

and Devel.,

Letter to

CG, AMC,

cited hereafter as Putt Letter. See also Report,

April

1,

1949,

"Outline on Status of

Supersonic Bomber Reconn. Studies," Sept. 2-6, 1950, Files, HO/ASD; cited hereafter as Supersonic Bomber Outline. 6. "Review of Project MX-1626," Feb. 16, 1951, Files, HO/ASD.

Flying Blind

was expected that the parasite bomber would carry one or more pods that would be dropped off over the course of its mission. These pods would carry the bomber's payload, extra engines for high-speed flight, extra fuel tanks, and other equipment not needed for the return flight. The parasite bomber, therefore, would have the horsepower it needed for supersonic flight into a heavily defended target zone, and it would also be able to shed excess weight once range became more important than speed. The Air Force raised the GEBO II design objectives when it reoriented the study around parasiting. The overall parasite system was to have a 4,000-5,000-mile radius of action, and the parasite bomber itself was to have a top speed of Mach 1.3-1. 5 in a target zone 575-2,300 miles deep. The parasite bomber's cruising speed was to be Mach 0.9, and top speeds of up to Mach 2.0 were to be sought. 7 The Air Force could afford to focus GEBO II on parasiting because Boeing was conducting design studies along much more conventional lines. Boeing's work in this area began in October 1947, when it won a development contract for the high-speed, medium-range XB-55 bomber. The XB-55 was, however, canceled in January 1949, when the Air Force was forced to trim its accounts in light of the fiscal year 1950 budget ceilings. Even so, Boeing was able to continue its work on high-speed bombers. At first it was authorized to complete the XB-55 wind tunnel tests and design studies that were already funded and under way. Later, its work was funded as the MX-1022 project. The MX-1022 bomber was supposed to have a 2,300-mile radius of action, 230 miles of which could be flown at Mach 1.3-1. 5. The Air Force specified that Boeing was to or cruise back to

its

base at a moderate speed.

It

follow conventional development lines as far as possible. 8 In late 1950, Boeing submitted its ideas for a new bomber the most promising being

—

Model 484-405B, a fairly conventional aircraft with a 47 degree swept wing and a conventional bomb bay. Boeing estimated that this aircraft would have a 2,200-mile radius of action, 200 miles of which could be flown at Mach 1.3. Its average cruising speed would be Mach 0.9. It would have a 2,600-mile radius if the entire mission was flown at subsonic speeds. 9

72

7.

Supersonic Bomber Outline; Thomas, History

8.

Supersonic Bomber Outline. Also, Col. Carl

of the

Development of the 6-58, pp. 65-

-

Div.,

AMC,

Letter to Dir. of Res.

F. Damberg, Chief, Ops. Off., Eng. and Devel., HQ/USAF, March 6, 1951, Files, HO/ASD;

Damberg Letter. Supersonic Bomber Outline.

cited hereafter as

Also, Lt. Col. E. N. Ljunggren, Chief, Bomb. Br., Aire, 9. and Guided Missiles Sect., AMC, Memorandum to Col. R. L. Johnston, Chief, Aire, and Guided Missiles Sect., Eng. Div., AMC, Dec. 6, 1950, cited hereafter as Ljunggren Memorandum; and Maj. Gen. Fred R. Dent, Jr., "Development of Strategic and Tactical Bomber Development Program," Letter to CG, ARDC, cited hereafter as Dent Letter; both in Files,

HO/ASD.

The Push

to

Develop Supersonic Capabilities

By way of comparison, Convair predicted in January 1951 that a parasite system would have a substantially greater radius of action than a conventional bomber and that a parasite bomber would be able to perform a greater portion of its mission at supersonic speeds. It estimated, for example, that a parasite system would have a 4,400-mile radius of

bomber flew 575 miles at Mach 1.3. system would have a 4,150-mile radius of action

action

if

the

It

estimated that the

if

1,300 miles of the

mission were flown at supersonic speed. Convair claimed, moreover, that top speeds of Mach 1.6-1. 7 were possible in a parasite bomber. It also estimated that a 4,600-mile radius was possible if the entire mission was flown at subsonic speed. 10

Aerodynamic Unknowns

The Boeing and Convair design studies were conducted in the shadow of some truly monumental unknowns about the aerodynamics of transonic and supersonic flight. Several basic scientific questions remained unanswered in the late 1940s and even in the 1950s. There was, for example, no fully developed mathematical theory in 1947 of the mixed airflow conditions that existed in transonic flight regimes. The equations that described these conditions were complex, and indeed they were not all solved until some time thereafter. For many years, therefore, aircraft designers did not have even a mathematical model to guide them through this particularly dangerous flight regime. NACA, moreover, did not focus its research program on this particular issue There was, quite simply, a lack of basic scientific knowledge about this flight regime long after the supersonic bomber development effort had begun. Even if the mathematics of transonic flight had been better understood in the late 1940s, it would have been difficult to translate theory into practice due to a lack of adequate wind tunnel facilities in the United States. Wind tunnel testing at high subsonic, transonic, and supersonic speeds was needed to check the validity of performance estimates and the aerodynamic stability of the various designs, but it until the late 1940s.

was

wind tunnel data for all these aerodynamic regimes, particularly in the transonic and supersonic regimes. Boeing's wind tunnel facility was limited to high subsonic and low transonic testing, and its supersonic data were based on wing and body combinations not directly applicable to bomber configurations. There was, therefore, a certain amount of guesswork in10.

generally difficult and frequently impossible to obtain

R. O.

Specifications for

HO/ASD.

R. L. Gabbe, Consolidated Vultee Aire. Corp., "Preliminary Model Long-Range Supersonic Bombardment Airplane," Jan. 19, 1951, Files,

Furlow and

Flying Blind

proposed aircraft. One Air Force review of Boeing's MX-1022 project noted that detailed wind tunnel data were "conspicuous by their absence." 11 The Air Force began tests in the AMC's high-speed wind tunnel in September 1950. George Schairer, Boeing's chief aerodynamicist, considered these tests to be extremely important because they focused on "the basic problem of deciding what types of airplane wings are optimum or even usable for supersonic bombers." 12 There was, in other words, no definitive answer even in 1950 to one basic question: what types of wings work best on supersonic bombers? Although several years of design studies had taken place by that time, there were real limits to their usefulness and validity in the absence of basic scientific knowledge about the aerodynamics of transonic and supersonic flight. These problems were compounded by the fact that the contractors had begun to feature delta wings-highly swept triangular wings with trailing edges roughly perpendicular to the fuselage in their designs. Delta wings were effective in overcoming the drag caused by compressibility at high subsonic speeds and also in avoiding the aerodynamic shock waves generated in supersonic flight. In addition, they were stronger and had more internal fuel capacity than conventional volved in

its

projections about the supersonic capabilities of

its

—

wings.

There were, however, several problems with delta wings, the most important of which was that American aircraft designers knew little about them. Although delta wings had been studied in the 1930s and early 1940s in Germany, little was known about their basic aerodynamic characteristics in the United States even in the late 1940s. The AAF did have the opportunity to conduct flight tests of German delta-wing aircraft, such as the Me-i63B, which were captured at the end of the war. Even so, no American delta-wing aircraft flew until the Convair XF-92A flight test program got under way in February 1949. Even then, the XF-92A lacked the power to fly at supersonic speeds, so it could not be used to investigate the supersonic characteristics of delta-wing aircraft. 13 This did not deter Convair and Boeing from gravitating toward delta-wing designs in their supersonic bomber studies. Delta-wing aircraft had a number of stability and control problems as well, many of which were not sorted out until long after extensive flight testing. One basic problem was described by the German aerodynamicist, Alexander Lippisch: "Our tests with aircraft with different u. Supersonic Bomber Outline, See

also

Thomas,

History of the Development of the B-58,

p. 86.

Thomas, History of the Development of the B-^ 8 p. 82. The XF-92A began a limited test program in September 1948 but did not begin 13. fledged flight test program until February 1949; see ibid., pp. 2-19, 27-43. 12.

Schairer quoted in

,

[166]

a full-

The Push

to

Develop Supersonic Capabilities

swept-wing configurations had shown

swept wing with angles larger than 30 degrees sweep showed severe wing tip stall in the low speed range and was therefore quite difficult to handle at take-off and landing. But in order to penetrate into the transonic speeds and proceed into supersonic flight

it

that a

was necessary to use a higher swept-back angle was inherently difficult to design a delta-wing

than 30 degrees." 14 It aircraft that would be efficient at supersonic speeds, which were said to be necessary for operational reasons, as well as stable at low speeds, which were necessary for taking off and landing. The optimal wing for supersonic flight was short and highly swept, like the delta; the optimal wing for taking off and landing was long and straight. This dilemma led some NACA scientists to conclude in 1949 that a variable-sweep wing might be best for supersonic aircraft. Such a wing could be swept out for taking off and landing and swept back for supersonic flight. One of NACA's reasons for supporting research on variable-sweep wings was, according to one NACA scientist, "we didn't have enough nerve to build a 60 degree airplane." 15 Both Convair and Boeing were reluctant to use the variable-sweep wing, ironically, on the grounds that it was too experimental. 16 Convair

was able to sidestep one design problem because its delta-wing parasite bomber would rely on a more conventional carrier for take-offs and landings. Boeing dealt with this problem by sweeping the wings of

its

proposed bomber 47 degrees, which enhanced its stability during takeoffs and landings but impinged on its high-speed performance. In general, the delta wing did not function well in low-speed flight regimes. A highly swept delta-wing aircraft would have to take off and land at very high speeds to avoid stability and control problems. This was intrinsically dangerous. Other problems with the delta wing emerged as the XF-92A flight test program proceeded. The delta had inherent handling limitations, even at high speeds; it was easy to roll and it had a high rate of roll due to the wing's poor damping characteristics. The AMC recognized that building a supersonic delta-wing

bomber would be

especially problematic. 17

The uncertainties that permeated the supersonic bomber program therefore, were formidable throughout the late 1940s and early 1950s. Basic scientific questions about transonic and supersonic flight remained unanswered. Wind tunnel testing, the first step in the design process, Quoted Quoted

14.

in ibid., p. 8.

Robert L. Perry, Innovation and Military Requirements A Comparative Study, Rand Corporation Research Memorandum, RM-5182-PR, August 1967, p. 47; see also pp. 15.

-

34 73 16.

in

-

Flight tests of prototypes

and 1952. Thomas, History

equipped with variable-sweep wings did not take place

until 1951 17.

of the

Development of the B-58, pp. 24-27, 74.

Flying Blind

even more, lhe sound barrier, after all, was broken only in late 1947. Research on deltawing aircraft was in an exploratory stage in the United States throughout this period, and several critical aerodynamic problems that confronted delta-wing aircraft remained to be resolved. The Air Force believed that the Convair and Boeing design studies would resolve these technological uncertainties and provide "realistic lagged due to inadequate

facilities.

Flight testing lagged

up-to-date information regarding the state of the

art."

18

It

was not

at all

these studies were supposed to do this, given that they were only paper studies supported by preliminary, uneven, and incomplete wind tunnel testing. None of the studies, moreover, involved pro-

clear

how

totypes or flight testing. They consequently generated little hard data themselves. In any event, simple design studies could not resolve underlying scientific uncertainties about the aerodynamics of transonic and

supersonic flight, which had to be addressed through basic research. Convair's and Boeing's design studies nonetheless gave the Air Force the impression that the basic questions about the aerodynamics of highspeed flight were being resolved and that the state of the art was being thoroughly assessed. Convair studied more than 100,000 aircraft configurations over the course of

its

GEBO

I

and

GEBO

II

studies;

perhaps

this

suggested that a comprehensive investigation of the supersonic bomber development problem had been undertaken. Both Convair's and Boeing's studies, however, made assumptions that were suspect until fundamental scientific issues were better understood. Convair's studies were also highly dependent on assumptions about parasiting. Although GEBO II indicated that a supersonic bomber with adequate range could be built, this was so only if parasite systems were used. Boeing's studies indicated that, in the absence of parasiting, range and speed capabilities

would

suffer appreciably.

Performance Requirements The Air Force began defining formal performance requirements for a supersonic bomber in late 1950, and it also intensified Convair's and Boeing's work on specific designs. It was not a coincidence that this surge in activity came shortly after the beginning of the Korean War. The likelihood of a general war between the United States and the Soviet Union seemed to be rising, and the effectiveness of advanced, Sovietbuilt fighters against the American bombers in Korea was particularly worrisome. In addition, the budgetary resources 18.

Ljunggren Memorandum.

for large-scale

develop-

The Push

to

Develop Supersonic Capabilities

became available, a complete turnaround from the situation in 1949. The one thing that had not changed between 1949 and 1950 was the fact that major scientific and technological unknowns merit projects suddenly

surrounded the program. The Air Force awarded expanded development contracts to Convair and Boeing in early 1951. It decided that new contractors should not be brought into what was becoming a head-to-head competition because it would take newcomers six to nine months to conduct preliminary design studies and catch up to Convair and Boeing. This would only slow things down. Convair's new MX-1626 study featured a parasite bomber with highly swept delta wings and a droppable pod that contained one of the aircraft's three engines. Convair estimated that this bomber would have a top speed of Mach 1. 8-2.0. Boeing's new MX-1712 study focused on a more conservative design along the lines of Model 484-405B. Although the Air Force acknowledged that many unknowns complicated the development effort, it estimated that the first flight of the supersonic bomber would take place in June or July of 1954. The Air Force hoped, moreover, that the development effort could be accelerated. The AMC asked NACA to focus on research problems relating to the supersonic bomber so that the engineering phase of the program could be accelerstill

ated. 19

Formal general operational requirements were issued in December, 1951. These assumed that Soviet air defenses would become increasingly formidable and efficient in the 1955-1960 period because they would include supersonic all-weather interceptors, guided surface-to-air missiles (SAMs), ground-based barrage rockets, and heavy antiaircraft

guns capable of delivering increasingly effective fire against targets at high speeds and altitudes. 20 The Air Force concluded that the next generation of strategic bombers would have to fly faster and higher than its predecessors, even though several Rand Corporation studies indicated that speed might be less important than maneuverability in the late 1950s. High-speed aircraft were not particularly maneuverable, though, so the Air Force would have to sacrifice speed if it wanted more maneuSee “Review of Project MX-1626/' Feb. 16, 1951; W. E. Lamar, Proj. Eng., Wright Air Development Center (WADC), “Project Status: MX-i 626," Aug. 27, 1951, Nov. 15, 1951; Capt. J. S. McCollom, Proj. Eng., WADC, “Project Status: MX-1626/' Feb. 15, 1952; Convair Report, "MX-1626 System," Feb. 4, 1952; Maj. J. D. Seaberg, Jr., Proj. Eng., WADC, “Project Status: MX-1712," Dec. 17, 1951; all in Files, HO/ASD. See also Damberg Letter; 19.

Dent

Letter.

Maj. Gen. M. R. Nelson, Dir. of Reqs., HQ/USAF, "Gen. Op. Req. for Strategic Bombardment System," Dec. 8, 1951, Files, HO/ASD; cited hereafter as General Operational Requirement. It was estimated in 1950 that Soviet air defenses would feature SAMs by mid-1958; see “Presentation for Senior Officers Board: Soviet Air Defenses, 1950 to 1958," Aug. 22, 1950, Report of the Seventh Meeting of the Board of Senior Officers, RG 20.

341,

HQ/USAF, NARS. [169]

Flying Blind

verability.

It

would

take

more than

a couple of studies

consultants to get the Air Force to shift

its

from

civilian

thinking on a basic issue such

High speed also adversely affected bombing accuracy, but the Air Force assumed that accuracy problems would ultimately be solved by new bombing systems of one kind or another. 21 The Air Force consequently stated that the new bomber had to fly supersonically at 50,000 feet. It also specified an unrefueled radius of 2,650 miles and, with one outbound refueling, a 4,600-mile radius. Since Soviet air defense capabilities were expected to be increasingly effective against subsonic bombers as the decade wore on, the supersonic bomber was to be ready for deployment in early 1957. The Air Force's goal was to field a bomber capable of delivering a 10,000-pound payload with great accuracy even in the face of advanced Soviet air defenses. 22 Parasiting was quietly dropped from the program at this time, for three main reasons. First, aerial refueling could be used to extend bomber range, although there is no indication that the Air Force made a careful analysis of the relative effectiveness of parasiting and aerial refueling as range-extension measures. Second, parasiting would be exorbitantly expensive, because each parasite bomber would need its own carrier. Two bombers could be serviced by a single tanker, however, and a fleet of tankers was already being built for the B-36, B-47, and B-52. Finally, parasiting missions would be complicated and hazardous. The carriers would be highly vulnerable to attack when they were carrying their bombers, since their speed, altitude, and maneuverability would not protect them against long-range interceptors. In addition, it would be difficult for the parasite bomber to reengage the carrier, if that was necessary. Locating and rendezvousing with the carrier would be problematic enough, but the severe aerodynamic turbulence caused by the carrier's wake would make the hook-up itself very dangerous. 23 Although there were good reasons for dropping parasiting, some important decisions about the supersonic bomber had been based on the assumption that parasiting would be employed in the program. The design objectives for GEBO II, for example, were raised in 1950 because parasiting was incorporated into the project. It was assumed that parasiting would provide the system with a total radius of 4,000-5,000 miles, as opposed to the 1,380-2,875 miles specified prior to that time. Convair later estimated that a parasite system would have a 4,600-mile radius, but only if the entire mission was flown at subsonic speeds. If the paraas this.

21.

Brig.

Gen. Howard G. Bunker, Chief, Maj. Gen. Street, Aug. 23, 1949,

randum to Memorandum.

USAF Files

Atomic Energy, Memohereafter as Bunker B-58

Field Office for

HO/ASD;

cited

General Operational Requirement. SAC later insisted that the B-58 needed as well, although this requirement was not specified in 1951. 22.

23.

Bunker B-58 Memorandum.

a tail

gun

The Push

to

Develop Supersonic Capabilities

bomber flew through the target zone at supersonic speeds, the system's radius would be between 4,150 and 4,400 miles. Even these site

though, were totally dependent on parasiting. Boeing estimated that, without parasiting, a bomber flying a subsonic mission would have a radius of only 2,600 miles. If part of the mission was flown at supersonic speeds, the radius would be only 2,200 miles. The Air Force's 1951 requirements were based loosely on Convair's and Boeing's performance estimates, but the conditions under which these estimates had been made were ignored. The requirements called for a 4,600-mile radius, with one outbound refueling, and a 2,650-mile unrefueled radius. Convair had estimated that a 4,600-mile radius was possible, with parasiting, and Boeing had estimated that a 2,600-mile radius was possible without parasiting. The Air Force, therefore, started capabilities,

by assuming that aerial refueling would substitute for parasiting, although it did not conduct a careful analysis of this issue. More important, it required supersonic capabilities for the new bomber, even though both Convair and Boeing had explicitly specified that these ranges were attainable only if the bomber's entire mission was flown at subsonic speeds. The Air Force did not base its requirements on what Convair and Boeing thought a supersonic bomber could do. These requirements were, therefore, beyond the state of the art. Convair's bomber, moreover, would have to be completely redesigned to meet these requirements. Its relatively small, parasite bomber was designed to be carried; it could not hold much fuel. Parasiting was important to Convair's design effort in one other effort. Its bomber featured a highly swept (60 degree) delta wing, so it would probably have stability and control problems when it took off and landed. This was not a problem when it relied on a carrier for taking off and landing, but now parasiting was proscribed. Convair would thus have to deal with serious aerodynamic problems unless it moved away from a highly swept wing. There was one other twist in the Air Force's requirements. The new bomber was also required to have a low-altitude capability, because Air off

would But since it was

Force intelligence forecasts indicated that Soviet air defenses

make

high-altitude penetration problematic after 1957.

hard to design a bomber to fly efficiently at both high and low altitudes, the Air Force subsequently issued separate sets of requirements for the high-altitude and low-altitude missions. The B-58 program focused on the former, and the latter were never pursued. 24 General Operational Requirement. Also, Brig. Gen. D. N. Yates, Dir. of Res. and Devel., Dep. Chief of Staff for Devel., "Devel. Dir. No. 34: High Alt. Strat. Bomber/ Reconn. Weapon System," and "Devel. Dir. No. 35: Low Alt. Strat. Bomber/Reconn. Weapanalyze the evolution of the low-altitude on System," Feb. 29, 1952, Files, HO/ASD. mission in Chapter 7. 24.

I

Flying Blind

By the end of 1951, the basic performance requirements for what would eventually be the B-58 bomber were in place, and the program was moving toward full-scale development.

Design Competition

A

formal design competition between Convair and Boeing was initiated in February 1952. Convair's MX-1964 design featured the highly

swept (60 degree) delta wing it had used in many of its parasite designs. Convair believed correctly, as it turned out that performance was the key to winning the competition, and highly swept wings promised better supersonic performance than straighter wings. Boeing's more conventional MX-1965 proposal featured a thin, swept wing that made more allowances for the stability problems associated with taking off and landing. Even so, Boeing's design involved technological advances that could be attained only in a high-priority program. Convair's design involved many advances in the state of the art and even more developmental risks than Boeing's, according to the Air Force. 25 The Air Force began to review these proposals in August. Its assessment was that both designs would meet its minimum requirements. Convair's bomber seemed to have better all-around performance in terms of cruising speed, acceleration, top speed, altitude, and range, but the Air Force believed that Convair's supersonic drag estimates were optimistic by 10-15 percent, and hence that its performance estimates were also optimistic. In addition, it believed that the weight estimates for both aircraft were optimistic. This was an important issue because it would not be possible to refuel two bombers with one tanker if the weight of the bomber (and, consequently, its fuel requirements) grew too much. The Air Force concluded that an operational capability could not be achieved by 1957 but estimated that the new bomber would be ready in 1958 if timely decisions were made. 26 The head of the Air Force's new Air Research and Development Command (ARDC) noted, though, that the cost of developing and producing the new bomber had received little attention in the review of the proposed designs. In addition, he estimated that the B-58 would not be ready until 1959. 27

—

25.

WADC

—

Report, “Design Approaches: Boeing MX-1965 and Convair MX-1964," Oct. Report; Col. Victor R. Haugen, Chief, Weapons Systems

WADC

8, 1952, cited hereafter as Div., WADC, Letter to CG,

Files,

HO/AFSC. Haugen Letter. See

ARDC,

Oct.

8, 1952, cited

hereafter as

Haugen

Letter,

both

in

Haugen, Chief, Weapons Systems Div., W. E. Lamar, Aug. 12, 1952, Files, HO/ASD; cited hereafter as Haugen Memorandum. See also Thomas, History of the Development of the B-58, p. 119. 27. Lt. Gen. E. E. Partridge, CG, ARDC, Letter to Dir. of Res. and Devel., Office of the Dep. Chief of Staff for Devel., HQ USAF, Oct. 31, 1952; cited in Thomas, History of the 26.

WADC, Memorandum

to

also Col. Victor R.

The Push

to

Develop Supersonic Capabilities

The Air Force issued a new and even more demanding set of performance requirements for the supersonic bomber in October 1952, after the contractors had submitted their designs for the source selection review. Speed requirements were specified in more detail: a top speed of at least Mach 1.7, with Mach 2 desired. 28 Although Convair had estimated in 1951 that top speeds of Mach 1. 8-2.0 might be attainable in a parasite bomber, the Air Force was now asking for comparable capabilities in a bomber that did not have the luxury of being carried halfway around the world. Once again, the Air Force institutionalized a performance estimate as

made

offs that

a

requirement while ignoring the design trade-

the estimate possible in the

first

place.

And

the contrac-

had thought the Air Force's requirements extremely demanding even before they were raised in October. 29 Although the Convair and Boeing design studies were not aimed at tors

the

new

requirements, the Air Force nonetheless finalized

source selection in November. The design competition was terminated, and its

Convair was awarded development and production contracts B-58.

The expectation

that Convair's

bomber would generate

for the

better per-

formance than Boeing's was a critical factor in this decision. 30 Boeing's George Schairer later observed that "Wright Field insisted on impossible performance for that size plane. We came in with a heavier plane and were honest as to what we could do. We wouldn't do what they required.

I

was

attacked for being uncooperative." 31

The B-58 source

selection

was consequently based on

the simple de-

126-127. Also, Haugen Letter. The Air Research and Developestablished alongside the Air Materiel Command in 1951, with primary ARDC having primary responsibility for research and development and responsible for all these had been areas responsibility for production and logistics.

Development of

the B-58, pp.

ment Command was

AMC

AMC

prior to 1951.

When

this reorganization

went

into effect,

AMC's engineering

became ARDC's Wright Air Development Center. For an overview

division

of these organizational

developments, see the appendix. 28. WADC, "Proposed Military Characteristics: High Alt. Strategic Bomber/Reconn. Weapons System," Oct. 8, 1952, Files, HO/AFSC. 29. See the interview with Boeing's George Schairer, Dec. 30, 1954, transcript in Boeing archives.

Chief of Staff, "High Alt. Strat. Bomber/Reconn. Weapons System," Letter to CG, ARDC, Nov. 1952, Files, HO/AFSC. See also Haugen Letter; WADC Report. It is possible that the Air Force's source selection was influenced by the fact that Convair had just lost the B-52/B-60 competition and that its B-36 production run was scheduled to end in 1954. Since Boeing was scheduled to produce large quantities of both the B-47 and the B-52, it did not need another bomber contract as much as Convair did. Although Convair did have 30.

programs (including the F-io2 and F-106) under way in 1932, its team of bomber specialists would have disintegrated if it did not receive another bomber contract. Therefore, this is not a good test of the relative importance of the Air Force placed on aircraft performance and the defense industrial base. These considerations were mutually reinforcing in this case; we have an overdetermined outcome. 31. Interview with George Schairer, Dec. 30, 1954, Boeing archives. See also Haugen several fighter

Letter;

WADC

Report.

Flying Blind

sign competition that took place in 1952. This competition was very different from the prototype competitions of the late 1940s. In the B-47

and B-52 programs, competing prototypes were

built,

flown, and com-

pared. Constructing the prototypes helped to refine cost estimates and flight testing

helped to pin

down

actual performance capabilities.

The

gave the Air Force a clear sense of the competitive advantages and disadvantages of each bomber and reliable data about aircraft performance, cost, and availability. None of this happened in the case of the B-58. Instead, a paper competition took place. Convair and Boeing simply refined their designs and elaborated on their performance, cost, and delivery estimates. In the absence of prototyping, the quality of the data fly-offs

available to the Air Force

when

it

made

its

source selection was highly

The data on stability and control at supersonic speeds, for example, were largely limited to theory and wind tunnel tests. Since no flight testing was conducted prior to source selection, the Air Force only later discovered the gap between actual flight characteristics and theoretical consuspect, especially in such a technologically adventurous program.

ARDC

acknowledged. Air Force cost estimates were inadequate. 32 Again, prototyping would have helped to provide more extensive and reliable information on the Air Force's procurement options. Why, then, was the source selection based on just a design competition? The Air Force argued that there were three good reasons for doing so. First, it maintained that prototyping was not necessary because the

clusions. In addition, as the

itself

design competition would provide "extensive data" for its deliberations. 33 The problem with this argument was that a design competition could not provide much hard data about technologically ambitious projects; it could only provide extensive speculation about aircraft performance and program costs. Second, the Air Force argued that a prototype competition would be too expensive, given the constraints imposed by its fiscal year 1953 budget. 34 This argument was disingenuous. It would not have been expensive to build developmental prototypes comparable to those built in the B-47 an d B-52 programs. Convair estimated in 1951 that two prototypes (including government-furnished equipment) and

an

cost $24 million. 35 This not less expensive than, the cost of the B-47

initial flight test

parable

to,

if

program would

was coman d B-52

prototype programs. Moreover, the defense budgets of the early 1950s were roughly three times those of the late 1940s. If the Air Force could 32.

Thomas, History

of the

Development of the B-58, pp. 24-25, 126-127.

Haugen Memorandum. 34. Ibid.; Haugen Letter; Albert E. Misenko and 1917-1978, HO/ASD, April 1979, p. 18. 33.

35.

"Review of Project MX-1626," Feb.

Philip H. Pollock, Engineering History,

16, 1951; Files,

HO/ASD.

The Push

to

Develop Supersonic Capabilities

afford a four-way fly-off in 1947,

it

certainly could

have afforded

a

two-

way

prototype competition in the early 1950s. Finally, the Air Force believed that a prototype competition would delay the program. 36 But, although prototype competitions appeared to slow the acquisition pro-

down, so much was learned in them that a great deal of time was made up in the production phase of the program. The Air Force's real concern was that a prototype competition would delay the production decision until both aircraft began flight testing. It was anxious to begin cess

production as soon as possible, while it had the budgetary resources to do so, and it did not want to wait for a fly-off. Preproduction funds for the supersonic bomber were included in the fiscal year 1953 budget, and the Air Force sible. 37

The

was

interested in allocating these funds as soon as posmilitary services had gone through several "feast-or-famine"

budgetary cycles over the course of the twentieth century, and the Air Force wanted to make an irreversible commitment to the B-58 while funds were available for it.

Procurement Strategy Underlying the source selection was the fact that the Air Force had already decided to structure the B-58 program around a highly concurrent procurement strategy. Along with the F-101 and F-102 programs, the B-58 program was one of the first to be affected by a series of new Air Force acquisition procedures, the net effect of which was to completely

abandon the sequential

strategies of the late 1940s.

The three main

fea-

new

concurrent approach were the weapon system concept, the Cook-Craigie procedures, and the prime contracting method. 38 tures of the

The weapons system concept was based on the assumption that modern weapons, jet aircraft in particular, had to be designed as total systems from the beginning. This was necessary, it was said, to ensure that the various pieces of the system

fit

together. In the case of a

bomber,

for

meant that the airframe along with the engines, navigation and communication equipment, and other critical military subsystems would be integrated into the design at the beginning of the development process. In addition, the weapon system concept held that training and operating equipment should be developed and produced concurrently with the weapon itself. This approach, it was said, would example,

36. 37.

this

Haugen Memorandum. Haugen Letter. For more on how concurrency was applied

programs in the 1930s, see W. D. Putnam, The Evolution of Air Force System Acquisition Management, Rand Corporation Report, R-868-PR, Aug. 1972, pp. 3-12. 38.

to acquisition

Flying Blind

expedite the transition from production to deployment. Maj. Gen. Donald Putt, vice commander of the ARDC, argued that "the complete

should be planned, scheduled, and controlled, from design through test, as an operating entity. 39 He believed that all the Air Force's major acquisition programs should be developed under the weapon system concept. A Convair report described why the weapon system concept was needed:

weapon system

.

.

.

an airframe designed to achieve the extreme performance of the B-58, a high degree of design efficiency must be achieved. The airplane must be designed with a minimum of duplication of function, with no unnecessary design features, and with complete marrying to the airframe of the various systems required to perform the mission. As a result, the operating systems 40 in the aircraft must be integrated closely from a functional standpoint. In

A memorandum

from Gen. Hoyt Vandenberg, the Air Force chief of outlined what were generally known as the Cook-Craigie pro-

staff,

cedures:

The

initial rate

that

minimum

of production of rate required to

new

aircraft or

equipment

will

be held to

produce adequate quantities of the

engineering, functional, and suitability testing.

Once

the testing

article for

program

has demonstrated the final aircraft or equipment configuration suitable for issue to the using agencies, the rate of production will be increased to the level needed to meet inventory requirements. 41

Another Air Force memorandum explained how the Cook-Craigie procedures would be applied the B-58 program: concept, sometimes referred to as the "Cook-Craigie" concept and the "Slow Build-Up" concept, visualizes an 18- to 24-month accelerated

This

new

beginning of the program, during which The major objecaircraft production is held at some low sustaining rate. tives are: (1) to proof test the airplanes produced during these first 18-24

development

test

period

at the

.

months

in

plant, fire

.

.

order to disclose any inherent deficiencies in the airframe, power control system, and other components of the weapon system; (2)

“General Policy Guidance on Use of Single Prime Contractor for Development of a Complete Weapon System," Memorandum to CG, WADC, Dec. 8, 1932, Files, HO/ASD; 39.

Memorandum. “Convair's Role as Weapons Systems Manager,"

cited hereafter as Putt

Convair,

40.

AFSC; 41. ter to

cited hereafter as

"Qualitative

HO/

Convair Report.

Changes

CGs, ARDC, AMC,

cited hereafter as

July 24, 1953, Files,

to Aircraft

Air Proving

Vandenberg

Letter.

and Equipment Programmed for Production," LetGround Command, Oct. 16, 1952, Files, HO/ASD;

The Push

to

engineer

to

Develop Supersonic Capabilities

necessary to correct these deficiencies; and (3) incorporate these fixes into the production line prior to the time the airplane is released for quantity production. 42 fixes

In the case of the B-58, the

first

thirty aircraft

were

be used for testing

to

purposes. 43

The 'slow build-up concept" was

a blueprint for highly concurrent

programs. In the first place, it eliminated prototype testing from the development phase of the program; instead, the first batch of production articles would be used for testing. Second, production would take place, though at a low rate, while testing proceeded and the system's design was being finalized; development and production would over-

The Air Force occasionally asserted, quite disingenuously, that the Cook-Craigie approach was really a plan to "fly before you buy." 44 But under the Cook-Craigie procedures, there was no way that one could lap.

begin flight testing until after a production commitment had been and production aircraft were built.

The

third of the Air Force's

new

acquisition procedures

made

was the prime be done under

contracting method. Since system integration needed to one roof and since the Air Force itself did not have the technical or managerial capabilities to do this, the Air Force decided that it would select one company to be the prime contractor on each of its projects. A

Senate report explained how the prime contracting method differed from previous procurement policies: In 1946,

when

the prototype B-36 flew for the

Government

the

first

time,

it

was customary

for

to furnish the aircraft contractor the engines, instruments,

wheels, brakes, and all but the simplest mechanical and electrical subsystems. For older aircraft, standardization of a sort was possible because many of them could use the same powerplants, armament and electires,

"Government Furnished Equipment" was concerned. The airframe contractor's effort was theoretically limited to installing this collection of equipment into the airplanes and hooking it into the necessary wiring and plumbing. 45 tronics insofar as this

42.

Col.

1953, Files,

J.

F.

McCarthy,

HO/ASD. See

J. T. Cosby, Research Center.

director,

43.

Bomb.

Jr., Chief, Aire. Br., Memorandum to Lt. Col. Bostwick, May 7, also the transcript of the interview with Convair's B-58 program

USAF

Aire. Div.,

Oral History Program, Dec. 1973,

WADC,

“Res.

and Devel.

Files,

45.

Historical

Proj. Report: Strat. Air. Devel.,

RB-58" Memorandum to HQ/USAF, May 1, 1954, Files, HO/AFSC. 44. See Dennis J. Stanley and John J. Weaver, Air Force Command

HO/AFSC, no

Simpson

for

R

&

YB/

D, 1949-1976,

date, p. 19.

The B-58 Program, Report of the Preparedness Investigating Subcommittee, Senate Services Committee, 86th Cong., 2d sess., June i960, p. 20.

Armed

—

Flying Blind

This approach became unworkable, the report continued, as more exacting requirements posed more difficult technical problems and higher performance machines needed more specialized subsystems. The unique and complex nature of jet aircraft seemed to require a new ap-

subsystem development and a more prominent role for the airframe manufacturer. The B-58 program was the first to employ the prime contracting method, and Convair was the prime contractor on this proach

to

project.

These formal procedures combined

to

make

the B-58

program highly

The Air Force relied almost totally on paper studies for its source selection and production decisions. Competition was cut off before full-scale development began. True developmental prototypes were not built. The Air Force allocated $15 million in fiscal year 1953 and $17 million in 1954 for preproduction work and production tooling and materials. Procurement of production tooling had to precede flight testing since, by definition, the first flight of a B-58 was to be made by a production aircraft. Production plans and a de facto production decision therefore had to be made virtually at the outset of the development program. 46 As a result, a substantial financial commitment was made to the

concurrent.

—

including the B-58 early in the acquisition process. Critical subsystems bombing system, navigation system, and defensive armament system were developed concurrently with the airframe, even though the exact

performance capabilities of the aircraft itself were not known. These subsystems were carefully integrated into the design of the system at the beginning of development, so any major changes in the design of the aircraft would reverberate throughout the entire system. Finally, the development and production schedule was quite compressed. One outside review observed that the program featured "by far the shortest time in which any complete weapon system of similar complexity (and none 4 The B-58 proothers compare in this respect) has been scheduled." gram therefore exhibited all eight of the features of concurrency outlined "

in

Chapter

1

(see Table

2).

Although the formal B-58 production decision was made in late 1953, after the mock-up of the aircraft was reviewed, a de facto production decision was made much earlier. Given that preproduction funding had been appropriated in fiscal year 1953 and that Boeing had been eliminated from the competition in late 1952, the Air Force was committed to giving Convair a production contract. Its only alternative would have been to delay the program, go back to design competition, and turn its back on the preproduction funding already spent. The Air Force's intention to begin production of the B-58 at the earliest possible date is reflected in "WADC Analysis of Design Approaches: Boeing MX-1965 and Convair MX-1964," Oct. 8, 1952, Files, HO/AFSC. 47. U. S. Department of Defense Ad Hoc Study Group (the Robertson commission). 46.

Program for Reducing

the

Time Cycle from Concept

to Inventory,

6 vols., July 1956, vol. 5, app.

p. 4.

I178I

2,

The Push

to

Develop Supersonic Capabilities

Hidden Challenges According

to

one Rand Corporation study, highly concurrent pro-

grams are successful only

weapon

is

the specific technical capability required of a firmly established at the outset of the program and careful if

and detailed planning can eliminate configuration fluctuations. 48 Neither of these conditions was met in the B-58 program. The performance requirements for the B-58 were still being defined and raised in late 1952, after the contractors had submitted their designs and proposals. These new requirements were not formally put into place until September 1953, almost one year after source selection. 49 Clearly, the specific technical capability required of the B-58 was not firmly established before the acquisition program began. Careful and detailed planning did

—

—

not eliminate configuration changes. Indeed, the design of the airframe itself had to be changed twice, in 1953 and 1954. Finally, Convair's performance estimates were consistently optimistic, and there were many

problems associated with subsystem development and integration. This completely disrupted the elaborate plans and integrated designs drawn up at the outset of the program. The underlying problem was that the B-58 program was exceedingly ambitious technologically. The root of this problem was the program's performance requirements, which were demanding in 1951 and raised in 1952. The new requirements were issued over the objections of the Air Force project officers who believed that they were unreasonable given the state of the art. 50 George Schairer of Boeing felt that the Air Force's requirements were "impossible," and he noted that B-58 project officers who agreed with this assessment were "all sent away" to other assignments. 51 One Air Force study concluded that "considerable doubt existed in the airframe industry" in 1951 and 1952 about the feasibility of meeting these requirements given "the state of the art in airframes, propulsion, and electronic systems." 52 Convair was asked in early 1953 48. Putnam, Evolution of Air Force System Acquisition Management, pp. 6-7. The Air Force recognized that freezing the aircraft design was critical under the Cook-Craigie procedures; see Vandenberg Letter; Lt. Gen. K. B. Wolfe, Dep. Chief of Staff for Materiel, Memorandum for Gen. Twining, March 28, 1951; Gen. Nathan F. Twining, Vice Chief of Staff, USAF, Memorandum to CG, AMC, April 11, 1951; and Lt. Gen. Orval R. Cook, Dep. Chief of Staff for Materiel, Memorandum for Gen. Putt, Aug. 29, 1951; all in Design and Configuration Freeze (Production and Inventory Aircraft, 1951), RG 341, HQ/USAF,

NARS. “Military Characteristics for the B-58 System," Sept. 11, 1953, Files, HO/AFSC. 49.

Thomas,

High Altitude

Strategic

Bombardment Weapon

History of the Development of the B-58, p. 152. 51. See the interview with George Schairer, Dec. 30, 1954, Boeing archives. Project Number 97," June 30, 1961, Files, HO/ASD. 52. ASD, “Report on 50.

OSD

Flying Blind

if

the

new requirements were

feasible.

It

replied that the majority of the

requirements could not be met. 53 The Air Force nonetheless formally established these requirements in September. As a result, major technological advances had to be made over the course of the B-58 program. Convair observed that the program involved "an extremely large advancement in the state of the art in three major design provinces, namely the airframe, propulsion, and electronics." It noted that further development in these areas would be required before the B-58 would be able to meet its requirements, a problem complicated by the short time available for development. 54 A Senate report agreed with Convair's assessment of the technological challenges that surrounded the B-58 program: "The B-58 weapon system will be the most complex, highly integrated, and mutually interpendent weapon system in the history of the Air Force. Extension of the 'state of the art' in many fields of endeavor will be required in order to meet the required operational environment." 55

acknowledged that the program was extraordinarily ambitious: "The B-58 program with Convair was one in which the aircraft industry and the ARDC were pushing the state of the art in practically every component area. The B-58 involved an unknown supersonic field, a realm where theoretical aerodynamicists had no equations to guide them, where many could predict the questions but few could come up with the correct answers." 56 Since the B-58 program involved new technology at the system level, it belongs in the second category on the nine-point scale of technological ambitiousness outlined in Chapter

The Air Force

1

(see Table

itself

1).

Two

design problems deserve special mention. The airframe and surfaces required new and exotic materials in order to withstand the high temperatures that would be generated by sustained Mach 2 flight. For example, aerodynamic friction would cause the leading edges of the wings to heat up to 26o°F. The wings were consequently covered with

sandwich panels, each with an outer skin of thin aluminum sheets and a fiberglass and aluminum honeycomb filler. The sandwich was bonded together with adhesives set under heat and pressure. The production process was complicated by the fact that each panel had to be individually crafted and built on its own special press. 57 The advanced electronics needed for the bombing and navigation sysspecial

53.

Thomas,

56.

ARDC/HO, History of the ARDC, Jan. to June 1954, vol. Senate Armed Services Committee, B-58 Program, p. 6.

History of the Development of the B-58, p. 152. 54. Convair Report. 53. Senate Armed Services Committee, B-58 Program, p. 21. 57.

1, p.

241-242.

[180]

The Push

to

Develop Supersonic Capabilities

tem also involved several advances. The

ARDC outlined

the basic prob-

lem:

The question of how a B-58 would find and hit its targets, particularly at the speed and altitudes at which the vehicle was designed to operate, was one fraught with

.

.

.

difficulties.

For example, in order to obtain a three-minute

bomb

run for a B-17 operating at 25,000 feet, the bombardier would have to get on his target 11 miles away. With a B-47 operating at 40,000 feet and an airspeed of 450 knots, the bombardier, to get his three-minute bomb run, would have to spot and track his target at least 25 miles away. But, to have a three-minute bomb run at the B-58's designed speed of Mach 2 and at an altitude higher than 50,000 feet, the bombardier would have to be on target some 65 to 70 miles away. 58

The B-58 therefore needed

improved bombing and navigation the system that was ultimately developed

a radically

system. Convair believed that represented a "quantum jump in technology for its time." 59 A Department of Defense review concluded that the B-58 "will be the fastest large aircraft of its

necessity, plex.

it

time and very advanced in other performance aspects. By will be one of the most 'dense' aircraft and the most com-

The major components contain much new

art in circuitry

technique." 60 In addition, substantial miniaturization

was required

the complex electronic systems into a relatively small fuselage.

It

and to

fit

was,

however, impossible to know exactly what bombing and navigation capabilities would be needed and how much space would be available for these systems until the exact capabilities of the B-58 were known and the design of the aircraft itself was established. Although the Air Force had assumed in late 1952 that the design of the B-58 could be frozen and that development and production could be undertaken concurrently, the magnitude of the associated technological advances suggested that this was risky. The subsequent history of the B-58 program bore this out: the design of the aircraft had to be changed in fundamental ways after the production program got under way, and major technological problems continued to plague the B-58 through the end of the decade. 61 In fact, basic aerodynamic research was being conducted at NACA even as the design of the B-58 was supposedly being

—

58.

—

ARDC/HO,

History of the ARDC, Jan. to Dec. 1953, vol. 1, pp. 439-440. Kaler, Manager of Aerodynamics, Fort Worth Div., General Dynamics,

"The 59. G. M. B-58 and FB-111: The Supersonic Cruise Precedents." Convair was purchased by General Dynamics in May 1953, and it officially merged with General Dynamics in April 1954. 60. Department of Defense Ad Hoc Study Group, vol. 5, app. 2, p. 3. 61. Based on the accounts in Thomas, History of the Development of the B-58, pp. 140, 147148; Alex Roland, Model Research, 2 vols. (Washington: National Aeronautics and Space Administration, 1985), vol. 1, pp. 280-281.

Flying Blind

frozen.

NACA

strong shock

research on transonic aerodynamics focused on the

wave formed where

the leading edge of an aircraft's

wing

met the fuselage. Wind tunnel tests confirmed that it was important to keep the total cross-sectional area of an aircraft including its fuselage, wings, engines, and droppable pods as constant as possible in order to minimize aerodynamic drag and the formation of this shock wave. According to this "transonic area rule," an aircraft's fuselage had to be indented to compensate for the cross-sectional area of the wings and engines. This produced a "coke bottle" shaped fuselage that had a greater diameter at the nose and tail than at the midsection.

—

—

Convair's design

was

inconsistent with the transonic area rule. This

is

not surprising, since Convair's proposal in the 1952 design competition was finalized before the discovery of the transonic area rule was publicized. The B-58's fuselage consequently had to be indented in March 1953. In addition, the leading edge of the wing was cambered or twisted,

and the

edge of the wing was swept

improve the aircraft's aerodynamics. As a result, the fuselage of the densest, most highly interdependent aircraft ever built had to be completely redesigned. The B-38 was a small and tightly packed aircraft to begin with; trailing

indenting

its

fuselage only

compounded

in order to

the space problem. In the end,

change reverberated throughout the system, affecting the basic design parameters of every subsystem. The delicate design package Convair had put together in 1932 had to be completely reworked. In addition, Convair had to redesign and rebuild its custom-fitted wing panels. this

NACA's

focus subsequently shifted to supersonic flight regimes.

It

speeds greater than Mach 1.2, the critical aircraft cross-sectional area was conical, corresponding to the conical shape of the supersonic shock wave, rather than a simple plane. According to the "supersonic area rule," therefore, the fuselage indentation should be influenced by the speed at which the aircraft was supposed to fly, since the angle of the supersonic shock wave (and, therefore, the angle of the critical conical cross section) varied with the speed of the moving body. This meant that the B-38's fuselage had to be redesigned again in August 1934. Changing the design at such a late date naturally delayed the overall program and added to its cost. Contrary to what the Air Force had assumed, elaborate planning exercises could not substitute for basic

concluded

that, at

research.

Persistent Problems

The B-38 program continued to be plagued by aerodynamic problems throughout the 1930s. Drag predictions, critical in a program confronted

The Push

to

Develop Supersonic Capabilities

Convair/General Dynamics B-58 (U.

S.

Air Force)

by extremely demanding speed and range requirements, had not been determined, even in March 1953. In addition, "flutter experience with the delta wing at that time was meager, although engineers had devised time-consuming and unproven theoretical methods for predicting it. The safety factors for the aircraft were consequently questionable." 62 Convair admitted in May 1954 that its drag estimates were based on a wind tunnel program that was not yet completed, and it acknowledged that some performance estimates could be off by as much as 15 percent. 63

By the end of 1954, the ARDC was highly skeptical about Convair's performance estimates. 64 A joint ARDC-AMC analysis indicated substantially greater subsonic drag than predicted by Convair. It also estimated the B-58's unrefueled radius to be only 1,950 miles (instead of the required 2,600 miles) and its refueled radius to be only 3,950 miles (instead of 4,600 miles). In addition, the costs of the program had increased beyond original estimates. Unfortunately, $122 million had already been spent on the program, and another $182 million was already obligated. 65 62. 63.

Thomas,

History of the Development of the B-58,

Joseph

McNarney,

T.

Pres.,

Convair

Div.,

p. 151;

see also

p. 139.

Consolidated Vultee Aire., Corp., Letter to

Commander, WADC, May 27, 1954, Files, HO/AFSC. 64. Brig. Gen. Howell M. Estes, Jr., Dir. of Weap. Systems Ops., WADC, Letter to Commander, ARDC, Sept. 30, 1954, Files, HO/AFSC. See also ARDC/HO, History of the

ARDC,

July to Dec. 1954, vol. 1, pp. 242-246. Briefing to the Air Council, "B/RB-58 Weapon System Prog. Re65. Joint view/' Feb. 3, 1935, Files, HO/AFSC. Also, Thomas, History of the Development of the B-58, p.

ARDC-AMC

194;

ARDC/HO,

History of the

ARDC,

July to Dec. 1934, vol.

1,

pp. 243-2433.

Flying Blind

The deterioration in performance estimates led SAC to criticize the program harshly. LeMay had been skeptical of the B-58 from the beginning. He had reservations about medium-range bombers in general, because they were highly dependent on aerial refueling and forward bases, and he was especially wary of the B-58 because its high speed would undoubtedly impinge on its range. 66 It is not surprising, then, that

SAC's

criticism intensified

when

the new, lower-range estimates

SAC's director of plans stated that the B-58 was no longer a suitable replacement for the B-47 and that it certainly was not an intercontinental bomber. In his opinion, range was a critical consideration as long as the Soviet Union, and not Canada, was the country's most likely adversary. 67 LeMay bluntly stated to the Air Force chief of

came out

in late 1954.

January 1955 that the B-58 was not desired in the SAC inventory because of its expected range limitations. 68 believed that the program was nonetheless The ARDC and staff in early

AMC

worthwhile because it was helping to consolidate technological advances. Others believed that the B-58 might still turn out to be a useful weapon system, although some thought it might be better suited for the short-range missions of the Tactical Air Command than for SAC. As for SAC, it was less than enthusiastic about the B-58, but it had no other along in early 1955. The Air Force's program to build a nuclear-powered bomber was floundering by this time, and the B-70, which was intended to be the successor of the B-52, was still in the requirements formation stage (see Chapter 6). 69 SAC was anxious to follow-on

bomber

this far

incorporate supersonic capabilities into the force structure, and the B-58

represented the only way of doing so in the near term. The Air Force, therefore, was ambivalent toward the B-58 at this time. It decided in June 1955 to buy only thirteen B-58S and to use these aircraft for aerodynamic testing only. This decision was reversed in August, however, and the program was once again oriented toward building a fully operational

ARDC/HO,

weapon system. 70 ARDC,

440-441. The Air Force believed that the B-58 would normally depend on forward bases planned for B-47 operations but could be operated intercontinentally on a limited basis; see "Tentative B-58 Operational Concept," Nov. 15, 1954, Files, HO/AFSC. 67. ARDC/HO, History of the ARDC, July to Dec. 1954, vol. 1, p. 244. 68. Gen. Curtis E. LeMay, Commander, SAC, Letter to Chief of Staff, USAF, Jan. 4, 66.

1955, Files,

History of the

Jan. to Dec. 1953, vol. 1, pp.

HO/AFSC.

Maj. Gen. Albert Boyd, Commander, WADC, Letter to Lt. Gen. Thomas S. Power, Commander, ARDC, Dec. 17, 1954; Bomb. Aire. Div., Dir. of Weap. Sys. Ops., WADC, "Evaluation Summary: System 302A, B-58 Tactical Bomber," June 27, 1955; Lt. Gen. C. S. Irvine, Dep. Chief of Staff for Materiel, Letter to Commander, ARDC, July 28, 1955; all in Files, HO/AFSC. Also, ARDC/HO, History of the ARDC, July to Dec. 1954, vol. 1, p. 244. 70. Brig. Gen. Howell M. Estes, Jr., Dir. of Weap. Sys. Ops., WADC, Letter to Com69.

mander,

ARDC, June

29, 1955,

and HQ/USAF, Telex

to

Commander, ARDC, Aug.

22,

The Push

to

Develop Supersonic Capabilities

he first flight of the first B-58 did not take place until November 1956, almost two and one-half years later than Convair had originally pre1

dicted and the Air Force had originally hoped. The Air Force decided in July 1957, over SAC's strenuous objections, to assign the B-58 to SAC.

LeMay's main concern was that deployment of the B-58 might prevent him from fielding advanced models of the long-range B-52 or the B-70. Fie consequently argued in mid-1957 that the B-58 should be limited to a test program. 71 He ultimately acquiesced on the 6-58, but only after he was reassured that it would not interfere with the B-52 or B-70 programs.

One

reason the Air Force proceeded with the B-58 in 1957, according to the assistant vice chief of staff in charge of implementing the decision, was protecting the $750 million investment that had already been made in the program. 2 The huge amount of money required by the B-58's concurrent program thus had a strong influence on production and

deployment decisions. It was difficult to write off this kind of investment, even if the military value of the program was marginal. It is significant, too, that the results of the long-delayed flight test program were not noted as a good reason for going ahead with the program. Although deployment plans were in the works, major design problems persisted in the late 1950s. Encapsulated ejection seats were added in February 1958 because standard ejection seats were not adequate at Mach 2. Not only was this problem discovered relatively late in the program, but the encapsulated ejection seat itself was a new development. In addition, integrating such a system into the aircraft was difficult, because of the constraints imposed by the size of the fuselage. As a result, the operational configuration of the B-58

March

The Air

was not

finalized until

assumption that the B-58's final technical specifications could be determined at the outset of the development process proved to be invalid. 73 In addition, there were numerous problems with the bombing and navigation system. It was not ready for large-scale production in 1958, and it still had reliability problems in 1959. Some of SAC's B-58S were missing critical electronic components, 1958.

1955; both in 200-211.

Files,

Gen. Curtis

Force's

HO/AFSC. See

also

Thomas,

History of the Development of the B-s8 ’

on

LeMay, Commander-in-Chief, SAC, Letter to Chief of Staff, USAF, June 26, 1957, Files, HO/AFSC. SAC's strong feelings about the B-58 are also reflected in Gen. Thomas S. Power, Commander-in-Chief, SAC, Letter to Lt. Gen. Samuel E. Anderson, Commander, ARDC, June 20, 1958, Files, HO/AFSC. 72. Maj. Gen. Jacob E. Smart, Asst. Vice Chief of Staff, Letter to Commander-in-Chief, SAC, July 26, 1957, Files, HO/AFSC. 73. Director of Research and Development/HO, History, Director of Research and Development, Deputy Chief of Staff/Development, July to Dec. 1957, pp. 35-36; D. G. Powell and D. J. Shapland, "B-58 Encapsulated Seat Escape System," Stanley Aviation Corporation Report, No. 873, Feb. 19-20, 1963; Senate Armed Services Committee, B-58 Program, p. 19. 71.

E.

Flying Blind

which had

to

course,

was

SAC complained

that

be retrofitted as they became available;

an expensive and time-consuming proposition.

this, of

the B-58 had "limited operational capabilities, questionable reliability, and dubious maintainability" for some time. 74 Even in i960, the bomb-

ing and navigation search radar was problematic. Concurrent development of critical subsystems did not go smoothly.

proved to be critical, contrary to what had been assumed at the outset of the program. The program was shaken by a series of crashes beginning in December 1958 and continuing into i960. An investigation of the B-58's design concluded that "extrapolation of subsonic flight data for use in supersonic flight Control surfaces during supersonic was at the heart of the difficulty. 7^ flight were only half as effective as during subsonic flight." The lack of developmental prototypes and extensive flight testing in the early stages of the program, therefore, was a fatal flaw in the Air Force's procurement strategy. Critical design problems were not even identified until flight testing began, and they could not be solved through paper analyses or extrapolations from existing data. These extrapolations were the source of the problem. Extrapolations from the subsonic flight regime were dubious at best and invalid at worst. This should have been recognized in the early 1950s, when the discovery of the transonic and supersonic area rules suggested that some mighty aerodynamic unknowns still surrounded the B-58 program. This lesson was not learned. Instead, each new technological bombshell was seen as Flight testing of the B-58 itself ultimately

.

the

last.

As

.

.

a result, supersonic flight testing

was not worked

into the

an early stage, and the B-58 program did not reduce technological uncertainties in a timely manner. In the end, eleven pilots died in the B-58 flight test program. Nine of the thirty test aircraft eventually crashed, as did 20 pefcent of all the B-58S that were ultimately built. 76 There were many hints over the years that the B-58 had these kinds of aerodynamic problems, but they were not rigorously investigated until

program

at

late 1959.

plan to buy 290 B-58S and develop an advanced, longer-ranged version of the aircraft, the B-58B. By the end of 1959, it had canceled the B-58B and cut back the produc-

The Air Force subsequently scaled back

tion plan for the

program as

a

whole

its

to 148 aircraft.

The Air Force

Maj. Gen. John D. Ryan, Dir. of Mat., SAC, Letter to Dir. of Ops., SAC, May 12, HO/AFSC. For more details on the electronic system's problems, see Gen. Thomas S. Power, Commander-in-Chief, SAC, Letter to Gen. Thomas D. White, Chief of 74.

1959, Files,

HO/AFSC.

ARDC/HO,

History of the ARDC, July to Dec. 1957, vol. 1; Thomas, History of the Development of the B-58, pp. 238, 256-257. 75. Thomas, History of the Development of the B-58, pp. 261-262. Staff,

USAF, Aug.

76. Jay Miller,

28, 1958, Files,

Also,

"History of the Hustler," Airpoiver (July 1976), 22-38.

[186]

The Push

to

Develop Supersonic Capabilities

inspector general investigated the program in early 1960, and his office concluded that the B-58 was about to enter the operational inventory

with "significant unresolved problems." 77 The B-58 was also being squeezed by other programs in the competition for limited procurement funds. It faced stiff competition from the B-5 2 H, which had impressive range and subsonic speed capabilities, and the supersonic B-70, which SAC saw as the mainstay of its bomber force in the 1960s.

The Eisenhower administration increased the B-70 occasions over the course of 1960, which inevitably

budget on several impinged on the funds available for the B-58. Given that SAC preferred long-range bombers to the B-58 and that the Eisenhower administration was harshly criticized in the i960 presidential campaign for neglecting U.S. strategic striking power, it is not surprising that the long-range B-52 and B-70 bombers were favored over the troubled B-58. These bomber programs, moreover, had to compete with the ICBM programs that were consuming more and more of the Air Force's budget. 78 Convair tried to improve its competitive position by proposing yet another version of the B-58, the B-58C, in early i960. The Air Force believed, however, that Convair's performance estimates were once again wildly optimistic. It concluded, for example, that Convair's range estimate was optimistic by at least 25 percent. 79 The bomber's range was

was to compete against the B-52H and B-70. The B-58C proposal was not pursued, and by the end of i960 the B-58 production critical if

it

plan had been cut back to 116 aircraft. The first B-58 was not delivered to SAC until August i960, and the first wing of B-58S did not become operational until June 1961, approximately four years later than the Air Force had originally hoped. The last B-58 was delivered to SAC on October 26, 1962, in the midst of the

Cuban missile crisis. Secretary of Defense Robert McNamara announced in December 1965 that the B-58 would be phased out of the strategic Maj. Gen. Charles W. Schott, Dep. Inspt. Gen., Letter and Report to HQ/SAC, March 22, i960; quoted in Thomas, History of the Development of the B-58, p. 256; See also pp. 240-241. Also, HQ/USAF, Telex to Headquarters, SAC, June 11, 1959, Files, HO/AFSC. 78. The budgetary pressures on the B-58 program are discussed in Gen. Griswold, "Force Structure," Telex to Gen. Blanchard, Sept. 1, 1958; Col. Ernest C. Hardin, Jr., Dep. Dir. of Plans, "Force Structure," Memorandum, Sept. 5, 1958; Lt. Col. William T. Wilborn, "Force Structure Briefing for Gens. White, LeMay, and Members of the Air Staff," Memorandum for Col. Townsend, May 7, 1959; all in Files, HO/AFSC. For an analysis of the relationship between strategic bombers and ballistic missiles in the force structure debate of late 1950s, see Chapter 6 of this book and Desmond Ball, Politics and Force Levels (Berkeley: University of California Press, 1980), pp. 248-251. 79. Gen. S. E. Anderson, Commander, AMC, Letter to Dep. Chief of Staff for Devel., HQ/USAF, March 4, i960, and Brig. Gen. Seth J. McKee, Acting Dir. of Plans, SAC, Memorandum for Gen. Power, Commander-in-Chief, SAC, Sept. 14, i960; both in Files, HO/AFSC. See also Thomas, History of the Development of the B-58, pp. 248-251. 77.

[187]

—

Flying Blind

inventory because it was not a particularly cost-effective weapon system. The last B-58 was retired from service in January 1970. 80

Program Outcomes The B-58 program experienced

cost,

schedule, performance, and

problems substantially more serious than those experienced by either the B-47 or B-52 programs. The irony is that the Air Force adopted a highly concurrent strategy for the B-58 program in part because of the problems it associated with these earlier, more sequential programs. The cost overrun on the contract for the first thirteen B-58S was, fitmore than twice the 28.6 percent overrun on the first tingly, 58 percent twenty B-52S. 81 One Rand Corporation study determined that, although the B-47 an(l B-58 programs were comparable in technological ambitiousness, the B-47 had an average cost growth factor of only 1.17 while the B-58 had an average cost growth factor of 3. 67. 82 The initial cost estimates for the B-58 were wildly optimistic, and the program experienced massive cost growth, which ultimately affected the number retrofitting

—

—

had more experience at cost estimating in the early 1950s, when the B-58 program was started, than it did when the B-47 anc B-52 programs were begun in the 1940s; if everything else was equal, one would expect the Air Force to have done a better job at estimating costs in the 1950s. 83 The critical difference between the B-58 program and the earlier programs was its highly concurrent procurement strategy. In the absence of prototyping, of B-58S the Air Force could buy.

The Air Force

actually

*

the Air Force's B-58 cost estimates were, inevitably, highly speculative.

SAC/HO, Development of the Strategic Air Command, 1946-1986, Sept. 1986, pp. 90, discuss McNamara's 1963 decisions about the strategic bomber program in 109, 133, 165. more detail in Chapter 7. 80.

I

81.

L. E.

Preston, Contract Negotiations and Results

Procurement: Case Studies of RM-3254-PR, Sept. 1962, pp.

in Aircraft

and B-58, Rand Corporation Research Memorandum, 66-67, 7°~7 l 82. A program's cost growth factor is the ratio of its actual costs to its estimated costs. Since different estimates were made over the course of a typical program, each program has several cost growth factors. The average of these factors is noted here; see Robert L. Perry et al., System Acquisition Strategies, Rand Corporation Report, R-733-PR/ARPA, June the B-52

-

1971, pp. 8-13. 83. The unit costs of the B-58

were also substantially higher than those of the B-47 ar>d were around $2 million; the B-52, $6-7 million; the B-58, B-52. The unit costs of the B-47 $14 million. The size of the production runs undoubtedly had an effect on this. One should also note that the B-58's unit costs were substantially higher than the Air Force had expected, although the cutback in the size of the production run might have affected its unit costs. See Secretary of the Air Force, Unit Costs of Aircraft, Guided Missiles, and Engines, Jan. 30, 1964, p. 2; John Greenwood, History of the Strategic Bomber since 1945 (Washington: Office of Air Force History, April 1975), apps. D, E, H.

[188]

The Push

to

Develop Supersonic Capabilities

The B-58 program determined

also suffered

many

delays.

that, in the early years of the

One

Air Force analysis

program,

its

schedule was

slipping at a rate of more than one year per year. 84 At this rate of slippage, the program would, literally, never be completed. Although the rate of slippage improved as time went by, the B-58 did not become operational in wing strength until 1961, four years later than the Air Force had originally stipulated and two years later than its revised esti-

mates had predicted. 85 Convair acknowledged in 1963 that the B-58 flight test program took six and one-half years, instead of three years as originally planned, because the serious aerodynamic problems that emerged in flight testing required major changes in the design of the aircraft. Problems with subsystem development also delayed the program. In addition, Convair noted that production activities were disrupted by the major improvements that had to be incorporated into the design of the aircraft late in the game. 86 Delivery of the final batch of B-58S to SAC was delayed in 1962 because strict flight restrictions were still

in effect.

The pace

of the B-58 program, moreover, does not

compare well

to

and B-52 programs. From the time the first major development contract was awarded, the B-47 program took seven and onehalf years to have bombers operational in wing strength, the B-52 program took nine and three-quarters years, and the B-58 program took ten and one-quarter years. 87 The B-47 program, which was widely viewed in the Air Force as a procurement nightmare, was actually very effective in achieving an operational capability quickly. The B-58 program, on the other hand, sputtered along for over a decade before reaching the same point, even though it was comparable to the B-47 n technological ambitiousness. The B-58 program, though, was supposed to do a better job of getting aircraft into the operational inventory; that was the whole point of compressing the B-58's development and production activities. Not only did concurrency fail to achieve this goal, it was counter productive because the program's rigid schedule was extremely sensitive to unexpected developments and technological disruptions. The B-58 program was often disrupted by technological complications. When this that of the B-47

i

"An Analysis of the Slippage of Estimated Completion Dates on 100 jects," no date. Files, HO/ASD. 85. Thomas, History of the Development of the B-58, pp. 127, 153. 84.

86. 3,

General Dynamics, Fort Worth Div., "B-58 History: Performance

to

ARDC

Pro-

Schedule," June

1963, pp. 10, 12, 14, 18.

The

major development contract for the B-47 was awarded in late 1944, after the requirements for the program were issued in November; the B-47 became operational in wing strength in early 1952. The B-52 development contract was awarded in June 1946 and a wing of B-52S became operational in March 1956. The first major development contracts for the B-58 were awarded in March 1951, and the B-58 finally became operational in June 1961. The B-58 was not actually combat-ready until August 1962. 87.

first

Flying Blind

happened, production tooling had to be discarded, elaborate schedules had to be scrapped, and substantial amounts of both time and money were lost. Performance problems have already been discussed in detail. Although the B-58 ultimately demonstrated impressive speed and altitude capabilities (and even fair range capabilities), it did not and could not do so early in the development process due to the absence of prototype testing. Whereas the timely B-47 flight test program enabled it to demonstrate its superior performance capabilities and stand out against its competition, the B-58's actual capabilities were slower to emerge. Even then, the Air Force contended that the aircraft's performance did not meet the criteria laid out in the initial B-58 contract. Convair failed to meet these requirements even though the Air Force emphasized that performance was more important than availability or cost. 88 Finally, the B-58 required no fewer than eight modification, modernization, and retrofitting programs, including one that began in August 1963, long after the aircraft had been deployed. Structural modification of some wing and fuselage assemblies was needed, for example, and retrofitting operational bombing and navigation systems into the aircraft be done. 89 The final operational testing program for the B-58 was consequently a complicated and time-consuming affair.

had

to

Conclusions Three questions remain to be considered. First, why did the Air Force adopt a highly concurrent strategy in the B-58 program? Some of the reasons have already been alluded to. The Air Force certainly misread the lessons of the B-47 experience.

It

believed that prototyping

was

too

expensive, even though there was no evidence for this conclusion and even though the defense budgets of the 1950s were much larger than those of the late 1940s. It believed that prototyping took too long, al-

though the

real

problem was that prototyping delayed the production

decision until after flight testing took place. Finally, the Air Force believed that the B-47 program did a poor job of integrating critical sub-

systems and producing operational weapon systems. Concurrency's emphasis on subsystem integration, however, could not solve the problem of subsystem immaturity, which grew out of the Air Force's in88.

dum

See George

E. Oster,

Dep. Chief of B-58/B-47 Sect. Bomb.

MemoranOster Memorandum.

Br., Aire. Div.,

to Col. Wetzel, Feb. 23, 1956, Files, HO/AFSC; cited hereafter as See also Preston, Contract Negotiations, pp. 78-79. 89. See General Dynamics, "B-58 History: Performance to Schedule," p. 20.

The Push

sistence

to

Develop Supersonic Capabilities

on dramatically improved performance

capabilities in

its

new

aircraft. 90

The Air Force's analytic case for employing concurrency, therefore, was not overwhelming. Powerful political and organizational considerations, however, also favored concurrency. The Air Force's political motivation was fairly straightforward: it was anxious to make an irreversible commitment to production before the unexpectedly large defense budgets of the early 1950s disappeared. It did not need or want reliable cost, schedule, and performance estimates. Instead, it wanted to make a quick production decision and begin production tooling. It wanted to make a substantial financial investment in the program as soon as possible. A sequential procurement strategy and extensive developmental prototyping would only have stood in the way. The Air Force's misreading of the lessons of the B-47 program was, therefore, a motivated misperception if not a willful distortion of acquisition history. 91 The organizational component of this decision is that concurrency radically simplified the acquisition process and eliminated a wide range of nagging uncertainties, at least from an organizational perspective. In ly concurrent program, for example, the Air Force did not have

a highto

wait

from the competition. In addition, it could begin to make elaborate production and deployment plans at once, which engaged many of the Air Force's main commands. A second question is, did the Air Force appreciate the risks associated with a highly concurrent strategy in such a technologically ambitious program? Clearly, it did not. In the first place, these decisions were made independently. The supersonic bomber development program had been extraordinarily ambitious from the beginning. Concurrency was injected into this development effort in the early 1950s for reasons that had nothing to do with technological risk. In the second place, the magnitude of the technological unknowns that surrounded the B-58 program was not appreciated when concurrency was introduced into the program. These unknowns seem to have been discounted because the strategic bomber was a weapon that fell within an established mission area. The problem of building a bomber that could fly faster and for a fly-off to eliminate contractors

growing complexity of weapons required early integration of subsystems and a weapon system approach instead of extensive prototyping was widely held in the Air Force in the 1950s; see Misenko and Pollock, Engineering History, p. 19; Senate Armed Services Committee, 6-58 Program, pp. 19-23; Lt. Col. J. W. Colopy, "Weapon Systems and the Weapon System Concept," Air University Quarterly Review 9 (Spring 1957), 105-114; "The WSPO Concept: Space Age Procurement," Armed Forces Management 90.

The

belief that the

(Feb. 1959), i4-!8. 91. See Robert Jervis, Perception and Misperception in International Politics, (Princeton:

Princeton University Press, 1976), chaps.

4, 10.

Flying Blind

predecessors fell within the realm of what was intuitively solvable. The Air Force consequently believed that its paper studies would adequately resolve whatever technological details remained. As a higher than

its

result, technological obstacles

were

virtually

assumed away. Concurren-

cy did not appear to be particularly risky under these conditions. did the Air Force emphasize aircraft performance over operational availability and program costs? Clearly, the Air Force had the latitude to emphasize one set of program outcomes over another. More-

Third and

over,

it

clear that

is

program

last,

it

intentionally sacrificed operational availability

and

The Air Force knew from the beginning, Convair's design for a supersonic bomber was riskier

costs for performance.

example, that than Boeing's, but it selected Convair's design because of its estimated performance. The priority placed on performance was also reflected in the incentives built into the B-58 contract. Convair's fee was to be based on the following weighted formula: one-half for performance; one-third for schedule; and one-sixth for cost. One can conclude that the Air Force was more interested in technical performance than cost and delivery, and twice as interested in the delivery schedule as in program costs. 92 When production problems eventually emerged, moreover, the Air Force was unwilling to make even "minor deviations" from its performance specifications in order to improve the delivery schedule. 93 Overall, then, the Air Force had a great deal of control over the outcomes of

for

the B-58 program. 92.

Preston, Contract Negotiations

93.

Oster

Memorandum. See

,

p. 75.

See also Oster

Memorandum.

also Preston, Contract Negotiations, pp. 78-79.

[192]

The Nuclear-Powered Bomber

and

the

B-yo

The Air Force made a concerted effort to build two new long-range bombers in the 1950s and early 1960s: a nuclear-powered bomber and the B-70. Both ventures were unsuccessful. The nuclear-powered bomber program was extremely adventurous from a technological standpoint, to say the least. Its success depended on the development of a small but powerful nuclear reactor as well as lightweight but durable shielding materials. These problems would have been formidable under the best of circumstances, but they were

compounded by

the Air Force's decision to develop the entire

weapon

system concurrently with the nuclear propulsion unit. As a result, resources were diverted from the program's most important technological puzzles, and a nuclear-powered bomber was never built. The B-70 was relatively conventional in that it was powered by jet engines and chemical fuels. Even so, it ultimately proved to be one of the most ambitious bomber development programs of the postwar era. The performance requirements set at the outset of the program in 195455 were simply unattainable, given the state of the art. A totally unforeseen breakthrough in basic aerodynamic research at NACA subsequently enabled the leading contractor on the project. North American,

bomber that was projected to fly its entire intercontinental mission at Mach 3 and altitudes of more than 70,000 feet. The design of the B-70 was radically different from that of contemporary aircraft. to design a

Equally important,

new

materials for the B-70's skin had to be developed

withstand the high temperature generated by sustained high-speed flight. In short, the B-70 program had to contend with several formida-

to

ble

development challenges. r

1

93]

Flying Blind

At the same time, the B-70 program was guided by a highly concurrent procurement strategy. Its schedule was compressed because the Air Force was anxious to deploy the B-70 in the early 1960s. Source selection was based entirely on paper studies. Competition was halted after North American was selected to be the prime contractor on the project in late 1957. Detailed work on the bomber's military subsystems began at an early stage, and production preparations commenced long before the first prototypes were completed. Finally, funding levels were very high throughout the early stages of the program, before prototypes took to the

air.

two B-70S were built. Although we do not have a B-70 production program to evaluate, the outcomes of the program as a whole are nonetheless clear. It failed miserably on cost, schedule, and In the end, only

performance grounds.

Origins of the Nuclear-Powered Bomber Program

The

origins of the nuclear-powered

bomber program can be

traced to

when the AAF initiated a series of feasibility studies collectively known as the Nuclear Energy Propulsion for Aircraft (NEPA) project. A 1946,

combination of strategic, bureaucratic, and technological considerations seems to have led the AAF to take this step. First, there were legitimate strategic reasons for beginning to work on an intercontinental bomber in 1946. It was becoming increasingly obvious that the Soviet Union was an aggressive geostrategic adversary. A bomber mission from the continental United States to the Soviet Union and back might involve a flight of 12,000 miles or more, depending on the target. An intercontinental bomber would not have to be based overseas, nor would it be dependent on potentially vulnerable forward bases or tankers for refueling. The vulnerability of forward bases and the problems associated with aerial refueling led General LeMay, who was deputy chief of staff for research and development in 1946, to push hard for long-range bomber capabilities throughout the late 1940s. 1 The range of a nuclear-powered bomber, it was said, would be limited "only by sandwiches and coffee for the crew." 2 Such a weapon had obvious appeal to those who shared these strategic concerns. The AAF also had an abiding bureaucratic interest in building longSAC's operational plans relied heavily on forward bases well into the 1950s, in part because, even then, most of the bombers in the force structure were medium-range air1.

craft.

Carlton Ward Jr., President of Fairchild Engine and Airplane Corporation, quoted in W. Henry Lambright, Shooting Down the Nuclear Plane (Indianapolis: Bobbs-Merrill, 2.

J.

1967), p- 3

-

The Nuclear-Powered Bomber and

range bombers. As discussed

bombardment was part

and

parcel of

its

in

the

B-yo

Chapter

raison d'etre

and

2,

it

believed that strategic

that long-range

bombers were

organizational essence. In addition, operations involving forward bases or tankers were complicated. Aerial refueling, its

example, was a nuisance because it involved several kinds of aircraft, each with different speed, range, and altitude capabilities. Strategic bombing missions would be much simpler to plan if aerial refueling and forward bases could be eliminated from the equation. Furthermore, the for

procurement and operating costs of tanker fleets and the operating costs of forward bases were a burden. The AAF would not have to spend money on forward bases or tankers if it had a force structure equipped with intercontinental bombers. A nuclear-powered bomber therefore also had enormous appeal from a bureaucratic perspective. The idea of building a nuclear- powered bomber had been toyed with in technical circles ever since the first nuclear reactor

was

built in 1942.

This idea was not pursued during World War II, though, because virtually all of the country's atomic energy expertise was devoted to building the atomic bomb. But once the war was over, other atomic energy

and the nuclear-powered bomber interested some in the technical community. That said, it is important to note that interest in the nuclear-powered bomber was not triggered by a technological breakthrough. Indeed, no systematic work was done in this area until the NEPA program got under way. The AAF took the lead in setting up the NEPA program. 3 It contacted the nation's leading aircraft engine manufacturers and directed them to select one company to be, although the phrase was not yet in vogue, the prime contractor on the project. The consensus of opinion was that Fairchild should play this role, and so in May 1946 the AAF awarded projects could be pursued,

Fairchild a contract to investigate the feasibility of building nuclear-

powered

Ten other companies subsequently became subcontractors on the project. In 1948, an outside review of the program concluded that it would be possible to build a nuclear-powered aircraft, although it might take fifteen years and $1 billion to do so. The Joint Chiefs of Staff decided in early 1951 that there was a military requirement for such an aircraft, and with that the program moved into engineering development. Over the next few months, the Air Force got the newly renamed aircraft.

Based on Lambright, Shooting Down the Nuclear Plane, pp. 2-7; U.S. Comptroller General, Review of Manned Aircraft Nuclear Propulsion Program, Report to the Congress, B-146759, Feb. 1963, pp. 122-125; Aircraft Nuclear Propulsion Program, Report of the Joint Committee on Atomic Energy, 86th Cong., 1st sess.. Sept. 1959, pp. 3-4; Aircraft Nuclear Propulsion Program, Hearings before the Joint Committee on Atomic Energy, 86th Cong., 1st sess., July 1959, pp. 113-117; Herbert York, Race to Oblivion (New York: Simon and 3.

Schuster, 1970), pp. 60-74.

[195]

Flying Blind

(ANP) program off the ground by awarding a series of development contracts. Convair and Lockheed were to build suitable airframes; Convair would modify a B-36 for this purpose. General Electric and Pratt and Whitney were to develop nuclear propulsion Aircraft Nuclear Propulsion

systems. In November 1951, General Electric estimated that it could deliver a propulsion system by May 1956. General Electric was not alone in its optimism about the program. The Air Force estimated in April 1952 that the first flight of a nuclear-powered aircraft

would

take place in 1956

or 1937.

ANP

program had to solve several difficult technological problems if it was to meet this or any other timetable. First, a new kind of nuclear reactor would have to be designed; it would have to be very

The

small but powerful. Second,

new construction

materials capable of with-

standing the intense heat and radiation generated by a small, powerful reactor would have to be developed. In addition to being lightweight, these materials would have to be durable. Otherwise, an airborne

meltdown might take place. Third, shielding to protect radiation would have to be developed; land-based shielded by several feet of concrete

— hardly practical

for

the crew from reactors

an

were

aircraft.

surrounded the ANP program soon proved to be unfounded. Technical problems at General Electric pushed plans for a flight test back to at least 1958. Pratt and Whitney's program progressed even more slowly. These technical problems, in conjunction with the Eisenhower administration's commitment to fiscal restraint, prompted a major cutback in the ANP program in early 1953. In March, the Air Force Scientific Advisory Board, a group generally quite supportive of military research and development projects, recommended cutting the program's budget by 50 percent because of its dimming tech-

The optimism

that originally

nical prospects. In April, the National Security

canceling the program because

May

it

was not

Council recommended

essential to national security.

program would be eliminated entirely from the fiscal year 1954 budget. The program was to be scaled back so that it could continue with unspent funds from previous fiscal years. The project to modify a B-36 for nuclear propulsion was canceled By

1953

it

had been decided

that the

entirely. 4

Origins of the B-70 Program

The Air Force was not about

to

abandon the idea

cessor to the B-32. In the midst of the

4.

ANP

of building a suc-

review in 1953, Maj. Gen.

Comptroller General, Review, pp. 127-128.

[196]

The Nuclear- Powered Bomber and the B-yo

Thomas Power, SAC's vice commander, insisted that the requirement for a long-range bomber would "continue to be valid and urgent for the foreseeable future, and well life

beyond the expected normal operational

of the B-52 force." 5

General Power also stepped up the Air Force's ongoing campaign against ICBMs. The Air Force had long recognized that ballistic missiles

posed a potential threat to the case for bomber forces. If it could be proven that missiles were accurate and reliable enough to carry out strategic attacks, many policy makers would undoubtedly see unmanned systems as cost-effective alternatives to manpower-intensive bomber forces. To guard against this possibility, the Air Force spent little on ballistic missile development in the late 1940s and early 1950s. Ballistic missile programs consequently made little technical progress; they were unable to demonstrate high accuracy, for example. The Air Force then used these technical shortcomings as an excuse for not spending more money on ballistic missile development. It was a Catch-22: ballistic missile programs could not make a compelling case for more money unless they

made

make progress

significant technical progress, but they could not

they were underfunded. 6 The hydrogen bomb, however, which the United States first tested in November 1952, threatened to make accuracy less important and ICBMs more competitive as strategic systems. if

was not a coincidence, therefore, that SAC's vice commander argued in early 1953 that building an ICBM was not feasible and that, even if it was, ICBMs were not reliable or accurate enough to replace or even supplement manned bomber forces. He concluded, "Regardless of the It

missile program,

it

is

the opinion of this headquarters that the continued

advance in the art of manned flight to high altitudes and long ranges should be at all times a priority objective of the Air Force's development programs. 7 In late 1953, though, the only successor to the B-52 in the development pipeline was the floundering ANP program. Several strategic and bureaucratic considerations came together in mid-1954 to reorient Air Force bomber development. The underlying strategic concern was that improvements in Soviet air defenses would make it increasingly risky to rely on the B-52 in the years ahead. The Air Force had believed for some time that substantial improvements in speed and altitude capabilities would be needed to penetrate Soviet air defenses in the late 1950s and early 1960s. This view was reinforced by 5.

Maj. Gen.

Thomas

S.

Power, Vice Commander, SAC, Letter to

Dir. of Reqs.,

USAF, March 30, 1953, Files, HO/AFSC; cited hereafter as Power Letter. 6. This argument is developed in Edmund Beard, Developing the ICBM, (New Columbia University Press, 1976), chaps. 3-6. 7. Power Letter.

HQ/ York:

Flying Blind

the discouraging operational experiences of the Korean War, in

which

advanced Soviet interceptors were surprisingly effective against older American bombers. It was with this operational problem in mind that the Air Force pushed ahead with the supersonic B-58 in the early 1950s. The B-58, though, was never intended to be an intercontinental bomber, and its range estimates started to deteriorate in the first half of 1954. It was becoming painfully obvious that the B-58 would be highly dependent on either overseas or aerial refueling for long-range missions. Reliance on overseas bases was becoming increasingly risky, however. A major Rand Corporation study, briefed extensively throughout the U.S. defense community in 1953 and early 1954, demonstrated that overseas bases would become increasingly vulnerable to preemptive attack as the Soviet Union's long-range bomber capabilities improved. An expanded version of this study was published (at the classified level) in April 1954. The Soviet Union's ability to strike U.S. overseas bases and the United States itself was growing by leaps and bounds. First, the Soviet Union

hydrogen bomb in August 1953; a relatively small number of Soviet bombers, armed with high-yield bombs, could now inflict staggering amounts of damage on their targets. Second, the Soviet Union unveiled its first intercontinental bomber during the 1954 May Day celebration (see Chapter 4). The Air Force concluded that it was vital to have an intercontinental supersonic bomber capable of penetrating Soviet air defenses and destroying this increasingly dangerous offensive force. 8 The B-58 could not fill the bill, and its deteriorating range estimates eventually drove SAC to announce that it did not want the B-58 in the force structure at all. At the same time, the nuclear-powered bomber was not expected to be particularly fast, even if new life could be breathed back into the ANP program, an increasingly unlikely prospect. tested a prototype of a

Several bureaucratic considerations also led the Air Force to reorient its bomber development efforts in mid-1954. First, the Eisenhower administration decided in late 1953 that its defense policy would emphasize strategic striking

power.

It

believed that formidable strategic ca-

would help to deter a wide range of threats to national security and that they could do so more cost-effectively than large conventional forces. Strategic forces seemed to provide "more bang for the buck," which was an important consideration in an administration committed to holding the line on defense spending. This meant that bureaucratic competition within the Pentagon would intensify and focus increasingly pabilities

8.

Reqs.,

The Air

Force's targeting priorities are outlined in Maj.

HQ/USAF, "General Operational

bardment Weapon System, No.

82,"

Gen. George

E. Price, Dir. of

Req. for a Piloted Strategic Intercontinental

March

22, 1955,

HO/AFSC.

[198]

Bom-

The Nuclear-Powered Bomber and

on

the

B-yo

programs. It was consequently more important than ever for the Air Force to have a viable successor to the B-52 under development. Otherwise, the long-range bomber could lose its dominant role in the strategic

nuclear force structure to other weapon systems. One of the bomber's competitors was the Navy's nuclear-powered aircraft program, which got under way in May 1953. The Navy, of course, had been anxious to acquire a significant role in strategic nuclear

operations since the late 1940s.

such an

aircraft,

bardment Still,

it

Navy was

the

successful in developing

would undoubtedly be promoted

for strategic

bom-

missions. 9

the main threat to the

ICBM development had been it

If

manned bomber was

the

ICBM. Although

neglected by the Air Force for

experienced an abrupt turnaround

many years,

1953 and 1934. One main reason high-level civilian intervention in the program. in

turnaround was Trevor Gardner, the new special assistant to the secretary of the air force for research and development, was concerned that ICBM development was not getting the attention it deserved from the Air Force establishment. He assembled a committee of outside experts to look into the for this

situation, and, in their report of February 1954, they recommended a major reorganization of ICBM research and development. In particular,

they recommended establishing a separate organization within the Air Force that would be responsible for and a true sponsor of ballistic missile development. The Western Development Division was estab-

—

—

lished for this purpose on July 1, 1934. Many believed that, as a result, operational ICBMs might be ready within five or six years. 10

The Air

Force's worst-case scenario

was

that the

ANP program would

grind to a halt, which would shut the long-range bomber out of the strategic sweepstakes altogether. But even under the best of circumstances, the ANP program was not going to produce an operational system in the 1930s, when many important decisions about missions and resources would be made. Thus, the Air Force needed a fallback option to the ANP as well as near-term option to complement the ANP.

was with these

and bureaucratic considerations in mind that the Air Force reoriented its long-range bomber development efforts on July 21, 1934. The director of research and development at Air Force headquarters called for the establishment of two high-priority programs, It

strategic

not surprising that,

when

Navy defined

the operational requirements and nuclear-powered aircraft in early 1955, it stated that the system's primary function included attacking shore targets; see Comptroller General, Re9.

It is

development

characteristics for

view, pp. 129, 134. 10. Beard, Developing the Strategies,

the

its

ICBM, chap. 6. See also Robert L. Perry, System Development Rand Corporation Research Memorandum, RM-4853-PR, Aug. 1966, pp. 54-100.

Flying Blind

11 The be based solely on nuclear propulsion. first, known as the Weapon System (WS) 125-A program, was a modification of the ANP program. Its goal was to build a bomber with both a nuclear propulsion system and a chemical engine. The chemical engine

neither of which

would provide

was

to

a supersonic

dash capability as well as

a

backup

in case

the nuclear reactor failed. It was hoped that nuclear propulsion technology would eventually progress to the point where the bomber could be powered entirely by the nuclear reactor. In the meantime, however, it

would be

a hybrid system.

It

was

felt

that this

was

way

the quickest

to

some way incorporated nuclear prosaw the nuclear-powered bomber as the ultimate successor to the B-52. The second, parallel program was explicitly designed to serve as a hedge against problems in the WS-125-A program. This bomber would have a relatively traditional chemical pro-

build an operational system that in pulsion. Many in the Air Force still

pulsion system and an airframe optimized for this engine alone. This

"chemical bomber,"

known

as WS-110-A, evolved into the B-70. Thus,

marked the beginning of the B-70 program. The WS-110-A program was not triggered by a technological breakthrough or an anticipated breakthrough. If there was a technological component to the WS-110-A decision, it was the emergence of several technical problems in the ANP and B-58 programs. The performance requirements for the WS-110-A program were set far beyond the state of the art, and there was a great deal of pessimism within the Air Force

July 1954

about the chances of meeting these requirements. Similarly, the economic well-being of the aerospace industry did not play a significant role in the Air Force's decision to initiate the B-70 program. The Air Force's concerns were both more global and more parochial.

Performance Requirements WS-110-A and WS-125-A programs were issued in late 1954 and early 1955. Both programs were to meet the same demanding requirements. For starters, the new bombers needed an unrefueled radius of 4,600 miles and a refueled radius of 6,325 miles, or ranges of roughly 10,000 miles unrefueled and 13,000 miles refueled. Since SAC wanted to reduce its reliance on tankers, an unrefueled radius of 6,325 miles was said to be "highly desirable." In addition, a cruising speed of Mach 0.9 was required. The top speed was increased as the requirements were formalized. At first, the aircraft were Formal operational requirements

11.

Brig.

Gen.

B. S. Kelsey,

for Devel., “Strategic

Acting

for the

Dir. of Res.

Weapon System

and Devel., Office of Dep. Chief of Staff CG, ARDC, July 21, 1954, Files,

Studies," Letter to

HO/AFSC. [

200

]

The Nuclear-Pozvered Bomber and the B-yo

to

have the

"maximum speed

possible." Later, they

were

to

have the

"maximum

supersonic speed possible." Still later, it was stated that the aircraft would have to be able to dash through a target zone of 1,0002,300 miles at a speed of Mach 2.0 or more. 12 To put these requirements in perspective, the Air Force's other Mach 2 bomber, the B-58, was required to have an unrefueled radius of only 2,650 miles. Furthermore, since heavily defended air bases were considered to be

important targets, the new bombers had to be capable of dropping half of their bombs within 1,500 feet of their targets. Alternatively, they could carry air-to-surface missiles, provided that these missiles could meet the

same accuracy standards. 13 Although bombing accuracy had been a centerpiece of strategic bombardment doctrine since the 1930s, it did not receive this much attention in the B-47, B-52, and B-58 requirements statements. One would think that, with the emergence of extremely powerful hydrogen bombs in the mid-1950s, pinpoint accuracy would not be so critical. But the Air Force was concerned that the emergence of ICBMs would undermine its case for the manned bomber, and consequently it was careful to specify a requirement that early ICBMs were unlikely to meet.

The

ability of the

new bombers

to operate at various altitudes

was

also

be investigated, but the Air Force insisted that low-altitude capabilities should not be incorporated at the expense of high-altitude capabilities. At a minimum, the new bombers were to fly at 60,000 feet, and an altitude in excess of 75,000 feet was desired. 14 The performance requirements for the chemical bomber and the nuclear-powered bomber were the same. The nuclear-powered bomber, though, would have virtually unlimited range, assuming that its nuclear propulsion system worked as it was supposed to. The chemical bomber, on the other hand, would have to contend with a brutal trade-off beto

12. "General Operational Req. No. 38 for Intercontinental Bombardment Weapon System," Oct. 14, 1954; Col. E. N. Ljunggren, Dir. of Weap. Systems, HQ/ARDC, "System Req. No. 22," Feb. 18, 1955; Brig. Gen. B. S. Kelsey, Dep. Dir. of Res. and Devel., Dep. Chief of Staff for Devel., HQ/USAF, "Characteristics of a Chemically Powered Strategic Bombardment Weapon System," Letter to CG, ARDC, Feb. 10, 1955; Col. E. N. Ljunggren, Dir. of Weap. Systems, HQ/ARDC, "System Req. No. 22 (Revised)," April 15, 1955; all in Files, HO/AFSC. See also "Program Development Plan," May 26, 1953, and "Resume of SAC Briefings: Advanced Strategic Weapon Systems"; both in Files, HO/ASD. See also Price, "General Operational Req. No. 82." 13. The bomber's standard payload was to be a 10,000-pound thermonuclear bomb, although alternative payloads of up to 40,000 pounds were to be considered. The air-tosurface missile's payload on a 300-mile flight was to be a 3,000-pound warhead, although alternative payloads of up to 20,000 pounds were to be looked into. See Price, "General

Operational Req. No. 82." 14. Ljunggren, "System Req. No. 22." Contractors were also supposed to consider equipping their bombers with chaff, decoys, electronic jammers, and air-to-air missiles as defensive measures.

Flying Blind

tween speed and range

— so

for

it

the requirements

were especially

chal-

lenging.

WS-110-A requirements were set far beyond the state of the art, as the Air Force itself acknowledged. The ARDC believed that these performance requirements had been "calculated to push the state of the art for chemical propulsion to the utmost" and that they To put

it

simply, the

"could not be completely fulfilled until major state of the art improvements had been accomplished." 13 The Wright Air Development Center's (WADC) assessment was, if anything, even more pessimistic. It concluded that the WS-no-A would not meet its mission requirements within the planned development period.

16

B-70

The Air Force moved quickly

to initiate a

Procurement Strategy

design competition for the

were invited to participate in the competition, but only Boeing and North American were interested enough to submit proposals. Of the others, some were unwilling to invest their time and engineering resources in a program that might be WS-110-A. In April 1955

six contractors

summarily canceled if the nuclear-powered bomber ever got off the ground. Others were already involved in the ANP program. Most simply was not possible to build a chemically powered bomber that could meet the Air 17 Force's demanding performance requirements. The Air Force was anxious, for both strategic and bureaucratic reasons, to deploy a chemical bomber as soon as possible. It was forced to acknowledge in early 1955, though, that its goal of having one wing of

some

important,

of the contractors believed that

it

chemical bombers operational in 1963 was unattainable and that a target 18 Theoretically, date of mid-1964 was the best that could be hoped for. the Air Force could have expedited this process by lowering its perfor-

15.

ARDC/HO,

16.

Bomb.

History of the Aire. Div., Dir. of

HO/AFSC;

ARDC,

Jan. to June 1955, p. 227. Weap. Sys. Ops., WADC, "System 110A,"

May

26, 1955,

WADC

110-A Report. 17. Col. Samuel C. Phillips et al., "Proposed Contractors for Conventionally Powered Strategic Bomber (System 110A)/' Memorandum, April 26, 1955, and Brig. Gen. Howell M. Estes, Jr., Chairman, Source Selection Board, Letter to Dir. of Procurement, AMC, July 13, 1955; both in Files, HO/AFSC. See also WADC 110-A Report. 18. Lt. Gen. D. L. Putt, Dep. Chief of Staff for Devel., "Development of a Chemically Powered Strategic Bomber," Letter to CG, ARDC, March 14, 1955, cited hereafter as Putt Letter; Brig. Gen. Howell M. Estes, Jr., Dir. of Weap. Sys. Ops., WADC, "Chemically Powered Strategic Bomber Program," Letter to CG, ARDC, April 7, 1955 Col. E. N. Files,

cited hereafter as

'

HQ/ ARDC, "Amendment to ARDC System Req. No. 22," HO/AFSC. See also Ljunggrem, System Req. No. 22;" WADC

Ljunggren, Dir. of Sys. Plans, Oct. 11, 1955; 110-A Report.

all

in Files,

[2021

The Nuclear- Powered Bomber and the B-yo

mance requirements; There

is

no evidence

it

could have traded performance for

that

it

even entertained

this idea.

Instead, the Air Force tried to accelerate the chemical

by employing within

ARDC

a great deal of

in

it.

bomber program

WADC,

the division

responsible for aeronautical systems development, ex-

plicitly stated that

methods"

concurrency

availability.

it

wanted

in this case.

19

avoid using "prototype manufacturing This attitude was reflected in the acto

WADC

quisition plans of 1955, plans that closely resembled those planned, first of years earlier for the B-58 program.

WADC

made all,

a

few

to build

twenty WS-110-A aircraft for testing purposes. These aircraft would not be true developmental prototypes similar to those utilized in the B-47 and B-52 programs, but production prototypes built on hard production tooling. The fact that twenty aircraft were to be built at this stage indicates that they were not true developmental prototypes; it would have been foolhardy to build that many prototypes if the point of the exercise was to refine and modify the design of the aircraft. In fact, WADC did not expect to see substantial change between the prototypes and the actual production aircraft. It also would have been foolhardy to build this many prototypes if the production decision really was up in the air. A decision to build this many prototypes was actually a de facto production decision, even though it was made virtually at the outset of the overall program. Since the only prototypes to be built were these twenty, the WS-110-A development and production programs had to overlap. An entire production line had to be set up before any aircraft could be built, and these prototypes had to be built before the development work associated with flight testing could even begin. According to WADC's original plan, approximately two years of preliminary development work would take place starting in mid-1955, which would lead to a design competition and a source selection. Preproduction work (including procurement of long-lead tooling and components) would begin in mid-1957, and the first flight of the aircraft would take place at the end of i960. It was expected that the twenty production prototypes would be engaged in a testing program that

would

mid-1963. This was a plan for a highly concurrent program. Paper studies were to be relied on for the design, source selection, and de facto production decisions. Competition was to be halted at the end of the preliminary last until

design phase. Perhaps most important, there was a substantial amount of overlap between development and production activities. WADC expected that three and one-half years of production work would take place before the first flight of the aircraft and that six years of production 19.

WADC

110-A Report.

1^03)

Flying Blind

work would take place before the main part of the development program was concluded. WADC stated explicitly that its intent was to expedite development and compress the acquisition schedule as much as possible. At the same time, it admitted that it would be extremely difficult to

meet

all

performance requirements given

opment schedule. 20 The WADC plan used

the

weapon system

this

compressed devel-

concept: subsystems

would

be developed concurrently with the aircraft itself and completely integrated into the design of the system from the beginning. This approach was deemed necessary because production would begin so early in the overall scheme of things. WADC acknowledged that subsystem schedules were tight and that the program would probably require interim bombing, navigation, and other critical subsystems. 21 Plans were even

made at have

the outset of the

to take place

even

program if

for extensive retrofitting,

the entire

which would

program proceeded smoothly.

The Air Force recognized at the outset that the chemical bomber's costs would be a problem. 22 It finessed this problem by charging the cost of several critical subsystems to other accounts. The WS-125-A program, for example, was to be charged with the cost of developing the bombing and navigation, defensive armament, decoy missile, and air-to-surface missile systems that were to be used on both bombers. Engine development and a high-energy fuels program were funded separately. Even so, it was expected that the highly concurrent WS-110-A program would require an enormous financial commitment early on. WADC estimated in 1955 that the budget for the program would skyrocket to over $300 million per year once preproduction work began and that over $800 million would be spent by the end of fiscal year i960, when the first flight of the aircraft would take place. The program would therefore acquire a great deal of momentum long before much was known about production costs or the performance capabilities of the aircraft itself. The Air Force's best estimate in early 1955 was that twenty production prototypes and a wing of thirty operational bombers could be built for approximately $1.6 billion. 23 One of the most remarkable aspects of this early planning exercise is that it was made in an informational vacuum. The performance requirements for the new bombers were extremely demanding, and enormous technological uncertainties permeated both the WS-110-A and the its

likely

20.

Ibid.

Also, Brig. Gen. Victor R. Haugen, Dir. of Labs.,

Defensive Subsystems

for

WADC,

Weapon System 110A," Memorandum, Aug.

“Offensive and 26, 1955, Files,

HO/AFSC. 21.

WADC

HQ/USAF, June 21, 1955, File, HO/AFSC. the ARDC, July to Dec. 1954, vol. 1, p. 249.

presentation to

22.

ARDC/HO,

23.

WADC

110-

History of Report.

A

(204]

The Nuclear-Powered Bomber ami the B-yo

WS-125-A programs. The nuclear-powered bomber program required an exotic propulsion system that was still in the basic research stage. The chemical bomber was, in early 1955, still a total unknown. Nothing was known about its design; the contractors who would participate in the design competition had not even been determined. It is not at all clear, therefore, how WADC could assume that the aircraft design could be frozen and that development and production activities could be compressed. Nor is it clear, under these circumstances, how WADC could make elaborate projections about the program's schedule and cost.

Return of the Nuclear-Powered Bomber The nuclear-powered bomber program, which was cut back in early 1953 and in a holding pattern until mid-1954, was revitalized by the WS-125-A approach. Although the Air Force had enthusiastically supported the ANP program throughout the early 1950s, it was even more

new

hybrid bomber.

expected the WS-125-A to have both extended range and supersonic speed while avoiding complete reliance on nuclear propulsion. The Air Force was not alone in its renewed optimism: a Department of Defense study group concluded in March 1955 that the WS-125-A approach made the goal of nuclear-powered flight more attainable. As a result, funding for the program increased dramatically, from $44.7 million in fiscal year 1955 to $91.2 million in 1956 and $180 million in 1957. 24 Important developments were taking place in both Moscow and excited about the prospects for the

It

Washington while the WS-110-A and WS-125-A programs were getting under way. The July 1955 Aviation Day celebration in Moscow featured an unexpected and alarmingly high number of long-range Bison bombers, which led some in the U.S. intelligence community to conclude that the United States faced an impending bomber gap. In Washington, Eisenhower declared in September that the ICBM development program was to receive the highest national priority. Special procedures were put into effect one month later to bypass established organizational channels in the Air Force, thereby cutting red tape and ensuring that the ICBM program would receive all the political support and financial resources it needed. 25 As 1955 progressed, the Air Force was confronted by what it perceived to be increasingly formidable strategic and bureaucratic threats. It became even more important from the Air Force's perspective to build a viable successor to the B-52 as soon as possible. 24.

Comptroller General, Review, pp. 110, 133.

Beard, Developing the ICBM, pp. 189-194; Perry, System Development Strategies, pp. 54-100. 25.

[205]

Flying Blind

B-70 Design Studies

Design work on the WS-110-A did not begin until November 1955, when Boeing and North American were awarded development contracts. 26 The main design problem was that the Air Force wanted both the range of the B-52 and the speed of the B-58 in its new bomber. The contractors did not have the luxury of trading one for the other. It soon became clear the new bomber could not perform its entire mission at supersonic speeds. Instead, it would have to fly a split mission cruising most of the way to its target at high subsonic speeds and then dashing through the combat zone supersonically. Even so, the chemical

—

bomber was required

Mach

which was substantially better than the B-52's top speed, and its top speed had to at least equal that of the B-58. Boeing and North American also had to ensure intercontinental range. Taken together, according to ARDC, these requirements placed extreme demands on the contractors. 27 The designs developed at both companies in early 1956 featured “floating wing tips" disposable fuel tanks attached to the ends of the wings. The bomber was to use the fuel in the wing tip tanks as it cruised to the combat zone, then discard them. The pilot would be left with an aircraft sleek enough to dash through enemy airspace at Mach 2. The main problem with these designs was that each of the floating wing tips was roughly the size of a B-47; the tanks actually needed wings of their own. Boeing's submission to the Air Force, Model 724-15, had a gross take-off weight of 710,000 pounds, including the two floating wing tips, each of which weighed 180,000 pounds (see Figure 6.) 28 North American's proposed bomber was projected to weigh 750,000 pounds. These bombers would therefore be 40-50 percent larger than the B-52. As LeMay, still the SAC commander, commented in reviewing one of these to cruise at

0.9,

—

designs, ''This

is

not an airplane. This

is

a three plane formation." 29

This approach had several obvious drawbacks.

work; although the idea of using floating wing 26.

March

Lt.

HO/AFSC;

might not

for the Record,

Memorandum. Weapon System 110A,"

cited hereafter as Beary

ARDC/HO, "The B-70 Story: The ARDC, Jan. to June 1959, vol. 2, p. 11. 27.

it

tanks dated back to at

AMC, Memorandum

Col. Kermit E. Beary, Chief, 110 Section,

11, 1957, Files,

tip

First,

Evolution of

History of the

Boeing's design studies are outlined in "110A Weapon System Trade-Off Studies," Document D2-1421, Jan. 21, 1957, and "History: Boeing Weapon System 110A," Document D2-2371, Dec. 26, 1957; both in Boeing archives. 29. LeMay quoted in Ed Rees, The Manned Missile (New York: Duell, Sloan and Pearce, i960), p. 99; see also The B-70 Program, Report of the Preparedness Investigating Subcommittee, Senate Armed Services Committee, 86th Cong., 2d sess., July i960, p. 4; Maj. Gen. 28.

Jr., Asst. Dep. Commander for Weap. Sys., Memorandum to Dep. Weap. Commander for Sys., ARDC, Dec. 31, 1936, Files, HO/AFSC; cited hereafter as Estes Memorandum. See also Beary Memorandum.

Howell M.

Estes,

The Nuclear-Powered Bomber and the B-yo

Figure

least 1949,

it

6.

Boeing floating wing

had never been put

tip

into practice

proposal

on

this scale.

Second,

it

would involve enormous production and operating expenses, especially if it proved to be impossible to land a bomber with tanks still attached. If this proved to be the case, the tanks would have to be discarded after every

full-scale test flight, training mission, or airborne alert

[207]

— the costs

Flying Blind

would be prohibitive. Finally, SAC would have to build new runways to accommodate these unusually wide bombers. Again, the associated costs were considerable. The only alternative to floating wing tip tanks was high-energy fuel, but exotic fuels involved more developmental risks than wing tip tanks. In addition, their development and production costs appeared to be nearly as imposing as the costs associated with disposable tanks. Given that the WS-110-A had to use either floating wing tip tanks or highenergy fuels, the contractors concluded in May 1956 that "it is impossible with today's state of the art to meet the general operational requirements." 30 Boeing and North American formally submitted floating wing tip proposals when the design competition ended in May. The Air Force reviewed these proposals and abruptly terminated work on floating wing but not because of technical risks or protip designs in October 1956 gram costs. Instead, it emphasized that the projected performance of the

—

bombers was inadequate. The bombers did not offer "a sufficiently marked advance in strategic capability over programs already in being" to warrant further development. In particular, they did not meet the Air 31 The contractors were sent back to their Force's range requirements. drawing boards and ordered to engage in "study, exploratory research, and component development directed toward maximum state of the art improvements." 32 If the state of the art would not permit the contractors to meet the Air Force's performance requirements, then the contractors would have to advance the state of the art itself: the Air Force's basic intent during the existing study program, therefore, to have the contractors re-examine every possible design in order to It is

attempt to find a better solution to the range versus penetration problem which will permit a greater range while still maintaining a high inherent Our interest does not lie in refinement of your penetration capability. previously submitted designs, but rather in the full investigation of every .

.

.

possible means, through complete redesign

desired performance parameters

if

necessary, of achieving the

33 .

The Air Force was simply adamant about meeting its performance requirements, and it was unwilling to sacrifice penetration speed for range, or vice versa. Finally, as if these requirements were not demanding enough, the Air Force also directed the contractors to explore the 30.

E.

Brig.

Gen. H.

W. Rawlings,

S. Jones,

May

2,

Dep.

Dir. of

1936, Files,

Procurement and Prod.,

Memorandum

to

HO/AFSC.

Memorandum.

31.

Estes

32.

Chief of

33.

Estes

Staff,

USAF, Teletype

Memorandum.

to

CG, ARDC, Oct.

18, 1956, Files,

HO/AFSC.

Gen.

The Nuclear-Pozvereci Bomber and the B-yo possibility of

having the bomber

fly

supersonically throughout

its

entire

mission.

Demise of the Nuclear-Powered Bomber

The fortunes of the nuclear-powered bomber sank considerably in 1955 and 1956, just as the WS-110-A program sputtered through its first round of proposals. A June 1955 Air Force Scientific Advisory Board report

estimated

development schedule for the nuclearpowered bomber was optimistic by three to five years. 34 In addition, several technical problems continued to haunt the program. The projected weight of the propulsion system and the shielding had not come down, so the bomber's top speed and the range of its dash zone would be limited. It was certain, moreover, that these problems could not be overcome in time to deploy a nuclear-powered bomber by 1963 or 1964. The prospects for the nuclear-powered bomber dimmed so much by 1956 that even LeMay, a vocal supporter of the program, suggested that the WS-125-A no longer be one of the Air Force's top priorities. In his view, the B-52, the WS-110-A, and even the Atlas ICBM were more important. 35 The Scientific Advisory Board concluded in October 1936 that, "while the present state of the reactor art is encouraging, it does not conclusively demonstrate that a useful vehicle can be built." 36 It believed that the program was geared too much toward building a weapon system at an early date, at the expense of basic research and development. The assistant secretary of defense for engineering recommended to the secretary of defense in October that the program be reoriented toward basic research. This would be the best way of achieving "the radical improvement necessary to make a nuclear-powered aircraft system which is a major advance over a chemically powered aircraft system." The assistant secretary of defense for research and developthat

the

ment came to a particularly harsh conclusion about the nuclear-powered bomber program: "It appears now that the probability of attaining the high performance desired period, is almost nil." 37

in the 125-A, in the originally established

time

The WS-125-A program was canceled in December 1956, and the ANP program was reoriented toward basic research at the same time. The WS-123-A program failed for two basic reasons. First, it was extraordinarily ambitious technologically, which even the Air Force ultimately

36.

Comptroller General, Review, pp. 134-135. Beary Memorandum. Quoted in Comptroller General, Review, pp. 138-139.

37.

Ibid., p. 139.

34. 35.

Flying Blind

recognized. The program's inherent technical problems were compounded by its extremely demanding performance requirements, such as the supersonic dash requirement. Second, the WS-125-A program had a great deal of concurrency built into it. A sequential program the key technological problems and endeavored to build a prototype of the reactor at the earliest possible date. The WS-125-A program, however, was geared toward the development of

would have focused on

the total

weapon system, with

predictable effects:

power [were] thus diverted from what

scientists

"Money and man-

argued were the basic

questions to such matters as the testing of tires for their resistance to ionizing radiation. From the scientist's point of view, this [was] placing 38 the cart before the horse, i.e., the plane before the reactor engine." Herbert York, the director of defense research and engineering, believed that this strategy set the program back five years. A different development strategy a more sequential one would have saved a great deal

—

—

of time

and money,

39 in his opinion.

Although the

ANP

program

itself

was canceled once and for all in 1961, the nuclear-powered bomber was never a serious contender for production and deployment after 1956. In the end, over $1 billion was spent on various nuclear-powered aircraft programs, even sputtered along for another five years before

though

it

40 a nuclear-powered aircraft never flew.

The B-70 Development Program and one that would dramatically improve 1956 made a major breakthrough supersonic aircraft performance. They found that the highly compressed air of the supersonic shock wave created by the forward fuselage of an aircraft could be carefully directed under the aircraft's wings, increasing its lift and improving its overall aerodynamic efficiency. It was estimated

NACA scientists engaged

in general aeronautical research in 1955

—

compression lift could carry about 30 percent of an aircraft's weight, thus improving aircraft performance, especially range. These that this

findings were

first

published in a

NACA

research

memorandum

in

by the Air Force to improve bomber performance. North American and Boeing design engineers came across the NACA study in literature searches conducted in late 1956. Early 1957 was consequently a time of considerable excitement for those involved in the WS-110-A program. It was estimated that a com-

March

1956. Driven

38.

Lambright, Shooting Dozen

39.

York, Race

to Oblivion,

the Nuclear Plane

,

p. 32.

pp. 63, 67-68.

Lambright, 40. pp. 66-74; Comptroller General, Review, pp. 110-112, 142-177; Shooting Dozen the Nuclear Plane, pp. 1, 12-32. Ibid.,

[

210

]

The Nuclear-Powered Bomber and the B-yo

bomber would have an unrefueled range of at least 8,000 miles, which eliminated the need for floating wing tip fuel tanks. In addition, it was estimated that such a bomber could fly at 70,000— 75,000 feet at a speed of Mach 3, substantially better than the Mach 2.3 projected for the floating wing tip designs. Finally, the contractors discovered that it was not only possible but more efficient to fly a compression lift bomber at Mach 3 (where the compression lift effects materialized) for its entire mission. As a result, the split mission was discarded in favor of an all-supersonic flight profile. The only drawback was that, once the aircraft was optimized for a high-altitude, all-supersonic mission, it would be inefficient at any other altitude or speed. It suddenly seemed that North American and Boeing could meet most of the Air Force's performance requirements. This was possible, though, pression

lift

only because of a totally unforeseen breakthrough in basic scientific research that materialized long after the requirements themselves were established. 41

The compression vast

lift

number of major

of performance.

The

bomber would, however, have

to consolidate a

technological advances in order to realize this level first technological hurdle was figuring out how to

maximize the compression lift effect itself. No aircraft had ever been designed with this aerodynamic principle in mind so the new designs would inevitably be radical departures from existing conventions. Another problem involved the skin of the aircraft. Aluminum, which was normally used in aircraft construction, could not withstand the heat generated by sustained Mach 3 flight. Stainless steel and titanium were better suited to high-temperature conditions, but they were too heavy to be used throughout the aircraft. The search for heat-resistant but lightweight materials led North American to develop stainless steel honeycomb panels for its new bomber. These panels were different from those in the B-58, where aluminum and conventional glues could be used. In this case, stainless steel and a more heat-resistant brazing process were needed. North American's honeycomb steel sandwich consisted of two face plates, between which a honeycombed stainless steel foil and a thin sheet of silver brazing foil were put. Each panel then had to be heated until the silver foil melted and bonded the honeycomb to the face plates. Obviously, manufacturing all the individual panels needed on a large aircraft such as the WS-110-A was enormously complicated and expensive. The apparently simple problem of deciding on construction materials for the WS-110-A therefore involved some fairly spectacular advances in materials development and fabrication. These seem41.

See Senate

ry/' PP- *7' 33-

[2u]

Armed

Services Committee, B-yo Program,

p. 12;

ARDC/HO,

"B-70 Sto-

Flying Blind

and manufacturing problems would ultimately 42 prove to be critical to the program as a whole. Major advances were also needed in propulsion technology. Existing engines, after all, had never been required to generate sustained Mach 3 speeds. The engine for the Mach 3 bomber would be confronted by efficiency and cooling requirements far more demanding than any established for previous programs. New materials would consequently have

ingly

mundane

materials

be incorporated into the engine. One Air Force study correctly observed that development of the bomber's advanced turbojet engine was 43 It is important to note that, once this engine was a high-risk venture. optimized for sustained Mach 3 flight, it would be relatively inefficient at off-design conditions. The engine inlets for the new bomber had to be highly advanced as well. They had to slow the air down to subsonic to

speeds before delivering it to the engines, and they had to do so efficiently. This had never been done under these challenging flight conditions. Again, a seemingly mundane aspect of the program involved critical technological advances and risks. Finally, the WS-110-A's bombing and navigation systems involved major technological advances. Meeting a 1,500-foot accuracy requirement in a bomber that flew at 70,000-75,000 feet at speeds in excess of

Mach

3

was

difficult, to

say the

least. Existing

bombing and navigation

requirements under these flight conditions. In fact, bombing and navigation system devel44 opers had never been confronted by such demands until 1957Many other advances had to be made before an operational Mach 3 bomber could become a reality. One study of the program observed that

systems could not come close

to these accuracy

involved "hundreds of state of the art advances ... in scores of component systems" precisely because it involved a "quantum jump in aircraft performance. The new bomber was, quite simply, too advanced 4r> for the industry which had to build it." Contrary to what the Air Force had assumed when it first outlined its acquisition plan for the WS-110-A, the design of its new bomber did

it

undergo major changes during development. The aircraft that North American and Boeing proposed in July 1957 were radically different

42.

See Rees, Manned

Missile, pp. 21-24;

ARDC/HO,

"B-70 Story,"

p.

30-35. See also lain

State of the Art Improver," Flight International, June 25, 1964, pp. 10551061; July 2, 1964, pp. 18-24. See also a series of reports from North American: "XB-70 Structure and Materials," May 11, 1964; "Structural Design of the XB-70," no date; "HistoPike, "B-70:

The

XB-70 Development Program," no date; "The XB-70 and the Future of Aeronautics," May 11, 1964; "Laboratory for Progress," no date; all in Files, HO, Air Force

ry of the

Museum. 43. ARDC/HO,

"B-70 Story," pp. 37-54. 44. Rees, Manned Missile, pp. 28-29. 45. Ibid., pp. viii, 12, 111-114.

[212]

The Nuclear-Powered Bomber and the B-yo

from the floating wing tip proposals of the previous year. Given that these proposals were based on many unproven and exotic technologies, there was no reason to assume that this round of proposals was any more immune to design changes than the last. Both proposed bombers featured a large, highly swept delta wing, but North American's design was the more adventurous of the two. Its wings were swept more than 65 degrees, and the wing tips could be folded down to trap the compression lift under the body of the aircraft. Stainless steel honeycomb panels were to be used extensively throughout the aircraft, and all six engines were packed into the aft fuselage, on the assumption that drag would be minimized if the engines were out of the way of the supersonic shock wave. Boeing's design was more conservative. The wings were swept only 57 degrees, and the wing tips were clipped off for greater structural strength. Boeing placed the aircraft's six engines out on the wings, as it had traditionally done, and it relied heavily on a titanium alloy as the major structural material. Each of the proposed bombers had a gross weight of around 500,000 pounds. 46 The Air Force was anxious to get beyond paper proposals, and it made a source selection decision quickly.

It

announced

December 1957 design competition and in

that

North American was the winner of the that construction of the B-70 would begin immediately. This decision reflected the Air Force's clear preference for proposals that promised outstanding operational performance over those that minimized developmental risks. In this case, though. North American was able to promise better speed and altitude capabilities because it optimized its aircraft for the high-altitude, high-speed flight regime the Air Force had emphasized since the early days of the program. Boeing's design, on the other hand, promised more operational flexibility. In retrospect, the Air Force would have been wise to opt for more flexibility and fewer risks, but it wanted its new bomber to fly as fast and as high as possible. 47

New

Timetables

,

Nezv Performance Requirements

The program's schedule had already slipped by over a year when Boeing and North American were sent back to their drawing boards in late 1956. The target date for one operational wing of bombers slipped ARDC/HO,

46.

"B-70 Story," pp. 30-35. See also the Boeing design studies cited in note

28.

was

North American needed the business more than Boeing. North Ameriprogram had been canceled in July 1957, and its F-108 program trouble (and would be canceled in September 1959). Boeing, on the other hand, was

In addition.

47.

can's

Navaho

in

cruise missile

midst of the B-52's long production run. Therefore, as in the case of the B-58, this is not a good test of the relative importance of aircraft performance and contractor need in Air Force source selection decisions. in the

Flying Blind

from July 1964 to late 1965. Still, the Air Force wanted to deploy a new bomber as soon as possible; it was concerned that the B-70 would be swamped by the expanding Atlas ICBM program if it was unable to promise operational capability before the mid-1960s. The competition with the Atlas became even more acute in the aftermath of the Soviet Union's ICBM test in August 1957 and Sputnik launch in October. Many believed that the United States faced an impending missile gap and that U.S. ICBM programs needed to be accelerated. The Navy's Polaris SLBM program, another strong contender in the growing competition for 48 funding, was making great strides by this time as well. The Air Force decided in March 1958 to accelerate the B-70's schedule by eighteen months, making it more competitive with the Atlas and enabling it to move along quickly while there was widespread concern in Congress over the state of the U.S. strategic forces in general. According to the Air Force's accelerated plan, one wing of B-70S (including thirty aircraft in service plus fifteen aircraft for testing) would be ready by May 1964. The B-70 would be fielded sooner under this plan than it would have been under WADC's 1955 plan, even though the program had been delayed by over one year in the interim. The highly compressed schedule for the program which WADC had insisted in 1955 could not be compressed any more without imposing high risks on the program would indeed be compressed still more. Development and production activities would overlap to an even greater degree, even though the B-70 was more technologically ambitious in 1958 than it had

—

—

been

earlier.

The new plan, like the old, relied on production prototypes. It therefore had to assume that the design of this exotic aircraft could be frozen and that production work could begin at once. Subsystem integration would also begin immediately. And competition, of course, had already been eliminated from the program. Finally, a substantial financial commitment would have to be made to the B-70 right away in order to begin gearing up for production. In all of these respects, the accelerated plan for the B-70 involved even more concurrency than the concurrent plan that had been proposed in early 1955. It was estimated that accelerating the B-70 program would add $165 million to its cost and that the first 45 aircraft would cost around $2. 5 billion. The Air Force also estimated that 250 B-70S could be built for $6.4

Desmond

49 billion, or $25.6 million each.

and Force Levels (Berkeley: University of California Press, 1980), pp. 3-58, 88-104; Harvey M. Sapolsky, The Polaris System Development (Cambridge: Harvard University Press, 1972), pp. 34-41. 49. Lt. Gen. Samuel E. Anderson, CG, ARDC, Letter to Chief of Staff, USAF, Feb. 7, 1958, and Maj. Gen. Jacob E. Smart, Asst. Vice Chief of Staff, Letter to CGs, ARDC, AMC, March 19, 1958; both in Files, HO/AFSC. See also Rees, Manned Missile, pp. 35, 108-111; ARDC/HO, "B-70 Story," pp. 19-20; Ball, Politics and Force Levels, pp. 214-215. 48.

See

Ball, Politics

[214]

The Nuclear-Powered Bomber and the B-yo

The Air Force

On

March.

also issued a revised set of performance requirements in

the whole, these requirements were substantially

manding than those

more de-

1954-55 because they codified the gains made A cruising speed and top speed of Mach 3 were required, and a top speed in excess of Mach 3 was desired. At a minimum, the B-70 was to cruise at 60,000 feet, although 75,000 feet was desired. in i 95 ^ _ 57

set in

-

Although the Air Force grudgingly permitted one aerial refueling over the course of an 8,740-mile mission, the minimum requirement, it hoped that a 12,650-mile mission could be flown without refueling. In addition,

bombing accuracy was reemphasized, thus differentiating the B-70 from the ICBM, which the Air Force assumed would be relatively inaccurate and usable only against large, soft targets. Finally, the B-70 had to be able to get away from its base quickly; this also differentiated it from the ICBM, which was immobile and therefore relatively vulnerable. As demanding as these requirements were, they were at least consistent with the direction the program had been going since 1954. 50 The Air Force also introduced a low-altitude requirement at this time, insisting that the B-70 needed operational flexibility. A capability of flying at 500 feet or less was required, and a top speed of Mach 0.9 at low altitudes was desired. Concern about the B-70's ability to penetrate advanced air defenses was also reflected in the emphasis placed on minimizing the radar reflectivity of the aircraft. The problem, of course, was that the B-70 had already been optimized for high-speed, high-altitude flight. In fact,

ble aircraft because

was designed

it

was

a particularly inflexi-

one special aerodynamic phenomenon, the compression lift generated by supersonic shock waves at high altitudes. Its low-altitude capabilities had been intentionally sacrificed. The B-70 would have been terribly inefficient at low altitudes because its large wing would have generated enormous drag whenever it plowed through denser layers of atmosphere. Its wing was designed to trap air beneath it for additional lift, rather than knife through the atmosphere smoothly. As a result, the B-yo's top speed and range would have suffered dramatically at low altitudes. In addition, the crew would have had a very rough ride at low altitudes because the aircraft had a long fuselage in addition to a large wing. This situation would have had serious practical implications for crew efficiency, navigation, and bombing accuracy. Finally, there was no way to minimize the it

radar reflectivity of a large, stainless steel panels.

The

flat

to exploit

delta-winged

sort of flexibility

covered with the Air Force sought in 1958 aircraft

had been designed out of the program. Maj. Gen. James Ferguson, Dir. of Reqs., "General Operational Requirement for a Supersonic Piloted Strategic Bombardment Weapon System, No. 82," March 7, 1958, Files, 50.

HO/AFSC. [215]

Flying Blind

The Air Force was obviously high-altitude air defenses in 1958.

It

bombing

trying to anticipate advances in Soviet

when

had been aware

it

issued

its

low-altitude requirements

of the potential importance of low-altitude

since the early 1950s; a low-altitude requirement

was issued

for

the B-58 in 1951 and then neglected as the B-58 was optimized for the traditional, high-altitude mission. A low-altitude capability was men-

tioned in the WS-110-A requirements statements in 1954-55, but the Air Force was unwilling at that time to sacrifice high-altitude capability for operational flexibility. For many years, the air defense threat to high-

bombers was simply too ambiguous

altitude

to

overpower the Air

Force's deeply held faith in the effectiveness of high-altitude operations.

was no conclusive evidence that operations were becoming dangerous. It was not until the

Throughout most high-altitude

of the 1950s, there

defense threat to existing high-altitude bombers became undeniable. But, even then, the Air Force believed that high-altitude attacks could be conducted successfully. What was needed, it felt, was a bomber with better high-altitude capabilities than late

1950s that the Soviet

air

where the B-70 came in. In selecting North American over Boeing in December 1957, the Air Force opted for the contractor that emphasized high-altitude performance over operational the B-52 or B-58, which

flexibility.

When

is

51

March 1958 that low-altitude capabilities might be needed, it refused to abandon the high-altitude mission and indeed still refused to compromise high-altitude performance. Instead, it imposed both high-altitude and low-altitude requirements on the B-70. But, since a complete redesign would have been needed to meet both sets of requirements, which would have delayed the program by a year or more, the Air Force once again allowed its lowthe Air Force finally admitted in

altitude requirements to

fall

by the wayside.

Technical

The B-70 program encountered 1959. Constructing the aircraft's

proved

a

and

Political

Problems

wide variety of problems

many

stainless steel

in 1958

and

honeycomb panels

New

manufacturing methods were needed for fabrication of the panels, and then the panels had to be welded together with the utmost precision. New welding techniques had to be perfected before this assembly process could get under way. North American ultimately developed a suitable electron-beam welder, but this kind of welding had never been done on this scale before; since it had to be done in a vacuum, it had only been used 51.

I

to

be especially

difficult, for

several reasons.

analyze the role of the low-altitude mission in Air Force thinking in Chapter

7.

The Nuclear-Powered Bomber and the B-yo

on very small parts

prior to this time. Welding the

huge wing panels of the B-70 together was more difficult to say the least, and it required a major advance in manufacturing technology. The B-70, moreover, required honeycomb panels on 75 percent of its surface. 52 It soon became clear that the B-70 could not be developed and produced on a compressed schedule. It also became clear that it would be an exorbitantly expensive aircraft, precisely because it involved new and time-consuming manufacturing methods. Unfortunately for the Air Force, these problems materialized just as interest intensified in deploying American ICBMs and SLBMs quickly and just as these missile programs began to produce concrete results. The range of the Atlas, for example, began to approach intercontinental distances in 1958. Deployment plans for the Atlas had already been set, and in October 1958 the Eisenhower administration decided to deploy 400 second-generation, solid-fuel Minuteman ICBMs as well. The ICBM program began to consume more and more of what the administration hoped would be a fairly flat budget. The amount of money spent on ICBM development and production increased from $3 million in

year 1954 to $2.15 billion in 1958. In addition, the administration was also planning to deploy SLBMcarrying submarines. 53 As a result, the B-70 program was decelerated in

October and November 1938, and on it. Its budget for fiscal year 1959,

fiscal

budget ceilings were imposed for example, was set at $221 million. One effect of this decision was that the target date for deploying the first operational wing of bombers was set back from May 1964 to August 1965; another was that total program costs would increase by $160 million because production activities would be stretched out. 54 Even more ominous developments took place in 1959, when two important programs related to the B-70 were canceled. Each cancellation dealt a severe blow to the B-70's prospects. First the high-energy fuel program was canceled in August. High-energy fuels had been counted on to extend the range of the B-70 by as much as 33 percent; eliminating strict

them obviously affected the B-yo's competitiveness as a strategic system. A more severe blow came in September when the high-speed, highaltitude F-108 interceptor was canceled. The B-70 and F-108 programs had been closely coordinated, and many of the costs of common subsystems had been shared. More important, the costs of developing advanced engines, fuel systems, and escape capsules which were common to both aircraft had been covered completely by F-108 funds. Cancellation of the F-108 meant that $180 million would be added to the

—

—

Rees, Manned Missile, pp. 111-115. 53. Beard, Developing the ICBM, p. 206; Ball, Politics and Force Levels, pp. 44-45. 54. Chief of Staff, USAF, Teletype to CG, ARDC, Nov. 27, 1958, Files, HO/AFSC. Also, ARDC/HO, "B-70 Story," p. 20. 52.

Flying Blind

B-70's costs at a time

when

it

was already

struggling to stay cost-effec-

tive. 55

These problems soon proved

be the least of the Air Force's worries, because the B-70 itself was virtually canceled in December 1959 by the Eisenhower administration. The Air Force was directed to eliminate everything from the program not necessary for prototype testing. Only to

was to be a stripped-down experimental aircraft, not a full-fledged weapon system. Development of the bombing and navigation system and the various defensive systems was canceled, as were all production plans. Funds for the program were reduced from $330 million to $150 million for fiscal year i960 and from $444 million to $75 million for 1961. The first flight of the B-70 was consequently pushed back from January to December 1962. 56 Given that the B-70 was extraordinarily ambitious technologically, the one prototype was

to

be

idea of demonstrating

built,

its

and

it

basic performance capabilities in an austere

prototype program was certainly not a bad one. But by late 1959 it was impossible to turn the B-70 program into an austere effort, simply because an enormous amount of money, variously estimated at $300 mil-

—

$500 million, had already been spent much on subsystem integration and production activities. 57 The Air Force had spent about one-tenth of this amount on the B-52 during the first five years of its development program. The decision to redirect the B-70 program in December 1959 was, in any event, not an attempt by the administration to adjust the program's lion

to

procurement strategy

to technological realities; rather,

it

was

the politi-

compromise the administration and the Air Force reached after the administration tried to kill the program outright. The Eisenhower administration was skeptical about the B-70's operational prospects, and it was very concerned about production costs, which appeared to be astrocal

now

be deployed as legitimate alternatives to the manned bomber, and they certainly appeared to be more cost-effective than the B-70. Keeping the defense budget under control in the face of mounting congressional pressure to expand U.S. strategic nuclear forces was one of the administration's overriding objectives at this point, and canceling the B-70 seemed to be the best way of doing this. The Air Force, for its part, naturally wanted to develop the B-70 as a nomical. Ballistic missiles could

55.

HEF

H. A. Storms, Chief Engineer, North American Aviation, “B-70 B-70," Letter to

CG, AMC,

April

1,

1959,

and B-70 Project

56.

Senate

Office, "B-70

28, 1958; both in Files, HO/AFSC. See also ARDC/HO, Armed Services Committee, B-70 Program, p. 51. Armed Services Committee, B-70 Program, pp. iii, 41, 51.

Summary," Nov. 21-22; Senate

Weapon System:

Manned

All

HEF Program

"B-70 Story," pp.

Dennis M. Sherman, The National Security Act" (PhD. Dissertation, University of Wisconsin at Madison, 1978), p. 201. 57.

Ibid., p. 1; York, Race to Oblivion, p. 55; Rees,

Missile, p. 119;

The Nuclear- Powered Bomber and the B-yo

weapon system, and it lobbied intensively to do enormous investment already made in the B-70 worked to full-fledged

Force

so.

The

the Air

advantage, vivifying the argument that it would be wasteful to walk away from the money already sunk into the program. It also created a formidable domestic constituency for the program; by this time, s

thousands of jobs were

literally

sional

and

at stake,

and i960 was both

a congres-

a presidential election year.

The Denouement of the B-70 The Eisenhower administration completely reversed itself on the B-70 before i960 was out, adding $60 million to the program's fiscal year 1961 budget of $75 million in July and another $20 million in August, just as the election campaign heated up. These increases were enough to add a second prototype to the program and, more important, to restore the program's weapon system status: the two prototypes were to be equipped with the military subsystems needed to demonstrate the full combat capabilities of the bomber rather than just the aerodynamic castripped-down aircraft. Finally, the administration added $110 million to the program's budget on October 31, just a few days before the election itself. Up to twelve prototypes could be built under this expanded program, and plans were made to build 200 B-70S. The Eisenhower administration's fiscal year 1962 budget, which was sent to Congress in January 1961, contained $358 million for the B-70, enough to lay the groundwork for a production program. 58 Democratic party candidates charged that the administration was blatantly trying to manipulate the election by buttressing its record on national defense and buying the votes of aerospace workers in California, where North American's main B-70 facility was located. It certainly was true that the administration had come under attack during the i960 campaign for what was said to be its neglect of the strategic balance. Critics charged that the administration had put its commitment to a balanced federal budget before national security. Restoring the B-70, it was said, was simply a way of belatedly addressing this general charge. pabilities of the

Many

believed that reviving the B-70 program

move because

was

was

a particularly blatant

key state in the electoral college. Eisenhower's vice president, Richard Nixon, ultimately carried Califorpolitical

California

a

nia in the election. It

was

also true that the administration

was under intense pressure

pp. 55-57; Sherman, National Security Act, p. 201; Ball, Politics and Force Levels, p. 215. Also, Chronology of Events Related to the Advanced Manned Strategic Aircraft Program, 1954-1965, no date. Files, Simpson Historical Research Center. 58.

York, Race

I219]

to Oblivion,

Flying Blind

North American B-70

(U. S. Air Force)

from both the Air Force and Congress to revitalize the B-70 program. The Air Force was careful to squelch an early i960 proposal to build an enhanced version of the B-58 in lieu of the B-70. Obviously, it did not want to jeopardize its chances of resurrecting the B-70. The Air Force was also careful to develop an array of contingency plans for reviving the B-70 production program, and it is not surprising that each of these plans was based on the highly concurrent framework that had guided its earlier procurement plans. 59 The Air Force had important allies in Congress, including Stuart Symington, a former secretary of the air force who sat on the Senate Armed Services Committee, and Lyndon Johnson, the senate majority leader. Both had endorsed the B-70, and both ran for the Democratic party's presidential nomination in i960. In addition, the B-70 had the support of Congress at large, reflected in the authorization of $265 million for the program in fiscal year i960, $190 million more than the administration had originally planned to spend. Gen. Thomas

Power, CG, SAC, Letter to Gen.

Thomas

D. White, Chief of Staff, USAF, March 31, i960, and Gen. Curtis E. LeMay, Vice Chief of Staff, Letter to Gen. Thomas S. Power, CG, SAC, April 29, i960; both in Files, HO/AFSC. See also Wright Air Development Division/HO, History of Wright Air Development Division, Jan. to June i960, 59.

vol. 2, p. 41.

S.

The Nuclear-Powered Bomber and the B-yo

North American B-70 (U.

It is

S.

Air Force)

clear that the decision to revive the B-70 in i960

any breakthrough

in the B-70

program

itself.

In fact,

was not based on the program con-

tinued to experience the kinds of technical problems that had plagued it since 1958. Nor did the program's cost estimates suddenly turn around. If anything, the projected cost of the bomber continued to climb as more

was found out about what

a B-70

production program would

entail.

developments did not bode particularly well for the B-70 program either. The Soviet Union shot down a high-altitude U-2 reconnaissance airplane in Soviet airspace in May i960, which suggested, if anything, that Soviet high-altitude air defenses were even more formidable than expected. Finally, the Minuteman and Polaris programs had no major setbacks in i960. Instead, American ICBM and SLBM deployments continued, and the U.S. strategic force structure evolved into a Strategic

triad. In short, the objective case for the

unfolded; indeed,

it

B-70 did not get stronger as i960 probably deteriorated. Accordingly, the Defense

Department's own Weapons Systems Evaluation Group, which conducted a comprehensive review of the strategic balance and the U.S. strategic force structure in

its

December i960

report,

opposed deploying

the B-70. 60

Although Senator John Kennedy had wholeheartedly endorsed the B-70 program during the i960 campaign, he decided just a few months after taking over the Oval Office that it should be scaled back. He announced on March 28, 1961, that only three experimental prototypes would be built and that they would not be developed as full-fledged weapon systems. A small number of bombing and navigation subsystems would be developed to assess the state of the art, though. 60.

Ball, Politics

|22lj

and Force

Levels, pp. 34-35*

Flying Blind

Kennedy also announced that the fiscal year 1962 budget for the program would be reduced from the $358 million Eisenhower had requested to $220 million and that the total cost of the program would be limited to $1.3 billion. It has been estimated, though, that by early 1961 61 almost $800 million had already been spent on the B-70. The Kennedy administration's case against deploying the B-70 was

multifaceted.

It

argued that the strategic rationale

for building a

bomber

with the B-7 o's particular capabilities was not convincing. Survivability, flexibility, and penetration capability were said to be the three key attributes of strategic systems, and, according to administration analyses, the B-70 did not measure up well on any of these counts. Supporters of

the B-70 argued that the B-70 could be launched on warning, which was something that could not be done with ballistic missiles. The administration countered that the problem with the B-70 was that it had to be

was vulnerable on the ground. In addition, the administration believed that the B-70 would have a difficult time seeking out and destroying mobile targets or targets the ballistic missile force missed, because it flew too high and too fast to conduct real-time reconnaissance or highly accurate attacks. The B-70, it was argued, was not a flexible system, and it would not be able to do anything that an ICBM could not do. The administration also argued that the B-70 would not be an effective penetrator because it was limited to high-altitude missions, which were becoming increasingly problematic. In the administration's view, the B-70 did not add enough calaunched on warning;

like

any bomber,

pability to the strategic forces that

it

were already scheduled

to

be de-

very high costs. 62 The administration estimated that a force of 200-250 B-70S would cost at least $10 billion above the $1.3 billion set aside for the prototypes and

ployed

to justify its

that the cost of a B-70 production

than $15

billion.

program could be considerably more

These estimates, moreover, did not include the procure-

ment costs of the fleet of tankers the B-70 would require, nor the operating and maintenance cost of the bomber force as a whole, which would be enormous due to the complexity of the B-70. Finally, it was estimated that the B-70 would not be available in operational numbers until 1968, 214-224; Sherman, National Security Act, pp. 213-255; York, Race to Oblivion, pp. 57-59; John F. Tarpey, "The RS-70 Controversy," (Ph.D. dissertation, Stanford University, 1972), pp. 65-71; Robert E. Hunter, "The Politics of U.S. Defense 1963: Manned Bombers versus Missiles," in Morton H. Halperin and Arnold Ranter, eds., Readings in American Foreign Policy (Boston: Little, Brown, 1973), pp. 191-202; Thomas S. Coffey, Iron Eagle: The Turbulent Life of General Curtis LeMay (New York: Crown, 1986), pp. 365-366, 37361.

See

Ibid., pp.

380, 405-411, 420, 432.

William W. Kaufman, The McNamara Strategy (New York: Harper and Row, 1964), pp. 220-229; Alain C. Enthoven and K. Wayne Smith, How Much Is Enough? (New York: Harper and Row, 1971), pp. 34, 44, 172, 243-251; Robert S. McNamara, Essence of Security: Reflections in Office (New York: Harper and Row, 1968), pp. 91-92; Ball, Politics and Force Lei'els, 62.

pp. 215-224.

[222]

The Nuclear- Powered Bomber and the B-yo

when

was expected to have a large ICBM and SLBM The B-70 therefore did not appear to be a particu-

the United States

force already in place. larly timely strategic

system, although

it

did appear to be a particularly

cost-ineffective one. Still,

tion

the Air Force did not give

program. The

new

up

the fight for a full-scale B-70 produc-

Air Force chief of staff

was

Curtis LeMay,

who

LeMay had been one of the staunchest and most forceful advocates of the manned bomber for many years, and he moved quickly to revitalize the case for the B-70. The new bomber moved

into the job in June 1961.

would be needed, he argued, to attack hardened targets (where accuracy was important), mobile targets (which could not be programmed into the guidance systems of ballistic missiles), and targets missed by ballistic missile attacks. According to this line of thinking, the B-70 was needed in the force structure to complement the ICBM. In July a revised requirements statement for the B-70 which emphasized this "reconnaissancestrike" mission was issued, and the Air Force subsequently referred to the B-70 as the RS-70.

It

maintained that the RS-70 could be operational

by 1967-70. Secretary of Defense Robert

McNamara was

skeptical about

the

mission was so challenging and because it would involve even more technological advances than the original B-70. To carry out the search and destroy missions envisaged by the Air Force, the RS-70 would require, as McNamara saw it, an extremely high resoluRS-70's prospects because

its

system and other highly advanced reconnaissance sensors, robust on-board data processing capabilities, sophisticated display systems, and a retargetable air-to-surface missile. The bomber's electronic systems would have to be advanced enough to recognize targets and tion radar

while the aircraft was flying at 70,000 feet at a speed of 30 miles a minute. Reconnaissance, moreover, would have to be conducted quickly if the RS-70 was to strike before it moved out of air-tosurface missile range. This was a daunting task, as McNamara exassess

damage

to targets

plained: Picture the RS-70 flying at 70,000 feet and moving at 2,000 miles per hour. The proposed mission would require the gathering of radar reconnaissance

data on the presence of

new

targets

—

— or known targets which may not have

been destroyed or neutralized and the prompt processing and analysis of these data in flight. The proposed radar, moving with the aircraft at 2,000 miles per hour would be seeing new area at a rate of 100,000 square miles per hour or 750 million square feet per second. We cannot state today with any assurance that satisfactory equipment to perform this processing and alone by display function in an RS-70 can be made operational by 1970/ human in1967, on the basis of any known technology, or whether the

^

63 terpretation job required of the operator can ever be done.

63.

McNamara quoted

in

Kaufmann, McNamara

Strategy, p. 224.

Flying Blind

McNamara

many key

had yet to be developed and were beyond present scientific knowledge. He concluded that the RS-70 was far from ready for production or even fullscale weapon system development. What was called for, he believed, was a prototype program that would determine whether the B-70 could stressed that

electronic subsystems

successfully perform this reconnaissance-strike mission. 64

and

McNamara

young systems analysts took the lead in the campaign against the RS-70. The Air Force, for its part, was especially bitter that McNamara's office would presume to make what the Air Force considhis

band

of

ered to be military judgments about the operational effectiveness of the bomber. The Air Force, LeMay in particular, resented what it saw as

domain. 65 The RS-70 proposal probably would not have gotten far if Congress had not become deeply involved in the issue. In 1961, Congress appropriated $180 million more for the B-70 than the administration had requested in its fiscal year 1962 budget request. The administration subsequently impounded these funds rather than spend them on the program. In early 1962, as the debate on the fiscal year 1963 budget shaped up. Congress once again took steps to provide more money for the program than the administration was willing to spend on it. The House Armed Services Committee, however, went one step farther in March 1962 when it “directed, ordered, mandated, and required" the administration to spend the full amount appropriated for the RS-70. The administration was not inclined to expand the scope of the B-70 prototype program, and it certainly was not inclined to allow Congress civilian intrusion into its professional

defense policy. A constitutional crisis loomed over the practice of impounding funds until Kennedy took the chairman of the House committee, Carl Vinson, for a walk in the White House Rose Garden and negotiated a compromise. Kennedy promised to restudy the B-70 in exchange for a withdrawal of the condictate the course

frontational

and content

wording

of

its

that “directed" the administration to

spend no

less

for the program. 66

than the full amount appropriated An RS-70 study group was consequently established at the end of March and it reported to McNamara in June. McNamara rejected all the development and production options suggested by the Air Force. From late 1962 through 1965, the Air Force routinely submitted proposals that called for production and deployment of the RS-70. McNamara rou-

Ibid., pp. 223-225, 227. Curtis E. LeMay, Mission with LeMay (Garden City: Doubleday, 65. 1965), pp. 6-12; Coffey, Iron Eagle, pp. 365-366, 373-380, 405-411.

64.

Katherine Johnsen, "Vinson Refuses to Drop Fight for B-70," Aviation Week and Space March 19, 1962, p. 28; Theodore C. Sorensen, Kennedy (New York: Harper and Row, 1965), pp. 347-349. 66.

Technology,

The Nuclcar-Pozvereci Bomber and the B-yo

them. In the spring of 1963 for example, the issue was brought up as the fiscal year 1964 budget deliberations got under way in Congress. McNamara was committed to confining the B-70 program to the prototype stage, and in the end he was successful despite the Air Force's concerted efforts. The fact that the Joint Chiefs of Staff did not support the Air Force's position on the RS-70 until after the March 1962 showdown strengthened McNamara's position. Congress ultimately acquiesced on the issue, and the administration continued to impound whatever excess funds were appropriated for the program. 67 Chronic technical problems also strengthened the hand of those who wanted to confine the B-70 program to exploratory, developmental, prototype activities. For example, development of the delicate engine inlet control system proceeded so slowly that the original subcontractor on the project had to be fired and North American had to take over. Another major problem was eliminating the fuel leaks that permeated the aircraft's honeycomb panels. Finally, there were problems welding the B-70's many honeycomb panels into wing assemblies. When the time came to join the wings to the fuselage, North American discovered a 3/4tinely dismissed

wing and the wing stub on the fuselage. The net effect of these and many other technical problems was that the first flight of the B-70 prototype was delayed again and again as the cost of the program escalated. In spite of these setbacks, the Air Force was still predicting in March 1962 that the B-70 would fly that December. By May 1963, its estimate had slipped to early 1964, and the date of the B-70's inaugural flight continued to slip thereafter. 68 The first flight of the B-70 did not take place until September 1964. Inevitably, the costs of the program grew while its development problems were sorted out. The Air Force and McNamara agreed in March 1964 to eliminate the third prototype from the program in order to keep the costs of the program to the $1.5 billion inch mismatch between the

from $1.3 billion) established for it. Supporters of the program naturally blamed these problems and delays on insufficient funding, but the fact is that technical problems prevented North American from getting even a prototype into the air and one could hardly argue that $1.5 billion was insufficient for a two-aircraft prototype program. Even though North American and the Air Force had a strong vested interest in getting a B-70 into the air as soon as possible, to demonstrate ceiling (up

—

67. Congress also acquiesced to the administration's decision to discontinue B-52 and B-58 production after October 1962, as well as to its cancellation of the Skybolt air-tosurface missile in December 1962. The evolution of the views of the Joint Chiefs of Staff on the B-70 and RS-70 is discussed in Ball, Politics and Force Levels, pp. 215, 219; Coffey, Iron

Eagle, p. 377; Kaufmann, McNamara Strategy, pp. 220-221. 68. "Save the RS-70 Drive," Aviation Week and Space Technology,

"Vinson Refuses

to

Drop

Fight."

May

20, 1963;

Johnsen,

Flying Blind

was

was simply impossible

do so. The program's technical problems were simply too stubborn. These problems, moreover, revealed how complicated and expensive a B-70 or RS-70 production program would be. Although the B-70 ultimately demonstrated the ability to fly for brief periods at Mach 3 and 74,000 feet, it never flew a simulated intercontinental mission. In fact, its longest flight was only 2,700 miles. The B-70 test program, which included only 129 flights, concluded in February that

it

1969.

a viable aircraft,

it

to

69

Conclusions Although the Air Force blamed the B-yo's problems on funding constraints and civilian opposition, it was given more than enough time and money to demonstrate that the B-70 was a viable system. The program was politically vulnerable because it was a developmental disaster, and it was a developmental disaster because of Air Force decisions about performance objectives and procurement strategies in the 1950s. These decisions determined the B-yo's fate. From the beginning, the B-70 had to meet demanding performance requirements, and these requirements became more demanding as time went by. Warnings from the ARDC about the technological implications of setting performance targets far beyond the state of the art were ignored by the Air Force hierarchy. The program consequently floundered until a totally unforeseen technological breakthrough enabled North American to project Mach 3 capabilities for its proposed bomber. The decision to build a Mach 3 bomber had profound implications for the program as a whole; for example, stainless steel panels had to be used

aluminum because aluminum could not withstand the heat generated at sustained speeds in excess of approximately Mach 2.3. The Air Force, though, did not carefully analyze the relative costs and benefits of Mach 2.3 and Mach 3 systems. Would a Mach 3 system be needed to penetrate Soviet air defenses in the 1960s? If so, why was the Air Force so slow to realize this? Would Mach 2.3 be sufficient? How much additional penetration capability would a Mach 3 system provide? Was it worth the added cost and development risks? The Air Force did not instead of

carefully address these issues.

Mach

3 bomber, cost-effective.

it

The B-70

it

might be possible

did not follow that such a

A Mach

69.

Although

flight test

3 capability,

program

is

to build a

bomber was necessary

or

moreover, might not even be desir-

reviewed

in

ASD/HO,

History of the Aeronautical

Systems Division, Jan. to Dec. 1965, vol. i-B, pp. 155-166.

[226]

The Nuclear-Powered Bomber and

able

if it

stood in the

way

the

B-yo

of a low-altitude capability,

if it

involved too

many

technological complications, or if it cost too much. Ultimately, it did all of these things. The B-70 program proved to be extraordinarily

ambitious technologically. It featured a new and radically different system design in addition to major advances in materials and virtually every major component and subsystem. The B-70 program therefore belongs in the first position on the nine-point scale of technological ambitiousness outlined in Chapter 1 (see Table 1).

At the same time, the Air Force was committed to a highly concurrent procurement strategy in the B-70 program. It compressed the development and production schedule even more in 1958, although the aircraft had become even more technologically adventurous when it was redesigned in 1957. In the end, the B-70 was a classic example of a concurrent procurement program. It exhibited all eight of the attributes of concurrency outlined in Chapter 1 (see Table 2). Why did the Air Force do this? First, it was important from a bureaucratic perspective for the Air Force to deploy a successor to the B-52 at the earliest possible date. The Air Force was deeply committed to the continuation of the manned bomber, and it was concerned about the emergence of thriving ICBM and SLBM programs in the mid-1950s. It assumed that compressing the B-70's development and production schedules would lead to early delivery, and that the B-70 would consequently be more competitive with the ballistic missile programs. At the same time, the Air Force knew that pushing the B-70 toward production would increase the program's momentum, endow it with a bigger domestic constituency, and therefore

make

it

more

difficult to cancel. Sec-

had to deploy a successor to the B-52 in 1963 or 1964 to stay one step ahead of Soviet air defenses. It never specified in any of its requirements statements, though, what kinds of Soviet air defenses were anticipated and why these particular requirements were necessary and sufficient with respect to this threat. Even so, given that Soviet air defenses would undoubtedly improve as time went by, there was at least a plausible strategic rationale for pushing the development of a new strategic bomber. Third, many in the Air Force genuinely believed that concurrency would work. Although it certainly was risky to employ concurrency in a technologically adventurous program, it was at least conceivable that things could work out as planned. What was bureaucratically necessary and strategically justifiable, thereond, the Air Force argued that

it

was seen as possible. The decision to employ concurrency compounded the problems were bound to appear under the best of circumstances. As it turned fore,

that

out,

the development and production programs could not be compressed

because

many

of the B-yo's

problems were simply

intractable, at least in

Flying Blind

More important, concurrency was actually counterproductive in several respects. The effort that went into subsystem development diverted attention and resources away from the program's central development problem: building the airframe. The lack of a prothe

short

run.

totype testing program played into the hands of those eager to cancel the B-70. Given that the balance of forces between those who supported the program and those who opposed it was a delicate one throughout

most of the program's history, a successful flight test in the early 1960s might have tipped the scales in favor of production. Accelerating the production schedule in 1958 also added to the program's total costs, which played into the hands of those critical of the B-70's cost-effectiveness. The program's delicate production plans were especially vulnerable to disruptions caused by accelerations, decelerations, and funding shifts. A great deal of time and money was lost whenever the program's schedule or funding profile was changed. Finally, concurrency involved a heavy financial investment in the early stages of the program, which meant that more money would be lost if the program was eventually canceled. Concurrency made it more likely that the program would fail and more costly when it did fail. The B-70 program ultimately failed miserably on the cost, schedule, and performance grounds. The Air Force estimated in 1955 that twenty test aircraft and one wing of thirty bombers could be built for $1.6

would be $32 million. In 1958 the Air Force estimated that fifteen test aircraft and one wing of bombers would cost $2.5 billion; unit costs would be $55 million. In the early 1960s, it estimated that fifteen test aircraft and one wing of bombers would cost around $5 billion; unit costs had grown to an estimated $110 million. Ironically, the two prototypes that were built cost around $1.3 billion roughly what the Air Force had estimated fifty B-70S would cost at the outset of the program. The Air Force's estimates about production costs also grew enormously as time went by. It estimated in 1938 that 230 B-70S could be built for $6.4 billion, or $24.6 million each. Most estimates in the early 1960s predicted that a fleet of 200 B-70S would cost between $10 billion and $13 billion, or $30-873 million each. At these prices and there was no guarantee that it could be built for $30-873 million the B-70 would have been substantially more expensive on a unit basis than either the B-32 or the B-38, costing four to five times as much as the B-38 and roughly ten times as much as the B-32. The B-70 would have been an exorbitantly expensive aircraft simply from a production standpoint, and its operating and maintebillion; that

is,

the unit cost for the

first

fifty

aircraft

—

— —

nance costs promised to be equally exorbitant. It is no wonder, therefore, that both the Eisenhower and Kennedy administrations decided to cancel the B-70 production program, even [228]

The Nuclear- Powered Bomber and the B-yo

though hundreds of millions of dollars had already been invested in it. The B-70 proved to be very difficult to cancel, though, despite its monumental technical problems and terrifying cost estimates, because it had built

up

a great deal of

momentum

as well as a powerful domestic

constituency.

The

B-70's schedule slipped repeatedly over the years.

WADC's

According

to

development plan, the first flight of the WS-110-A would take place at the end of i960. The first flight of the B-70 ultimately took place in September 1964. Almost nine years elapsed from the signing of the first major development contracts to the first flight of the aircraft. By the way of comparison, the B-47 program took only three years to go from contract award to first flight, the B-52 and B-58 programs took less than six years. There is no doubt that technical problems with the airframe were primarily responsible for the delay in the B-70 flight test program. There is also no doubt that these problems could have been resolved more quickly if the Air Force had concentrated more of the program's resources on them. Contrary to what the Air Force continually argued, it was not a lack of funding that held the program back but the way the available funding was spent. These funding decisions were driven by the Air Force's concurrent procurement strategy. Finally, the B-70 never did demonstrate the performance capabilities that so captivated the Air Force in 1957 and 1958. Even if it had, it would not have solved the Air Force's penetration problems in the 1960s because it could not fly efficiently at low altitudes and because it had a large and distinctive radar cross section. Although the Air Force toyed with the idea of low-altitude operations off and on throughout the 1950s, it did not require low-altitude capabilities in the B-70 until 1958, when it was too late to modify the design of the aircraft. The Air Force subsequently came to the realization that, if low-altitude capabilities were needed, they would have to be integrated into the designs of its bombers in the earliest stages of the acquisition process. This realization led to the development of the B-i. the

original

[7]

Low -Altitude

The

program was

Penetration: The B-i

departure for the Air Force in three respects. First, unlike its predecessors, the B-i was designed to perform lowaltitude missions. Although the Air Force thought about incorporating low-altitude capabilities into the B-58 and B-70, it did not follow through on its tentative plans to do so. It was not until 1961, when the first design studies for what eventually became the B-i were commissioned, B-i

a

committed itself to building a bomber capable of flying at very low altitudes. Even then, the Air Force did not abandon its traditional, high-altitude orientation. Instead, it developed a bomber capable of both high- and low-altitude operations. Second, the B-i was not particularly adventurous aerodynamicallv, and on the whole it was that the Air Force

not as ambitious technologically as

gram was guided by

a

its

predecessors. Third, the B-i pro-

mixed procurement

strategy,

which came about

in

complicated fashion. This strategy was adjusted several times in the late 1960s and early 1970s, when the program finally moved into full-scale development. In the end, the B-Ts procurement strategy involved both concurrent and sequential elements. In assessing the B-i program's

held view that cost

it

was an

acquisition debacle.

and schedule problems,

many

outcomes,

it

was not

take issue with the widely

I

Although the program had

the unmitigated disaster that

made it out to be. It experienced relatively few technical problems, and some of its cost problems can be traced to funding instability and flawed projections about the rate of inflation. One could in the 1970s

-

even say

that the B-i

was

a qualified success,

which

is

what one would

[230]

Low-Altitude Penetration

expect from a program with

mixed procurement

fairly

modest technological objectives and

a

strategy.

Emergence of the Low-Altitude Mission Air Force thinking about the relative merits of high- and low-altitude bombing operations had a long and complicated history. Starting in the

American strategic bombardment doctrine emphasized daylight operations, which meant that bombers would have to contend with air defenses when they were at their best. It was obviously advantageous to fly beyond the range of enemy antiaircraft guns and, if possible, air defense interceptors, which led to a requirement for high-flying bombers. Because air defenses were constantly improving, each new generation of bombers was designed to fly higher than its predecessor. In the early 1930s, the B-10 flew at 21,000 feet, a marked improvement over the biplanes of the 1920s. The B-17 and B-29 flew at 30,000 and 32,000 feet, respectively, which allowed them to attack targets defended by antiaircraft guns in World War II. Postwar bombers were designed to have even more impressive altitude capabilities. The B-36 and B-47 had service ceilings above 40,000 feet, and early models of the B-52 flew at 47,000 feet. When it set performance requirements for bombers in the 1950s, the Air Force stuck with the formula it had followed for years, a 1930s,

formula

being deeply ingrained in Air Force thinking, had worked well in the past. It consequently required the B-58 to fly at 50,000 feet and the B-70 at 60,000 feet 75,000 feet if possible. that, in addition to

—

Although American air force leaders emphasized high-altitude operations throughout the 1930s, 1940s, and 1950s, low-level options were not completely overlooked. Most of the fire-bombing raids on Japan were flown at low altitudes. The AAF switched to nighttime attacks in the Pacific theater because Japanese air defenses were fairly effective during the day but weak at night (see Chapter 2). In fact, Japanese air defenses were so weak at night that American bombers could approach at 7,000 feet, allowing them to carry heavy bomb loads and deliver their payloads with great accuracy. The Air Force also toyed with the idea of low-level of the Aircraft

1947 that a high-altitude penetration, but nothing came of this study.

1.

in

Heavy Bombardment Committee, USAF

Heairy Bombardment, Nov. 7, 1947,

war.

and Weapons Board study be carried out to compare low- and

The Heavy Bombardment Committee

recommended

tactics after the

p. 6.

Aircraft

1

and Weapons Board, Report on

Flying Blind

The 1951 and 1954-55 requirements statements for the B-58 and B-70 called for low-altitude capabilities. The B-58 was to perform at low and high altitudes because Soviet high-altitude air defenses were expected to improve significantly by 1957. The Air Force discovered, however, that the only

was

way

to sacrifice

unwilling to do.

impressive low-altitude capabilities high-altitude performance, which it was

to give the B-58

some

of

its

subsequently issued separate sets of requirements for high-altitude and low-altitude bombers, instead of building one bomber capable of performing both missions, as it originally intended. The B-58 program focused on the Air Force's high-altitude requirements, and the It

idea of building a low-altitude

Chapter

5).

The

bomber was eventually dropped

B-70's first sets of requirements noted that the

(see

bomber's

perform low-altitude missions should be investigated, but the Air Force insisted that low-altitude capabilities should not be pursued at the expense of the high-altitude mission. The B-70 was subsequently optimized for high-altitude flight (see Chapter 6). There were several reasons why the Air Force failed to incorporate ability to

low-altitude capabilities into

its

bombers

in the early- to

mid-1950s.

There was no simple technical solution to the problem of building a bomber with both low- and high-altitude capabilities. Different kinds of wings were needed for the two flight regimes: the optimal wing for high-speed, low-altitude

was relatively short and highly swept, to best wing for high-altitude flight was longer

flight

minimize drag, whereas the and straighten An aircraft designed specifically for high-altitude operations would not perform efficiently at low altitudes, so the Air Force was unlikely to get impressive low-altitude capabilities out of the B-58 and B-70 for free, as it had hoped. One solution to this problem would have been to strike an aerodynamic compromise, to design an aircraft optimized for neither flight regime but capable of performing reasonably well in both. But any such compromise was unacceptable to the Air Force, which believed that maximizing high-altitude performance was the key to operational effectiveness and that acquiring low-altitude capabilities at the expense of high-altitude performance would be a net loss. In addition, striking this kind of aerodynamic compromise would have broadcast to the rest of the American defense policy community that the Air Force was not certain about the future of high-altitude operations. This would have undermined its case for new bombers. Another solution to the twin-mission problem would have been a technological fix, such as the variable-sweep wing. A movable wing that could change angles as the flight regime demanded would have softened some of the harsh aerodynamic compromises that confronted bomber designers. The main problem with the variable-sweep wing was

Low-Altitude Pc net ra tion

that

it

would not have completely eliminated these compromises;

it

would not have allowed the Air Force to optimize high-altitude performance. Again, striking such a compromise was something the Air Force resisted for both operational and political reasons. In addition, the variable-sweep wing was still experimental at this time; the first American aircraft equipped with variable-sweep wings flew in 1951 and 1952. 2 Although the Air Force was often willing to gamble on new technologies that promised better bomber speed, altitude, and range, it emphasized the risks associated with

new

technologies that

fell

outside the tradi-

which the variable-sweep wing did. the Air Force failed to emphasize low-altitude capabilities be-

tional mission area, Finally,

deeply held beliefs about the viability of high-altitude operations could only be overturned by unambiguous evidence to the contrary. Long-term intelligence forecasts about Soviet air defenses were inherently speculative in the early 1950s, however; the Soviet Union was highly secretive and the United States had no overhead reconnaissance capabilities. These forecasts were not compelling enough to trigger an overhaul of the Air Force's operational doctrine and organizational belief system. The Air Force ignored these warnings in the early 1950s because it could ignore them. They were ambiguous enough to allow the Air Force to do what it was predisposed to do: build high-altitude bombers. 3 To be sure, the Air Force made pessimistic assumptions about the operational environment in the early 1950s, but its pessimism had limits. It recognized that improving Soviet air defenses would cut into the effectiveness of existing high-altitude bombers, and it assumed that the next generation of bombers would have to fly much higher and faster than its predecessors, but it refused to conclude that it would be impossible to conduct high-altitude operations, or that the B-58's and B-70's high-altitude capabilities should be compromised. The Air Force was predisposed to ignore intelligence forecasts even its own which suggested that its conception of strategic bombardment operations would have to be revised in fundamental ways. From the Air Force's perspective, such warnings were both intuitively wrong and organizationally troublesome. Therefore, the Air Force had to walk a fine line. Its projeccause

its

—

—

wing is Robert L. Perry, Innovation amt Military Comparative Study, Rand Corporation Research Memorandum, RM-5i82PR, Aug. 1967, pp. 34-73. See also Stuart M. Levin, "Why the Swing Wing?" Space/ Aeronautics 50 (Nov. 1968), 69-75. 3. The literature on the stability of belief systems is analyzed in Robert Jervis, Perception and Misperception in International Politics (Princeton: Princeton University Press, 1976), chaps. 4-7; John Steinbruner, The Cybernetic Theory of Decision (Princeton: Princeton University Press, 1974), chaps. 3-4. The role belief systems play in intelligence failures is analyzed in Richard K. Betts, "Analysis, War, Decision: Why Intelligence Failures Are Inevitable," World Politics 31 (Oct. 1978), 61-89. 2.

The best

Requirements:

history of the variable-sweep

A

[233]

Flying Blind

be pessimistic enough to justify building new bombers, but not so pessimistic that high-altitude operations could no longer be seen as viable. tions about Soviet air defenses

The Air late

had

to

Force's interest in low-altitude operations

reemerged

in the

1950s as hard evidence about Soviet high-altitude air defenses began

accumulate. U-2 overflights of the Soviet Union, which began in 1956, generated mounds of photographic evidence about the status of the to

Soviet air defense program. In addition, these flights were tracked

on

and U-2S were shot at by SAMs and interceptors armed with air-to-air missiles. As the years went by and U-2 flights continued, the Air Force was able to monitor the Soviet Union's air Soviet air defense radar,

defense improvements. 4 In 1958 the Air Force tried to reinsert a low-altitude requirement into

The B-70 had already been optimized for a specific high-speed, high-altitude flight regime by that time, and it would not fly efficiently at low altitudes unless it was completely redesigned. Redesigning the B-70 would have delayed the program and jeopardized its future, which the Air Force was loath to do. The idea of incorporating low-altitude capabilities into the B-70, therefore, was once again allowed to fall by the wayside. the B-70 program.

defenses continued to improve, SAC began low-altitude training exercises with its B-47S and B-52S in 1959. 5 The B-47 and B-52 were not designed for low-altitude operations, howSince Soviet high-altitude

air

and their low-altitude capabilities were consequently limited. One problem was that their range suffered when low-level flight was incorporated into their mission profiles; range, of course, was one of the Air Force's main concerns. Another problem was that, although the B-47 and B-52 could fly at fairly low altitudes or at fairly high speeds, they could not do both well at the same time; flying these aircraft at very low altitudes at speeds in excess of Mach 0.55 was risky. 6 In challenging either ground-based or airborne defenses, it was to the bomber's advantage to fly fast as well as low, in order to minimize its exposure time. Ground-based air defense radar had only a limited opportunity to detect fast-moving bombers that were able to hug the terrain and stay out of the defense's line of sight. Airborne systems eventually became the key to low-altitude air defense, but airborne deever,

4.

George

J.

Seiler, Strategic

Nuclear Force Requirements and Issues (Maxwell Air Force

Base, Ala.: Air University Press, 1983),

p. 109.

SAC/HO,

Development of the Strategic Air Command, 1946-1986, Sept. 1986, p. 84. 6. Lt. Col. John J. Kohout III, "A Post B-i Look at the Manned Strategic Bomber/' Air University Review (July-Aug. 1979), 24-31; Bernard Kovit, “Low-Altitude Penetration," Space/ Aeronautics 43 (May 1965), 76-83. 5.

[234]

Low-Altitude Penetration

fenses faced problems of their own: they had to be able to look

down

and pick out low-altitude bombers from all of the ground clutter that bombarded airborne radar and confounded pilots trying to make visual identification.

Low-level flight at high speeds was of course treacherous. Pilots had to avoid being clobbered by hills, trees, and man-made objects when flying at 200 feet or so. The alternative was to fly somewhat higher, but

gave ground-based radar a better viewing angle, and it made it easier for airborne radar to overcome the clutter problem. There was, this

The B-47 and B-52 were not designed with this problem in mind. Another problem the B-47 an d B-52 faced was that low-level flight placed a great deal of stress on their airframes. Neither aircraft was designed for the wear and tear of lowaltitude operations, and consequently both suffered stress and fatigue problems once they became involved in regular low-level exercises. Even the U-2 incident did not discredit high-altitude operations in the eyes of the Air Force leaders. The Air Force continued to support the B-70 and RS-70 programs with great vigor throughout the early 1960s. It therefore, a clutter-clobber trade-off.

maintained that high-altitude operations could still be conducted successfully because penetration corridors could be cleared out by bombers equipped with air-to-surface missiles and because bombers equipped with electronic countermeasures could confound whatever air defenses remained. Although many policy makers began to question the viability of high-altitude bombers and even the manned bomber itself. Air Force leaders were not among them. The U-2 incident was not a catalytic event as far as the Air Force was concerned. One sign of this is that the Air Force did not initiate any major studies of low-altitude bombers in the last eight months of i960, in the immediate aftermath of the U-2 incident.

event from the Air Force's point of view was the Kennedy administration's cancellation of the B-70's production run in March 1961. It was in the latter half of 1961 that the Air Force began the design studies that ultimately evolved into the B-i program. The cancellation of the B-70 was critical for the Air Force because it left no successor to the B-52, which was about to complete its production run, in the acquisition

The

catalytic

problem in i960; the Eisenhower administration had actually expanded the scope of the B-70 program in the latter half of i960, and Kennedy had supported it during the presidential campaign. Although the Air Force's faith in the operational viability of the high-altitude bomber was not shaken by the U-2 incident, it was starting to become clear in 1961 that the B-70 was no longer viable politically (see Chapter 6). The Air Force continued to pipeline.

The Air Force had not

[235]

faced

this

Flying Blind

lobby for the B-70 and the RS-70 in the early 1960s, but it also began to explore the idea of building a bomber with low-altitude capabilities; it needed a fallback option in the event its lobbying efforts failed.

The original impetus behind the B-i program was thus both strategic and political, with the latter especially important. The Air Force reacted to strategic developments in that it began to consider low-altitude options in the 1950s, but it moved slowly and took tentative steps because these developments called for a doctrinal shift. Strategic and operational concerns and even the U-2 incident were not enough by themselves

—

—

to get the Air Force to accept the necessity of building a low-altitude

bomber. Although Kennedy and McNamara were not trying to get the Air Force to reassess its ideas about low-altitude operations when they canceled B-70 production in 1961, their actions led the Air Force to undertake such a reassessment as it searched for ways to keep the

manned bomber It is

alive.

significant that the Air Force's design studies for a

new bomber

in

the early 1960s ultimately led to a dual-mission airplane. The Air Force's plan was to build a hybrid aircraft capable of performing both high- and low-altitude missions. The traditional, high-altitude mission, therefore,

was not discarded in the aftermath of the B-70's supplemented. The Air Force's refusal to abandon tions

even

political

in the face of

opposition

is

cancellation;

it

was

high-altitude opera-

growing operational problems and staunch

a testament to the durability of

its

beliefs

about

its

core military mission.

Preliminary Studies

As

the political problems surrounding the B-70 creasingly insurmountable in the early years of the tion, the Air Force

One was

began

to explore several

program became

in-

Kennedy administralow-altitude bomber options.

redesign the B-52 or the B-58 to improve their low-altitude capabilities. Since the production lines for both bombers were still open in 1961 and 1962, this option had the advantage of being relatively inexpensive. It also held out the possibility of providing SAC with a new

bomber

to

in short order,

which the Air Force

insisted

was

vitally

impor-

tant.

The idea

of deploying recycled B-52S

and B-58S did not appeal to the Air Force, however, because both aircraft had inherent low-altitude limitations. The B-58, for example, had limited range when flying at low altitudes and its compact airframe had little room for additional radar, bombing, and navigation equipment, which it would need in low-level flight. These limitations could not be overcome unless its airframe was

Low-Altitude Penetration

completely redesigned, which would negate many of the advantages of relying on a warm production line. Even then, the B-58 would not be particularly effective or efficient at low altitudes. The Air Force also recognized that building a modified B-52 or B-58 would probably foreclose the option of building an entirely new bomber,

which

it

had been trying

to

do

for years.

Given

a choice

between early and better per-

(which a modified B-52 or B-58 would offer) formance (which a new bomber would offer), the Air Force preferred to pursue the latter course. SAC, in particular, was not enthusiastic about settling for a modified B-52 or, even worse, a stretched version of the

availability

medium-range

B-58.

The

technical

commands

within the Air Force also

supported the idea of starting up an altogether new program. A new program would provide them the opportunity to develop new technologies, engineer new hardware, and design an entirely new system from

ground up. The Air Force attitude was reflected in a proposal made by the Air Force Systems Command (AFSC), which suggested that the solution to the bomber vulnerability problem was to build a new airplane capable of flying even higher and faster than the B-70. 7 One of AFSC's designs was the

for

an exotic

aircraft

capable of

Mach

8 flight at altitudes in excess of

feet. 8

The consensus of opinion in the Air Force was that it needed a new and modern, though not necessarily interplanetary, manned bomber. Although the Air Force hierarchy was still deeply com100,000

concluded that this AFSC proposal was too risky from a developmental and therefore a political perspective. One of the lessons the Air Force learned from the B-70 experience was that exotic technologies were counterproductive if they made programs vulnerable to cancellation. A second lesson the Air Force learned from the B-70 experience was that operational flexibility was an important political asset. When highmitted to the idea of high-altitude operations,

it

—

—

became problematic, the B-70 became vulnerable, politically if not operationally. A dual-mission capability would make a new bomber easier to sell to the civilian leadership. The Air Force consequently embraced the concept of operational flexibility, which it had rejected in the 1950s for the B-58 and B-70 programs. The Air Force did altitude operations

The Air Force Systems Command succeeded the Air Research and Development in 1961. It assumed responsibility for applied research, development, and weapon system production. Logistics, supply, and maintenance activities were assigned to the newly created Air Force Logistics Command. For more details, see the appendix. 8. Bernard C. Nalty, The Quest for an Advanced Manned Strategic Bomber: USAF Plans and Policies, 7961-1966, USAF Historical Division, Aug. 1966, p. 21, and USAF, Directorate of Advanced Engineering/HO, History of the Directorate of Advanced Engineering, July-Dee. 7.

Command

1962, p. 28; both in Files,

USAF/HO.

Flying Blind

not become an advocate of operational flexibility, therefore, until came politically expedient for it to do so.

A

and

it

be-

lesson the Air Force took from the debates of the late 1950s and early 1960s was that prelaunch survivability was an important third

final

design consideration. The vulnerability of American bomber forces to a Soviet ICBM attack was a growing concern. 9 Secretary of Defense McNamara, in particular, considered this a most serious liability. The B-52, for example, could not be dispersed to many airfields because it needed a long, wide runway. In addition, it was slow in getting away from its bases and it was not particularly resistant to the blast and thermal effects of nuclear weapons. The Air Force's next bomber would

have

to

do much

better

on these counts

if it

was

to get past the office of

the secretary of defense.

The Air Force was guided by these three considerations when it began to think about a new strategic bomber in the early 1960s. It conducted a series of design studies between 1961 and 1963 aimed at defining a new mission profile, major performance requirements, and some of the most likely system options. No hardware was developed at this stage; these studies were merely paper exercises. The first of these studies and the B-i's most ancient, direct ancestor was the Subsonic Low Altitude Bomber study, which was initiated in 1961 and completed in March 1962. This in-house study conducted by the Aeronautical Systems Division (ASD) of the AFSC focused exclusively on the problems associated

—

—

with low-altitude operations. 10 A September 1961 Rand Corporation study took a broader look at the bomber development problem. It argued that the Air Force had three basic system options: a hypersonic, boost-glide bomber similar to what the AFSC proposed; a supersonic, high-altitude bomber similar to the

and

bomber, which was seen as the most promising of the three. The Rand study argued that a low-altitude bomber capable of penetrating at high subsonic speeds would be better than a bomber capable of penetrating at supersonic speeds, because a supersonic bomber would probably not have intercontinental range. Building a suB-70;

a low-altitude

bomber would also take longer and cost more. According to Rand, a low-altitude bomber would be capable of only short supersonic bursts in any event. Moreover, the ability to dash at Mach 1.2 as opposed to Mach 0.9 would add little to the bomber's survivability, since personic

9. Fred Kaplan, The Wizards of Armageddon (New York: Simon and Schuster, 1983), chaps. 7-10; Albert Wohlstetter, "The Delicate Balance of Terror," Foreign Affairs 37 (Jan. 1959), 211-235. 10. Subsonic Low Altitude Bomber, ASD Report TDR-62-426, March 23, 1962; "B-i Strategic Bomber," Office of Information, AFSC, no date. In 1961, the Aeronautical Systems Division took over the aeronautical systems development activities previously conducted by the Wright Air Development Division of the ARDC.

Low-Altitude Penetration

the

main

defense threats it would face were high-Mach SAMs and interceptors armed with air-to-air missiles. The Rand study concluded that a Mach 0.9 capability was optimal at low altitudes, and a December 1962 Rand study reaffirmed this conclusion. 11 Given that the Air Force placed a premium on range and wanted to deploy a new bomber within air

five to eight years, before the effectiveness of its existing force

badly, the subsonic low-altitude penetrator

The Air

had much

to

eroded

recommend

it.

Force, however,

had two other factors to consider. First, its high-altitude mission would be cast aside if it built a bomber optimized exclusively for low-altitude flight. If it had to build a bomber with lowaltitude capabilities, the Air Force preferred to build a

bomber capable of Second, a bomber

performing both low- and high-altitude operations. optimized for low-level operations would probably have short, stubby wings, which would not be well suited for taking off and landing on short runways. This would limit the extent to which such an aircraft could be dispersed in times of crisis, which, in turn, would cut into its prelaunch survivability. These considerations led the Air Force to favor a variable-sweep wing for its new bomber. The aircraft's wings could be swept out (that is, made roughly perpendicular to the fuselage) for taking off and landing, which would reduce the bomber's runway requirements. They could also be swept out for high-altitude flight and long-distance cruising at moderate altitudes, which would boost the bomber's range. The wings could be swept back, however, for lowaltitude operations. This line of thinking began to dominate the program in 1963, when the Air Force initiated a second round of studies. The relatively simple, subsonic, low-altitude bomber that had been the focus of the analytic effort in 1961 and 1962 gave way to a variable-sweep, dual-mission bomber that eventually featured supersonic capabilities at both low and high altitudes.

The

first

of these studies

was

the

Low

Altitude

Manned

Penetrator

ASD. Although this bombwould probably penetrate at very low altitudes, it was also designed cruise at high altitudes. The study concluded that these compound

study, another in-house effort conducted by er to

requirements could best be met by an aircraft equipped with a variablesweep wing, which would also enable the bomber to take off in a distance of around 5,000 feet, roughly half that required by the B-52. 12 The Air Force's Advanced Manned Penetrator study, which involved outside 11.

R. B. Johnston et

A

Study of Future Low-Altitude Intercontinental Bomber Designs, 1961, pp. v, 1, 8-9, 29; E. P. Bomber, Rand Corporation RM-3183-PR, Dec. 1962, pp. viii-x.

al.,

Rand Corporation Research Memorandum, RM-2797-PR, Sept. Oliver, Summary Report of a Study on the Penetrating Manned Research 12.

Files,

Memorandum,

Proposed System Package Plan for a

HO/ ASD.

Low Altitude Manned

Penetrator,

AFSC,

Sept.

1,

1963,

See also "Low-Altitude Penetration Plane Studied," Aviation Week and

Space Technology, Sept. 16, 1963,

p. 26.

Flying Blind

contractors,

came

same conclusion about the variable-sweep focused on subsonic, low-altitude operations, but the

to

wing. This study also it considered the possibility of incorporating supersonic, high-altitude capabilities as well. 13 The third and final study initiated in 1963 was the Advanced Manned Penetrating Strategic System study. This study,

which involved Boeing, General Dynamics, and North American

in ad-

commands, originally focused exclusively on subsonic aircraft designs. By August 1963, however, its 14 scope had been expanded to cover Mach 2.2 and Mach 3.0 systems. dition to the Air Force's technical

The

A

AMSA Program

March 1964 to coordinate the Air Force's low-altitude design studies and related development activities. It was initially known as the Advanced Strategic Manned Aircraft program, but its name was changed to the Advanced Manned Strategic Aircraft (AMSA) program in April 1964, presumably because "ASMA" was not a dynamic-sounding acronym. The new bomber's basic performance requirements were set at this time. The Air Force called for a bomber with a high-altitude, Mach 2.2 capability as well as low-altitude penetration capabilities at speeds up to Mach 1.2. The traditional highspeed, high-altitude mission was alive and well in the AMSA program. In addition, the supersonic speed requirement had trickled down to the low-altitude level. The Air Force's standing faith in supersonic capabilities overrode the analytic arguments made against a low-altitude supersonic capability forcefully and repeatedly in 1961 and 1962. The Air Force's one concession to moderation was to set the highaltitude speed requirement at Mach 2.2 instead of Mach 3.0, as it had for the B-70. The AMSA program would thus not have to grapple with the program

office

was

set

up

in

technological challenges that ultimately helped to derail the B-70 (see

Although the Air Force was never shy about demanding exceptional performance from its bombers and state-of-the-art technology from its contractors, it was not willing to run excessive developChapter

6).

mental risks

this

time around.

clearly a function of

its

Its

decision to sidestep these risks

was

experience with the B-70.

Development Planning (DDP)/HO, History of the Directorate of Development Planning, July-Dee. 1963, pp. 37-39; "AMP Study Contracts," Aviation Week 13.

USAF, Directorate

of

and Space Technology, Nov. 11, 1963. 14. Nalty, Quest for Advanced Manned Strategic Bomber, pp. 22-25. Also, DDP/HO, History of the Directorate of Development Planning, Jan. -June 1963, pp. 34-35; July-Dee. 1963, p. 39; Jan. -June 1964, pp. 38-39; AFSC/HO, History of the Air Force Systems Command, July 1964 to June 1965; George C. Wilson, "USAF Stronger in Latest Aircraft Fight," Aviation Week and Space Technology, Feb. 24, 1964,

p. 28.

Lou^- Altitude Penetration

The Air Force's complex performance requirements led it to give the AMSA bomber a variable-sweep wing. The variable-sweep wing, it should be noted, had been around for over ten years at this point and, although it had never been used on an aircraft as large as a strategic bomber, it was not particularly risky technologically. One should not conclude, however, that development of the variable-sweep wing pushed the Air Force's bomber program along. The variable-sweep wing was not even featured in the design studies conducted in 1961 and 1962, when the developmental effort got under way, and it was not a new, exciting technology when it was formally incorporated into the AMSA program. The Air Force adopted the variable-sweep wing for a variety of operational and political reasons, as discussed earlier; it did not set up the new program to take advantage of variable-sweep technology. The AMSA program's most serious problem in 1964 was obtaining funds to get the development effort off the ground. This problem would dog the program as long as Robert McNamara was secretary of defense. McNamara believed that ICBMs and SLBMs should constitute the backbone of the strategic force structure and that strategic bombers only reinforced the deterrent capabilities provided by more cost-effective unmanned systems. Since American ICBMs and SLBMs were being deployed in great numbers in the mid-1960s, McNamara saw no urgent need to add a new strategic bomber to the force structure, and he was not prepared to rush the AMSA program into engineering development, as the Air Force repeatedly proposed. McNamara's decisions on the strategic bomber program in general and the AMSA program in particular

clashed sharply with the Air Force's preference throughout the 1964-

68 period. 15

One

of

December

McNamara's most 1965,

when he

controversial decisions

told the Air Force to begin

was announced

in

phasing out the 80

B-58S in the active inventory as well as 345 older B-52S. He also announced that 210 FB-ms, long-range models of the F-111 fighter, would

be deployed

in their place.

McNamara

felt

that this

modernized,

re-

—

vamped force would provide an adequate strategic capability and much more cost-effectively than other options. A fleet of FB-ms was expected to cost $1.9 billion, whereas a fleet of 200 AMSA bombers, for example, was estimated to cost $9-11 billion. Although AMSA bombers would be more effective than FB-ms, they would not be as costeffective as the smaller aircraft, according to McNamara. 16 15.

The deployment

of U.S.

outlined in Desmond Ball, 1980), chaps. 7 and 10. 16.

FB-m

ICBM and SLBM

Politics

forces in the early and mid-1960s is and Force Levels (Berkeley: University of California Press,

The FB-111 never became popular with SAC. Secretary procurement back [241]

to 76 aircraft in 1969

when

the

of Defense Melvin Laird cut Nixon administration began to

Flying Blind

and Congress disagreed. They argued that cutting the strategic bomber force structure from 680 to 465 operational aircraft was unjustified and dangerous. Moreover, McNamara planned to replace B-52S and B-58S, full-fledged strategic systems, with modified fighters. The FB-111, it was said, lacked the range, payload, avionics capability, and operational flexibility to be a useful strategic system. Although the Air Force conceded that the FB-111 was better than nothing, it maintained that the AMSA bomber should be pursued instead of or in addition to the FB-111. The Air Force certainly had nothing against modernization per se. Indeed, it had lobbied for a new strategic bomber for many years. It felt, however, that older bombers should be replaced by legitimate strategic bombers, and on a one-for-one basis. McNamara nonetheless strongly opposed proceeding too far or too fast with the AMSA program. In 1964, for example. Congress appropriated $52 million for the AMSA program for fiscal year 1965. The Air Force planned to spend $15 million of this on what it called the system's project definition phase; that is, the Air Force wanted to begin designing the bomber itself. According to the Air Force's plan, the rest of the program's budget would be devoted to advanced development of the propulsion and avionics subsystems. McNamara planned to proceed in a different manner. Although he recognized the value in research and development of propulsion and avionics, he opposed spending a lot of money on system definition. He recognized that this was the first step toward full-scale development and that the program would quickly build up momentum if much work was done on the bomber's airframe. With this in mind, McNamara decided to spend only $28 million on the AMSA program in fiscal year 1965, only $5 million of which would be devoted to system studies. 17 This pattern was repeated for several fiscal years (see Tables 8 and 9). McNamara determined, not just how much money was spent on the program, but how available funds were allocated; he spent what he wanted, when he wanted, and how he wanted. McNamara was especially careful to concentrate spending on propulsion and avionics. Work in these areas was always transferable to other programs, and it was important to continue this kind of basic research and development in any event. Only about 20 percent of the money spent on the AMSA

Many

in the Air Force

on McNamara's 1965 decision, see DOD Decision to Reduce the Number and Typies of Manned Bombers in SAC, Report, House Armed Services Committee, 89th Cong., 2d sess., April 1966. See also Cecil Brownlow, “Bomber Slash Follows B-111 Fund Rebuff," Aviation Week and Space Technology, Dec. 13, 1965, pp. 24, 27; "Improved B-111 Proposed to SAC," Aviation Week and Space Technology, June 6, 1966, p.

move forward with

the B-i program. For details

3i17.

46;

DDP/HO,

ASD, HO,

History of the Directorate of Development Planning, June-Dee. 1964, pp. 45History of the Aeronautical Systems Division, Jan. -Dec. 1965, pp. 81-88, 91-92.

Low- A l tit tide Penetra tion Tabic

8.

AMSA

budgets

(in millions of dollars)

Appropriated by Congress

Fiscal

year 1965 1966

$ 52.0 22.0

1967 1968 1969

22.8

Approved by

McNamara $ 28.0

46.0 18.8 26.0

47.0

30

a

23.0

$148.8

$143.8

Source: Department of Defense Appropriations for 1970, Hearings before the House Appropri-

ations Committee, 91st Cong., 1st sess., P- 539 a Carried over to

pt. 4,

-

fiscal

year 1970.

program between 1965 and 1968 was for system studies, which was exactly the way McNamara wanted it. It would have been politically difficult for him to cancel the program outright, given strong Air Force and congressional support for a new strategic bomber, so he kept the program alive, but barely. 18 By refusing to spend what Congress had appropriated on the AMSA program, McNamara kept the sunk costs in the program to a minimum, and he kept it from building up any appreciable

momentum. was anxious to move AMSA into full-scale frustration only mounted as McNamara repeatedly

Naturally, the Air Force

development, and rebuffed

its

its

appeals to expand the scope of the program.

McNamara

stood firm even against the recommendations of the Joint Chiefs of Staff in 1966 and 1967 to move the program into the contract definition phase. This step would have involved some engineering development, although not the initiation of full-scale development itself. 19 McNamara held his

ground year

after year,

and by 1967

it

was beginning

to

appear that the

Congressional support for AMSA continued to be fairly strong throughout the 1964"House Unit Increases AMSA Funding," Aviation Week and Space Technology, June 21, 1965, p. 25; George C. Wilson, "Rivers Renews Drive to Curb McNamara," Aviation Week and Space Technology, Jan. 17, 1966, p. 38; George C. Wilson, "Congress Fights for Defense Policy Role," Aviation Week and Space Technology, May 23, 1966, pp. 26-27; C. M. Plattner, "Variable-Geometry AMSA Studied," Aviation Week and Space Technology, Jan. 16, 1967, p. 26. More details on the AMSA program's funding history can be found in USAF, Directorate of Operational Requirements and Development Plans (DORDP)/HO, History of the Directorate of Operational Requirements and Development Plans, March-June 1963, pp. 123125; July-Dee. 1965, pp. 136-138; Jan. -June 1966, pp. 239-242; July-Dee. 1966, pp. 22718.

68 period; see

230; Jan. -June 1967, pp. 1-4. Also, ASD/HO, History of the Aeronautical Systems Division, Jan. 1967 to June 1968, pp. 318-321. 19.

Brownlow, "USAF Plans Crucial

AMSA

the Directorate of Operational Requirements

Bid in July,"

p. 28;

DORDP/HO,

and Development Plans, Jan. -June 1967,

History of p. 2.

Flying Blind

Table

AMSA

9.

element funding

(in millions of dollars)

Propulsion

Avionics

Systems

Year

development

development

studies

Total

1965 1966

$16.0

$ 3.0

$ 28.0 46.0

1967 1968

5.8 18.0

$ 7.0 14.3 10.3

Fiscal

24.0

$63.8

a

$31.6

7-7 2-7 8.0

$23.4

18.0

26.0 $118.8

DORDP/HO,

History of the Directorate of Operational Requirements and Development Plans, July-December 1967. a Funding for the avionics program came out of the systems studies

Source

budget

:

year 1968.

in fiscal

AMSA program was permanently stalled.

Boeing came to this conclusion in late 1967 and dropped out of the system study effort. Had McNamara stayed in office past February 1968, it seems clear that the program would have stayed in this developmental no-man's land.

Initiation of the B-i After

McNamara

the Pentagon in February 1968, the Air Force program into engineering deefforts to move the left

redoubled its velopment. It soon discovered a Clifford,

Program

AMSA

much more receptive audience in Clark McNamara's successor: a July memorandum from the Office of

and the director of Defense Research and Engineering directed them to study ways of speeding up the AMSA program in the event that full-scale development and production decisions followed. According to this directive, a proposal for cutting development time and moving expeditiously into production was to be one of the four options explored. 20 Whereas McNamara had strenuously resisted every attempt to move the program toward engineering development, Clifford was thinking of moving ahead. The problem for his staff and the Air Force was to determine the the Secretary of Defense to the secretary of the air force

best

way

to proceed.

The 1968 Development Concept Paper

Determining the best way to proceed was the purpose of the AMSA Development Concept Paper (DCP), which was circulated through the

20.

DORDP/HO,

Historx/ of the Directorate of Operational

Requirements and Development

Plans, Jan. -June 1968, pp. 177-181; July-Dee. 1968, pp. 180-183.

[244 1

Low-Altitude Penetration

Department

of Defense

stipulated that the

dash

capability, at

would

and the Air Force

new bomber

fly at

an altitude of 200

in

Mach

feet

Mach it was

also be required to fly at

November

1968.

The

DCP

Mach 1.2 The bomber

0.85-0.95, with a

over rolling terrain.

2.2 at higher altitudes.

The

DCP

acknowledged, however, that difficult to quantify the military value of supersonic dash capability at low altitudes, which seemed 'To be a matter of judgment." The costs of this requirement were not difficult to quantify, however. Studies conducted by Rand and the Planning Research Corporation estimated that an all-altitude supersonic capability would increase the procurement and life-cycle costs of the program by 15-30 percent. The DCP nonetheless continued to impose this requirement on the program. 21 The DCP assumed that it would take around eight years to develop, produce, and deploy the new bomber, depending on the development and production program selected. The heart of the DCP was an analysis of four different acquisition options, each with advantages and disadvantages.

The

option called for a three-year design competition that would not include the construction of competing prototypes. After three years first

of trade-off studies,

wind tunnel

tests,

and detailed design work,

deci-

would be made about whether production was warranted and, if so, what the design of the system and who the prime contractor would be. This option would provide a fairly prominent decision point located between the design and engineering development phases of the program, and it would also provide decision makers with a fair amount of information about the program's cost and technical feasibility for their production, design, and source selection decisions. The main disadvantage of this option was that it did not provide for a full-scale prototype competition, so critical program decisions would not be based on the sions

kind of hard data that only prototypes could provide. This option, therefore, was sequential in form, but not in substance.

The second option allowed for only a fifteen-month contract definition phase. At the end of this period, the prime contractor would be selected, the design of the system would be frozen, and development and production contracts would be negotiated. The cost of supporting contractors for three years would not be borne under this option. If everything went according to plan, this option would lead to an early operational capability because it would move quickly into full-scale development Deiwlopiment Concept Papier for AMSA, Nov. 6, 1968, Files, Systems Analysis Technical Library, Bolling Air Force Base. The evolution of AMSA's performance requirements can be tracked in Dir. of Op. Reqs. and Devel. Plans, Characteristics, Cost, and Effectiveness of 21.

1965, and Dir. of Op. Reqs. and Devel. Plans, Summary Report of the Concept Formulation Studies of AMSA, Oct. 1966; both in Files, HO/ASD.

AMSA, Aug.

Flying Blind

and production. This plan was risky, though; it contained considerable concurrency. The design of the bomber would be frozen at the end of the contract definition phase, so it would be difficult to incorporate subsequent technological developments into the system. The system's specifications would be set eight years before the bomber became operational, so the Air Force would have to make accurate, long-range intelligence forecasts about the bomber's operational environment. According to the

DCP, this could not be done with any degree of confidence. The third option would defer the decision on system development for one year. In the meantime, the program would stay on the same course it had been on for several years. Although this option would not involve a major increase in expenditures in the short run, it would not do much address the program's technological puzzles either. And it did not address the lead time problem that concerned the secretary of defense. Finally, this option did not represent any progress toward full-scale development or production; it was merely a continuation of McNamara's policies. to

The fourth option was like the first, except that the contractors would study subsonic bomber designs in addition to supersonic designs. Although the DCP was not explicit, it seemed to suggest that a simpler, cheaper, fixed-wing subsonic bomber would be considered if this development option was pursued. The consensus of opinion in the Pentagon was for the first option, which called for a three-year design competition. It was endorsed by John Foster, the director of Defense Research and Engineering; Harold Brown, the secretary of the air force; and Paul Nitze, who signed for Clifford. The only dissent came from Alain Enthoven, the assistant secretary of defense for systems analysis; he favored the fourth option on the grounds that alternative mission profiles should be considered by the contractors. Foster and Brown argued, however, that adequate mechanisms existed within the framework of the first option for studies of alternative subsonic designs. In deciding the matter, Nitze split the difference between these two positions by insisting that trade-off studies of a subsonic system be

made and

that the supersonic-subsonic issue be

Although a strong consensus among the civilian leadership in the Pentagon could have overridden the Air Force's low-altitude supersonic requirement, a consensus of this kind did not materialize in 1968. The DCP even went so far as to allude to the rich bureaucratic heritage of the all-supersonic requirement and the effort that had already gone into meeting it. No votes were cast for the third option, the "status quo" option. It is interesting that Foster, Brown, and Enthoven, none of whom was a periodically reexamined.

[246]

Low-Altitude Penetration

strong supporter of the

AMSA

program while McNamara was in office, all voted to move the program along in 1968. Foster and Brown were especially eager to push ahead. Foster argued that it was important to field the new bomber because its low radar cross section, infrared signature, and speed capabilities would make it a formidable penetrator; this was the first time that what would later be known as "stealth" capabilities played a

major role in deliberations about strategic bomber options. Brown also argued that the penetration capabilities of the new bomber were needed, and he stressed that reducing the program's lead time was an urgent consideration. Even so, no one recommended pursuing the second option, the highconcurrency option. This is especially interesting given the dominant role played by concurrency in most of the major weapon procurement programs of the McNamara era, such as the F-111 and the C-5A programs. Although the first option did not call for a perfectly sequential program, it featured some competition during the development stage of the process and it did call for a discrete production decision. The 1968 DCP therefore represented a shift away from the kind of concurrency that had been favored by the McNamara Pentagon and by the Air Force for many years. It is significant that this step was taken by the civilian leadership in the Pentagon not by the military and it preceded the highly publicized "Packard reforms" of the Nixon administration. Concurrency's promises were met with skepticism even in 1968. 22 The 1968 DCP was also noteworthy because decision makers consid-

—

—

ered sequential and concurrent options at the same time. High-level attention to the issue of choosing a procurement strategy was missing

from most of the bomber development programs of the postwar era. In most previous programs, in fact, no real choice existed; the dominant procurement strategy of the day was simply put into effect. Although

was not much

there

of a debate in 1968,

options were at least formally listed side

To implement

two very by side.

this decision, the secretary of

different acquisition

defense requested $97

program for fiscal 1970. Although this request was later $85 million and ultimately to $77.2 million, the AMSA budget

million for the

lowered to nonetheless roughly tripled as a result of the 1968 DCP decision to move into system development. 23 Thus, the program began to negotiate the transition toward full-scale development before the Nixon administration took office.

The problems that concurrency created were understood by this time.

22.

ple,

23.

DORDP/HO,

and C-5A programs,

for

exam-

History of the Directorate of Operational Requirements and Development

Plans, July-Dee. 1968, p. 183.

(247]

in the F-111

Flying Blind

The 1969 Development Concept Paper

The Nixon administration,

by Deputy Secretary of Defense David Packard, moved quickly to review the AMSA program. By the end of 1969 the performance requirements for the bomber had been modified, a new development concept paper had been issued, and a new, accelerated development program had been put into place. Packard was briefed on the program's status within a few weeks of taking office. He subsequently asked for an alternative development plan based on full-scale development in fiscal year 1970; as it stood, fullscale development would follow the three-year design competition. Packard recommended on March 7 that $23 million be added to the program's funding for fiscal year 1970; this would increase funding from led in this area

$77.2 million to $100.2 million. 24

Packard was anxious to solicit full-fledged proposals from the aerospace industry, but three problems had to be dealt with first. One problem concerned the size, complexity, and cost of the new bomber's avionics. Packard decided that the Air Force would follow a phased approach to avionics development; it would upgrade the avionics package as time went by. As the program jelled, therefore, allowances would have to be made for avionics growth. 25

To keep the bomber's weight and cost under control, Packard lowered its low-altitude speed requirement from Mach 1.2 to Mach 0.95 and its payload capacity from 32 to 24 short-range attack missiles. The lowaltitude speed requirement for the bomber was ultimately set at Mach 0.85 for sustained operations over rolling terrain. These decisions were implemented by the Air Force in late July, but only reluctantly; a lowaltitude supersonic capability had been a basic feature of the program for years. Packard felt, however, that the marginal operational value of a supersonic dash capability was outweighed by its costs. He believed that it was important to keep the program's costs under control and that this might have to be achieved at the expense of performance. 26 This philos-

ophy was

A

third

rewritten

alien to the Air Force.

problem was if it

was

to

development plan had to be development in 1970. A new

that the program's

move

into full-scale

24.

Ibid., Jan. -June 1969, p. 100.

25.

Ibid., p. 101.

Development Concept Paper for B-i, Oct. 1, 1969, p. 2, Files, Systems Analysis Technical Library, Bolling Air Force Base. More details on the discussions over the program's low-altitude speed requirements can be found in B-i: Request for Proposal, System Specification, Sept. 17, 1969, pp. 7-8, and B-i Requirements and Characteristics Summary, Report ASZD 68-1, Sept. 18, 1969; both in Files, HO/ASD. Also, Technical Development Plan: B-i AMSA, Oct. 1969, Files, Systems Analysis Technical Library, Bolling Air Force Base. DORDP/HO, History of the Directorate of Operational Requirements and Development Plans, July-Dee. 1969, pp. 86-87. 26.

Low-Altitude Penetration

Table 11. B-i

Option

development options Competition ends

First

Estimated

flight

development costs $2,086 billion 2.148

1

May

1970

Nov. 1973

2

Oct. 1970

3

Jan. 1972

4

Feb. 1971

Apr. 1974 July 1974 July 1974

Source:

Development Concept Paper

2.384 2.170

for B-i, 1969, p. 15.

development concept paper was consequently issued in October 1969. Once again, four development options were considered (see Table 10). The first option called for a six-month design competition, which would be immediately followed by a source selection decision. This option had several distinct advantages, given Packard's interest in accelerating the program. It would get the program into full-scale development by the end of fiscal year 1970, and it would therefore reduce the program's development costs, perhaps by as much as 13 percent. It would also reduce the paper competition, which Packard thought was cost-ineffective and time-consuming. The main drawback to this option was that it was fairly risky. Design and source selection decisions would have to be made on the basis of preliminary paper studies; little hard data would be available for critical decisions. The second option featured a two-stage source selection process. First, proposals would be solicited from the aerospace industry in a preliminary three-month competition, and then two contractors would be selected to develop more elaborate proposals over the course of a sixmonth design competition. A single contractor would then be selected for engineering development. This option would provide the Air Force with more and better information for its design and source selection decisions than the first option. On the other hand, it was estimated that it would cost 3 percent more and proceed more slowly, an important consideration for both Packard and the Air Force. The third option extended the competition into hardware development. Design, source selection, and production decisions would be made only as data about program costs and technological feasibility became fairly reliable. This option was therefore the most sequential of the four. One of its disadvantages was that development costs were expected to be about 10 percent higher under this option than under any of the others. Even so, an extended hardware competition was not prohibitively expensive. The DCP pointed out, moreover, that a hardware competition had many important advantages: it usually resulted in a "superior system design," and "experience has shown in many cases

Flying Blind

development program cost and schedule realism tend to emerge only after [such] a competition has ended." 27 The DCP also pointed out that development costs usually constituted less than 20 percent of total procurement costs. This was significant because prototype programs tended to be relatively inexpensive in the long run; extensive hardware testing led to smoother production runs. These long-term savings, however, unlike short-term development costs, were difficult to quantify at the outset of the program. This option's other main disadvantage was that it would interfere with the Air Force's long-range planning. It would not resolve the Air Force's nagging concerns about whether production would be authorized, and, if so, who the program's prime contractor would be. The Air Force could not begin its elaborate planning activities until these issues were resolved, which was a major consideration from an organizational perspective. This option would put off these key decisions until the results of the prototype competition became clear. The fourth option postponed the design competition while the Air Force and Office of the Secretary of Defense continued to study performance requirements and development risks. This option was seen as relatively expensive, time-consuming, and inappropriate given the strong interest in accelerating the program. Going back to the policies of the McNamara Pentagon was not what most of the decision makers in the new administration had in mind. The key decision makers in the administration could not agree on the that

best

way

to

as

was

called starting in 1969, disrupted the cautious

it

proceed, however. Packard's interest in accelerating the B-i,

existed in 1968 for a three-year design competition.

were some

As

consensus that a result, there

fundamental disagreements in the 1969 DCP. John Foster, still director of Defense Research and Engineering, favored the second option because its nine-month design competition most closely approximated the course of action chosen in 1968. The Office of Systems Analysis voted to postpone the design competition altogether; it recommended the fourth option on the grounds that the Air Force had not yet adequately analyzed the program's performance requirements, system fairly

specifications, costs,

and

would allow the commitment was made to

technical risks. This option

Air Force to conduct these analyses before a

development. By the same token, the Office of Systems Analysis preferred the second and third options to the first, because the risks of limiting the design competition to six months were seen as excessive. The new secretary of the air force, Robert Seamons, favored the first option, however, because seven years of studies had already generated a lot of paper analyses and because hardware competitions tended to be full-scale

27.

Development Concept Paper for B-i, 1969,

p.

16.

[25°]

Low- Altitude Penetration

expensive. The other options would only add to the pile of paper and to the program's delays, in his opinion. Packard ultimately decided in favor of the first option, despite the strong dissents, but with two stipulations: the Air Force would be required to identify milestones during the devel-

opment program where production

cost estimates

would be updated,

and the possibility of continuing with two contractors into engineering development would be left open. The most sequential of the four alternatives received no votes. The immediate costs of a hardware competition and other short-term considerations outweighed the long-term benefits, such as design flexibility and production savings, which would accrue in a sequential program. As it turned out, design flexibility was sought later in the B-i program, when design changes were difficult and expensive to implement. Even Packard, later known as the champion of sequential development and prototype competitions, did not endorse

His interest in year 1970 pushed

this option.

moving the B-i into full-scale development in fiscal him in the direction of the first option. Ironically, this alternative contained more concurrency than any of the others. Although Packard tempered this concurrency (by inserting program milestones and keeping the possibility of extending competition into engineering development),

he did not make the program a model of sequential development. Instead, he merely modified a concurrent option by incorporating sequential elements. The program would still rely heavily on paper studies for design, source selection, and even production decisions; probably have no competition beyond the design stage; freeze the design of the system at an early stage; and require a considerable funding long before a production decision could be made. Packard did eliminate one of the B-i's riskier performance requirements, the low-altitude supersonic speed requirement. In doing so, he cut back the technological adventurousness of the program and, wittingly or unwittingly, brought the technological objectives of the program more in line with the mixed procurement strategy that

of the 1969 DCP. solicited from the aerospace industry in

came out

Proposals were

November

North American Rockwell, Boeing, and General Dynamics submitted B-i designs, and in June 1970 it was announced that Rockwell had won the competition. 28 Rockwell came out on top in terms of both cost and performance, which was still the Air Force's main consideration. 1969.

Rockwell was especially bold

in its integration of the aircraft's

various

North American Aviation merged with Rockwell Standard in 1967. Although the new company was formally known as North American Rockwell until 1973, when it refer to it as Rockwell throughout this changed its name to Rockwell International, 28.

I

section.

Flying Blind

subsystems, which appealed to many in the Air Force. 29 Both Boeing and General Dynamics hedged their bets and submitted fixed-wing designs in addition to variable-sweep options. Although the fixed-wing designs did not meet the Air Force's performance requirements, they were low-cost, low-risk alternatives. The Air Force, however, was not prepared at this point to sacrifice any more performance for cost savings.

The

B-i

Development Program

Although Packard was anxious to proceed with the B-i, he believed that the program's development objectives and procurement strategy still needed to be refined. A major concern was the program's estimated procurement costs: $11.2 billion for 244 aircraft. Many believed this estimate to be wildly optimistic. With this in mind, Packard ordered two separate reviews of the program in the last half of 1970, before full-scale development got under way.

Project Focus

The

first

of Packard's reviews. Project Focus,

nical specifications. Its goal

was

examined the

B-i's tech-

design of the aircraft to This, in turn, was expected to

to simplify the

reduce the risk associated with building it. help keep the program's costs under control. The main change was a reduction in the amount of titanium in the airframe from 38 to 19 percent.

Although titanium was

a relatively strong, lightweight material

— and

therefore ideally suited for an aircraft with the B-i's low-altitude, long-

—

range requirements it was also expensive. In addition, it had never been used on this scale in an aircraft as large as the B-i. Although substituting one material for another might seem like a relatively mundane matter, this was actually a major design change. The design of the airframe would have to be modified to accommodate the use of steel and aluminum instead of titanium, and changes in the airframe would inevwas also true that Rockwell needed the business. Rockwell's F-100 production run had ended; work on the Apollo program was ending; its commercial Sabreliner was a flop; and it lost the F-15 contract to McDonnell Douglas in December 1969. Boeing, on the other hand, had contracts for the Minuteman ICBM and its 747 commercial carrier, and General Dynamics would inevitably get the contract to build the new Trident submarine. Since Rockwell came out ahead in the B-i design competition on cost and performance grounds, 29.

It

it is impossible to say if these economic considerations were influential in the Air Force's source selection. For details on Rockwell's plight, see Nick Kotz, Wiki Blue Yonder (New York: Pantheon, 1988), pp. 90, 94-95. For more details on the B-i proposals, see Michael P.

London, "B-i: The Last Bomber?" Space/ Aeronautics 53 (April 1970), 26-33; Cecil Brownlow, "Major Incentives Offered for B-i," Aviation Week and Space Technology, June 15, 1970, pp. 12-13.

Low- Altitude Penetration

Rockwell B-i (U.

S.

Air Force)

many

subsystems. It was expected, though, that these changes would have only a negligible effect on performance. 30 The B-i was fairly ambitious technologically even after these and other changes were made in its design. Many assumed that it was techitably affect the designs of

nologically unambitious because

aerodynamically.

not a

It

is

it

was not

particularly

adventurous

certainly true that the variable-sweep

wing was

new development. Work on movable wings had been conducted

the United States since the end of World

War

II,

and prototype

in

aircraft

with variable-sweep wings had flown in the early 1950s. Important technical advances were made in the late 1950s, and the major obstacles to using variable sweep were in the process of being overcome. 31 The Air Force went on to build the F-111 with a variable-sweep wing in the early 1960s, which pushed the learning curve even farther along. Even so, a was estimated design changes, would 30.

It

that the aircraft's increased weight, in conjunction with other

increase the B-i's take-off distance by 9 percent. This was not expected to have a significant effect on the number of runways to which the B-i could be dispersed in a crisis. For more details on Project Focus, see Project Focus Configurations, 1970, and B-i Development Concept Paper, revised Aug. 20, 1976; both in Files, HO/ASD. Also, USAF, Directorate of Development (DD)/HO, History of the Directorate of

Sept.

2,

Development, July-Dee. 1970;

DOD

Appropriations for 7972, Hearings before the

Appropriations Committee, 92nd Cong., 1st sess., 31. Perry, Innovation and Military Requirements, p.

pt. 6,

67.

pp. 322-324.

House

Flying Blind

variable-sweep wing had never been used on an aircraft as large as the B-i, and the movable wing added to the aircraft's complexity and weight. Although the variable-sweep wing did not pose great risks for the program,

Some

it

clearly complicated the B-i's design.

were confronted by more formidable technical challenges. The avionics, in particular, would have to be highly advanced. The offensive avionics suite would have to navigate and guide an aircraft that weighed almost 400,000 pounds as it flew at high subsonic speeds just 200 feet above the ground. This had never been done before, and highly advanced terrain-following radar, terrain-avoiding radar, flight control systems, and weapons delivery systems would consequently have to be developed. On the defensive side, advanced electronic countermeasures would be needed to deal with the Soviet Union's increasingly capable air defense forces. Since developing advanced offensive and defensive systems seemed problematic, a decision was made at the outset of the program to rely on off-the-shelf components as much as possible, thus providing an interim capability while more advanced systems were brought on line. 32 The engines and engine inlets also had to be highly advanced. They had to perform efficiently in several radically different flight regimes, so their designs were novel and quite complex. Advanced cooling techniques and new high-temperature alloys had to be used in the B-i's engines, for example. 33 One of the most adventurous features of the B-i's design was its crew escape module. Instead of using conventional ejection seats in the aircraft, Rockwell planned to put all four crew members in an enclosed module that would be ejected from the aircraft in an emergency. The design problem was to ensure that the module would be aerodynamically stable under a wide variety of flight conditions and give the crew a soft landing even if ejected from the aircraft at low altitudes. It was recognized from the beginning that developing this module would be a major challenge. 34 of the B-i's subsystems

Project Innovation

The second procurement

strategy. Its objectives

that flight testing 32.

33.

were

to

would begin long before

the production decision

was

GAO,

6-1 Weapon System, March 1974, pp. 50-38. Leon H. Dulberger, “Advanced Strategic Bombers," Space / Aeronautics 45 (June

pp. 62-75;

June

on the program's restructure the program so

B-i review. Project Innovation, focused

"GE

Tests

New

Technology Engine

for B-i," Aviation

1966),

Week and Space Technology,

26, 1972, pp. 61, 64.

ASD/HO,

History of the Aeronautical Systems Division, July 1973 to June * 974 v ol. 1, pp. 150-151. Also, Gen. John B. Hudson, Vice Commander, AFSC, “B-i Program Cost Reduction," Memorandum to HQ/USAF, Oct. 11, 1984, Files, Office of the Deputy Chief of 34.

Staff for

Programs.

/

Low-Altitude Penetration

made, and

minimize investment prior to the production decision. The idea, in other words, was to reduce concurrency. Two main steps were taken in January 1971 with these objectives in mind. First, the program's schedule was revised so that twelve months of flight testing would take place prior to the production decision instead of just six. Packard emphasized that it was important to "fly before you buy," to assure that major technological obstacles had been overcome and relatively reliable cost and performance estimates were in hand before a commitment was

made

to

Under the revised plan, the first flight was scheduled for March 1974, and the production decision would be made in April 1975. Second, the number of prototypes to be built was reduced from five to three, saving $300 million and minimizing costs while flight testing was taking place. These prototypes, moreover, were to be built on soft development tooling rather than hard production tooling. The three prototypes would therefore not necessarily represent the thin edge to production.

wedge. 35 It is important to emphasize that the impetus to simplify the B-i's design and shift to a more sequential procurement strategy came from the civilian leadership in the Pentagon and not from the Air Force. The Air Force opposed the move away from concurrency; the "fly before you buy" strategy withheld a production commitment until the B-i actually proved itself in the air. Packard took steps to make the production decision a real decision, one that could go either way. The Air Force had a vested interest in seeing the B-i production decision go a particular way. Some in the Air Force argued that the "fly before you buy" strategy of a production

was inappropriate

because it would create a time lag between the first flight and the first production order. 36 The problem, as the civilian leadership in the Pentagon recognized, was that "if you built airplanes and then went away and tested them, the people who built them would have gone away by the time you wanted to build some more." 37 Many in the Air Force and at Rockwell argued that building more prototypes would smooth the transition from development to production. Taking steps to smooth the transition to production, though, in this case

See U.S. Air Force Flight Test Center/HO, History of the USAF Flight Test Center, July 1971 to June 1972, pp. 135-138; DD/HO, History of the Directorate of Development, Jan. -June 1971; DOD Appropriations for 1972, pt. 6, pp. 324-327; 8-1 Development Concept Paper, 1976, p. 10; Cecil Brownlow, "USAF Tightens B-i Management," Aviation Week and Space Tech35.

nology, Feb. 22, 1971, pp. 56-59; 1973 to June 1974, p. 83.

AFSC/HO,

History of the Air Force Systems

Command,

July

36. Lt. Col. Bunting, "Talking Paper on Fly-Before-Buy," Jan. 25, 1974, Files, HO/ASD. See also John W. Rustenburg et al., "B-i Structural Test Program," Defense Systems Management Review, Winter 1977, pp. 149-160. 37. Testimony by Grant Hansen, Assistant Secretary of the Air Force for Research and Development, DOD Appropriations for 1973, Hearings before the House Appropriations Committee, 92d Cong., 2d sess., pt. 4, pp. 664-665.

Flying Blind

assumed that a production run was a given. Taking these steps, moreover, would make a production commitment more likely by contributing to the program's sunk costs and momentum. Pushing the program toward production would actually be counterproductive if serious technical problems emerged during development; this was not out of the question with a system as complex as the B-i. The Air Force's overriding interest, however, was not in avoiding technical problems but in getting a production commitment. The B-i's procurement strategy was revised in 1971, as it had been in 1968 and 1969. Packard tried to walk a fine line, pushing the program into full-scale development while taking steps to avoid a commitment to production. The balance between sequential and concurrent development continued to fluctuate throughout the course of the program. The key point, though, is that the program always contained a combination of sequential and concurrent elements, even though the mixture itself changed from time to time. This was not an unreasonable way to proceed, given that the B-i involved a few, but only a few, technological challenges. For this kind of program, the goal

was

to achieve a

"balance

between concurrency and step-by-step, which gets the maximum advantage of testing before you buy, but minimizes the impact of producing and then shutting down and then starting up again." 38

Development Problems Technical problems began to appear as the B-i

moved

into full-scale

development. The crew escape module proved to be especially troublesome. Its builders were unable to get it to work reliably under all flight conditions. The Air Force nonetheless decided in November 1972 to continue to work on the module rather than replace it with conventional ejection seats. 39 It also decided to change the engine inlets from a mixed compression design to a simpler, external compression arrangement to optimize the aircraft for subsonic rather than supersonic

—

flight.

This suggests that, until

ing the B-i for supersonic

late 1972, the

— that

is,

Air Force

high-altitude

was

still

optimiz-

— missions.

The Air

was therefore slow to acknowledge that the subsonic, low-altitude mission was primary. The switch to the simpler inlets cut 1,400 pounds from the weight of the aircraft, which was an issue of growing concern. In addition, the switch to the simpler inlets would save $46 million in production costs and an additional $12-17 million in long-term operating costs. Keeping the cost of the program under control was also a Force

38.

Ibid.

39.

ASD/HO,

History of the Aeronautical Systems Division, July 1972 to June 1973, pp. 173-

174.

[2S6|

Low-Alt tude Penetra tion i

growing concern. This switch would, however, lower the top speed of the B-i from Mach 2.2 to Mach 2.1. 40 More serious problems began to emerge in 1973. Rockwell had problems constructing the first prototype, and it consequently fell behind schedule. In particular, the carry-through structure in the fuselage for the variable-sweep wing proved to be troublesome. Subsystem integra-

behind schedule, in part because Rockwell had to divert manpower from subsystem development to airframe construction; in a concurrent program, many things had to be done simultaneously, which was not always easy. In addition, many subcontractors fell behind schedule in delivering subsystems. This was one of the risks associated tion also

fell

with trying to build a highly complex, highly interdependent system on a tight schedule. The Air Force believed that management deficiencies at Rockwell only made matters worse. 41

As

the program slipped farther behind schedule, the Air Force was left with two basic options. It could try to get the program back on schedule

by spending $134 million on overtime for Rockwell's workers, which is what Rockwell wanted to do. Or it could revise the program's schedule to take these problems into account, which would save it from having to ask Congress for more money in the short term. In the long run, though, this option would add to total costs, since stretching the program out would obligate the Air Force to pay contractor overhead costs for a longer period of time. The Air Force estimated that this option would ultimately add $78.8 million in costs spread over several years. Since the B-i

was already beginning

to attract a great deal of critical attention in

Congress, the Air Force did not want to aggravate the situation by asking for a bigger annual budget. Instead, it decided in July 1973 to stretch the program's schedule. The B-i's first flight, which had already slipped from March to April 1974, was rescheduled for mid-1974. Construction of the second and third prototypes would be delayed to keep the program's budget down, but the first flight of the second prototype would therefore be delayed from April 1975 to January 1976. The flight test B-i Development Concept Paper, 1976, p. 11; USAF, Directorate of Development and Acquisition (DDA)/HO, History of the Directorate of Development and Acquisition, July-Dee. 1972, pp. 123-124; "B-i Engine Inlet Design Simplified," Aviation Week and Space Technology, 40.

Oct.

9, 1972, p. 13.

41. The Air Force has also been accused of trying to cover up Rockwell's problems; see Kotz, Wild Blue Yonder, pp. 108-110. For more details on the B-i's problems, see ASD/HO, History of the Aeronautical Systems Division, July 1973 to June 1974, pp. 123-125; AFSC/HO,

History of the Air Force Systems Command, July 1973 to June 1974, pp. 68-79; B-i Development Concept Paper, 1976, p. 11; GAO, B-i Weapon System, March 1974, p. 15; "B-i Prototype Production Stretched," Aviation Week and Space Technology, July 16, 1973, p. 16; Cecil Brownlow, "Proposed Cuts Could Cripple B-i Project," Aviation Week and Space Technology, Aug. 27, 1973, p. 18; DDA/HO, History of the Directorate of Development and Acquisition, Jan.-

June 1973,

p. 109.

Flying Blind

whole would thus be set back a great deal. The production decision, which had already slipped from April to July 1975, was consequently pushed back, to May 1976. 42

program

as a

Outside Reviews

The decision to stretch the B-Ts schedule led to two outside reviews of the program. The first of these was conducted by the Senate Armed Services Committee, which held a special set of hearings on the B-i in July 1973. The committee decided to cut $25 million from the program's fiscal year 1974 budget, reducing it to $448.5 million. The Air Force, which already believed that Congress was partially responsible for the delays in the program because of its unwillingness to fund it at the levels requested, noted that this cut would delay the first flight of the B-i to 43 late 1974 and the production decision to November 1976. The second outside review of the B-i was initiated by the secretary of the air force, John McLucas. He asked Raymond Bisplinghoff, deputy director of the National Science Foundation and former dean of engineering at the Massachusetts Institute of Technology, to determine whether the B-i program was likely to meet its cost, schedule, and

put together a small committee, which reported to the secretary in October. The committee could see no major technical obstacles to development and production, but there were some problematic areas. The aircraft's estimated gross weight, for example,

performance

targets. Bisplinghoff

was growing from 360,000 to 395,000 pounds, which would inevitably impinge on the bomber's range and take-off distance. It was not expected, though, to prevent the B-i from performing the mission for which it was designed. More worrisome was the new electrical multi-

plexing system, the electrical guts of the view, this

out

it.

was an

aircraft.

In the committee's

area of grave concern, and the B-i could not fly with-

44

ASD/HO, History of the Aeronautical Systems Division, July 1973 to June 1974, pp. 123AFSC/HO, History of the Air Force Systems Command, July 1973 to June 1974, pp. 71-79;

42.

124;

DOD

Appropriations for 1974, Hearings before the House Appropriations Committee, 93d Cong., 1st sess., pt. 7, pp. 978-979; Fiscal Year Authorization for Military Procurement, Hearings before the Senate Armed Services Committee, 93d Cong., 2d sess., pt. 7, pp.

4084-4091.

ASD/HO,

Histori/ of the Aeronautical Systems Division, July 1973 to 124; B-i Development Concept Paper, 1976, pp. 11-12. 43.

June 1974, pp. 123-

Committee Report, Oct. 4, 1973, pp. 1-10, and John L. McLucas, Letter to John C. Stennis, Chairman, Senate Armed Services Committee, Feb. 6, 1974; both in Files, HO/ASD. For more details on the B-i's electrical multiplex system, see Philip J. Klass, "Multiplex System to Be Tested in B-i," Aviation Week and Space Technology, March 5, 1973, pp. 37-41. The weight issue is also discussed in "B-i Weight Gain to Raise Cost," 44.

Bisplinghoff

Aviation Week and Space Technology, April 9, 1973,

p. 23.

[258]

Low-Altitude Penetration

The Bisplinghoff committee's main concern, however, was not

tech-

concluded that the July 1973 decision to stretch the B-i's schedule, in conjunction with the program's various delays, would create an undesirable two-year gap between the beginning of the flight test program and the production decision. At least part of Rockwell's work force would be idle during this period, and layoffs would probably be inevitanical.

It

ble. In addition,

highly skilled engineers and workers would drift away from the project while it marked time. Production would therefore face additional start-up costs. The committee recommended adding a fourth

prototype to bridge the gap between development and production. The Air Force endorsed this recommendation, and in February 1974 McLucas petitioned Congress for funds to begin construction of a fourth prototype in 1973 and possibly a fifth in 1976. In addition, he argued that

enough

flight testing

would be completed by 1975

to justify long-lead

funding for the first production aircraft in 1976, even though a formal production decision would not have been made at that point. Congress subsequently approved construction of the fourth prototype and the procurement of some long-lead production tooling. 45 The additional prototype represented a big step toward concurrency. It increased sunk costs, contributed to the program's momentum, and made an eventual production run more likely. This, of course, was why the Air Force enthusiastically supported the idea of building more prototypes. The Air Force considered the fourth prototype to be a preproduction model, which it expected to find its way into the operational inventory. In addition to supporting the recommendations of the Bisplinghoff committee, the Air Force took steps of its own to smooth the

way

for production.

It

established a directorate of production in the

B-i project office in February 1974,

study

in

November

it

began

a

production planning

one month before the first flight of the B-i its part, had been engaged in production plan-

1974,

prototype. Rockwell, for

ning since

and

at least 1972. 46

McLucas

Letter, Feb. 6, 1974; Bisplinghoff Report, pp. 1-3, 8-11; AFSC/HO, History of the Air Force Systems Command, July 1973 to June 1974, pp. 80-84; ASD/HO, History of the

45.

DDA/HO, History of the Directorate of Development and Acquisition, July-Dee. 1973, p. 150; Cecil Brownlow, "Major Restructuring Set for B-i Project," Aviation Week and Space Technology, Feb. 18, 1974, pp. 2223; Fiscal Year 1975 Authorization, pt. 7, pp. 4084-4093. Aeronautical Systems Division, July 1973 to June 1974, pp. 126-133;

ASD/HO,

History of the Aeronautical Systems Division, July 1973 to June 1974, p. 137; July 1974 to June 1975, p. 188. Also, AFSC/HO, History of the Air Force Systems Command, July-Dee. 1975, vol. 1, p. 111; DDA/HO, History of the Directorate of Development and Acquisition, Jan. -June 1976, pp. 86-87; Donald E. Fink, "Work Begins on Fourth B-i Test Aircraft," Aviation Week and Space Technology, Aug. 25, 1975, pp. 16-18; "Air Force Seeks to Avoid 46.

Break in B-i," Aviation Week and Space Technology, July 15, 1974, pp. 117-122; "Bid for Production Funds Critical to B-i," Aviation Week and Space Technology, Feb. 1, 1975, p. 14; Robert R. Ropelewski, "North American Gears to Produce B-i," Aviation Week and Space Technology, June 26, 1972, pp. 53-59.

[259]

Flying Blind

Table 11.

B-i cost estimates (in millions of then-year dollars)

Date of estimate

Development

Production

$2,683.0

$ 8,533.8 10,884.2

June 1970 Sept. 1973

June 1974

2,787-7 3,615.6

Sept. 1975 Sept. 1976

3,884.0 4,222.0

Compiled from GAO, B-i March 1975, p. 68; March 1976, p. Source:

program

$11,218.8 13,671.9 18,632.6

15,917.0 17.312.0

21.196.0

18,667.5

22,889.5

Aircraft Program, 53;

Total

March

March

1974, p. 61;

1977, p. 34.

The B-i drew more flak from Congress in 1974 as its cost estimates grew and delays accumulated. Although the actual cost of development was definitely on the rise, it was the estimated cost of the entire program, including production, that attracted the most attention from Congress and the media. The General Accounting Office's annual reports on the B-i indicated that its cost estimates grew radically over the course of the 1970s, from $11.2 billion to $22.9 billion (see Table 11). Some analysts believed that each B-i would cost $100 million, a figure that made even

the B-i's strongest supporters blanch. 47

As these estimates appeared,

more questions were raised in Congress about the desirability of building the B-i. Was a penetrating bomber needed in the force structure? If so, was the B-52 adequate? Would the B-i be able to penetrate against advanced look-down/shoot-down air defenses designed to guard against low-altitude systems, or would a more capable system be needed? Would the Air Force be forced to rely on hard-to-detect airlaunched cruise missiles (ALCMs) in the face of these defenses? How cost-effective was the B-i compared to a force of B-52S, B-52S armed with ALCMs, or a system designed for cruise-missile carriage? Even if it was needed, was the B-i affordable, given its escalating cost estimates?

The Corona Quest Review

The Air Force recognized that the situation was becoming critical: cost problems were placing the B-i in jeopardy. This problem would have to be confronted even if production was approved, though, because B-i procurement would consume too much of the Air Force's budget and hurt other vital programs, such as the F-15 program. It was with this looming problem in mind that the Air Force vice chief of staff established

47.

Clarence A. Robinson,

and Space Technology, March

]r.,

"USAF Manned Bomber Need

24, 1975, pp. 18-19.

Challenged," Aviation Week

A

Lout-

ltitude Penetration

a high-level B-i review board in

board, as

it

came

to

September 1974. The Corona Quest

be known, was to scrutinize the B-Ts operational

requirements, system specifications, and procurement strategy to identify expensive elements that could be eliminated, replaced, or reconfigured. 48

The board completed

its

review

in

December and issued some

extraordinary recommendations.

main recommendation, which was subsequently implemented, was to replace the B-i's movable engine inlets with fixed inlets. This would simplify the system and save $230 million. In addition, it would reduce the weight of the aircraft by 1,200 pounds, which would help preserve its range. Switching to the simpler inlet, though, would subIts

reduce the B-Ts top speed at high altitudes from Mach 2.1 to Mach 1.6. It was virtually unprecedented for the Air Force to suggest compromising capabilities in such a dramatic way. What made this step even more remarkable was that high-altitude speed had been a top Air Force priority for decades. In the 1930s, the Air Force had been unwilling to compromise high-altitude performance even if other performance parameters, such as low-altitude capabilities, were enhanced. In this case, the Air Force sacrificed performance for cost considerations, alstantially

did not take this step entirely of its own volition; its back was against the wall. Cutting the B-Ts speed proved to be a double-edged

though

up

it

sword, however. Although

saved money, it invited criticism that the B-i was not the bomber it had once been. And if a Mach 1.6 capability was adequate, why was the requirement set at Mach 2.2 in the first place? This suggested that the strategic rationale for the bomber was fuzzy at best and organizationally driven at worst. 49 The Air Force also decided at this time to abandon the idea of building a crew escape module for the B-i. The module was still not as aerodynamically stable as it needed to be, and work on advanced ejection seats was proceeding unexpectedly well. Eliminating the module simplified the cockpit design and saved a considerable amount of weight. Most important, it saved $70 million in production costs and $270 million in operating costs. 50 The pressure to cut costs was so intense that there was even talk of 48.

Gen.

R.

AFSC, SAC,

H.

Ellis,

it

Vice Chief of

Staff,

"B-i

Program Evaluation," Memorandum

to

Sept. 23, 1974.

ASD/HO, History of the Aeronautical Systems Division, July 1974 to June 1975, pp. 171175, 191-194; AFSC/HO, History of the Air Force Systems Command, July 1974 to June 1975, vol. 1; "USAF Presses B-i Cost Effort," Aviation Week and Space Technology, Oct. 28, 1974, p. 22; "USAF Cuts B-i Maximum Speed," Aviation Week and Space Technology, June 16, 1975, 49.

DOD

Appropriations for 1976, Hearings before the House Appropriations Committee, p. 18; 94th Cong., 1st sess., pt. 9, pp. 193-197; Fiscal Year 1976 Authorization, Hearings before the Senate Armed Services Committee, 94th Cong., 1st sess., pt. 10, p. 5533. 30. See the sources cited in note 49. Also, John L. McLucas, Letter to John C. Stennis, Chairman Senate Armed Services Committee, Oct. 23, 1974, Files, HO/ASD.

Flying Blind

wing with a fixed wing. This change would design enormously and saved a great deal of

replacing the variable-sweep

have simplified the

B-i's

would have eliminated the B-Ts high-altitude supersonic capability altogether. Although at least 15 percent and perhaps as much as 40 percent of the B-Ts cost could be traced to the movable wing, there was a limit to how much of this investment could be recouped at this late stage. Switching designs would still have saved some money, but it would have also disrupted the development effort that was already well under way. The Air Force consequently decided against making this weight, but

it

design change. 51

The Production Decision

The net

design changes was that the procurement costs of the B-i were cut by an estimated $1 billion. 52 Even so, the political fortunes of the program continued to fluctuate. The first flight took place in December 1974, and as the flight test program proceeded the debate over production heated up. Although the Air Force believed effect of the late

that the results of the flight test

program confirmed the wisdom

of going

ahead with the B-i, production cost estimates continued to rise through 1975 and 1976. As a result, the cost-effectiveness of the bomber came increasingly into question.

A

critique of the

51.

DORDP/HO,

program published by the Brookings Institution proved especially influential in the debate, and in Congress, Senators William Proxmire and George McGovern took the lead in opposing the B-i. The Air Force weighed in with its own joint Strategic Bomber Study, while Senator Barry Goldwater pushed the case for the B-i on the Hill. 53 Although the lame-duck Ford administration formally decided in December 1976 to build the B-i, Congress was more equivocal. It placed a Histon/

of the Directorate of

Operational Requirements and Development

Plans, Jan. -June 1970, p. 270. 52. B-i Development Concept Papier, 1976, pp. 11-12; Systems Division, July 1974 to June 1975, pp. 171-175.

ASD/HO,

History of the Aeronautical

For an overview of the B-i debate, see John F. McCarthy, Jr., "The Case for the B-i Bomber," and Archie L. Wood, "Modernizing the Strategic Bomber Force without Really Trying: A Case against the B-i Bomber," along with their responses to each other, in 53.

The Air

Bomber Study can be found in DOD Appropriations for 1976, pt. 9, pp. 162-185. The highly influential Brookings study is Alton H. Quanbeck and Archie L. Wood's Modernizing the Strategic Bomber Force (Washington: The Brookings Institution, 1976). See also the Air Force's response to the Brookings study and the testimony in Fiscal Year 1977 Authorization, Hearings before the Senate Armed Services Committee, 94th Congress, 2d sess., pt. 11, pp. 6077-6108; Hearings on Military Posture and H.R. 3689, House Armed Services Committee, 94th Cong., 1st sess., pt. 1, pp. 1663-1706. Also, U.S. Comptroller General, Study of Alternative Courses of Action for the Strategic Manned Bomber, Report B-178845, April 1974. International Security (Fall 1976), 78-122.

Force's Joint Strategic

[262]

Low-Altitude Penetration

spending

limit

on

B-i production so that the

incoming Carter admin-

would have an opportunity to review the production decision. After several months of deliberation. Carter decided in June 1977 to rescind Ford s production commitment and cancel the program, rejecting the option of building just 120-150 B-is at the same time. His assessment was that a new penetrating bomber was not urgently needed and that B-52S armed with ALCMs constituted an adequate, cost-effective istration

alternative to the B-1. 54

Assessing the Development Program

Although the B-i was seen by many as a debacle, a few points about its development program should be emphasized. It experienced relatively few technical problems, and it did an adequate job of meeting its critical performance requirements. The kinds of catastrophic problems that plagued the B-58 and B-70 programs did not occur. Rockwell's problems with prototype assembly were more managerial than technical. And, although the crew escape module did have fundamental problems, it was possible to jettison this subsystem without derailing the whole program. Some performance capabilities were intentionally sacrificed to keep the cost of the program politically acceptable, but it would be a mistake to categorize these developments as performance failures. The main problem in this regard was a function of the B-i's weight, which grew from 360,000 to 395,000 pounds over the course of development. This extra weight reduced the B-i's range by 8 percent and increased its take-off distance from 6,500 to 7,500 feet. 55 These were not negligible considerations, but even outside reviewers such as the Bisplinghoff committee acknowledged that the B-i would still be able to perform the mission for which it was designed. The program was not a failure from a performance standpoint. The B-1 development schedule did not fail catastrophically either. Although the program was delayed by assembly and subsystem installation problems, funding instability also accounted for some delays. In any event, the first flight of a B -1 prototype took place just four and onehalf years after full-scale development began, which compares well to the B- 52 's timetable. This is not to say that the B-1 program was enor54.

George C. Wilson, "Ford Orders

B-i

Bomber Production," Washington

Post,

Dec.

3,

1976, p. 1; U.S. Comptroller General, Status of the B-i Aircraft Program, Report PSAD-77-35, Feb. 1977; Thomas P. Cronin, Options Available to the President in Deciding the Future of the B-i

Bomber Program,

CRS

Report 77-1

tion," Washington Post, July

1,

Dec. 16, 1976; Austin Scott, "Carter Halts B-i Produc1977; Fred Barnes, "Carter Switched Twice before Calling

Halt," Washington Post, July

1,

1977.

55.

1F,

Quanbeck and Wood, Modernizing

[263]

the Strategic

Bomber

Force, pp. 11, 28.

Flying Blind

meeting its deadlines; both the first flight and the production decision were delayed repeatedly by assembly and funding problems. But the B-l program did not experience the kinds of massive, chronic delays that bedeviled the B -70 program throughout its ten-year development ordeal. The B-i program has also been unfairly lambasted for its cost history. Most of the attention has focused on the cost estimates for the overall development and production program, which did indeed go ballistic in the 1970s, from $11.2 billion in 1970 to $22.9 billion in 1976 (see Table 11). As a result, the program's estimated unit costs rose from around $46 million to almost $100 million. These figures are misleading in two critical respects, however. First, they are simply estimates of what the program would have cost if a complete production run had taken place. They should not be confused with the actual costs of the development program or with actual cost overruns. Second, these estimates do not take inflation into account. The inability of economic forecasters to accu-

mously successful

in

rately predict the rate of inflation in the 1970s accounts for a great deal of

the growth in these estimates. After correcting for inflation, the cost

estimates for the procurement program rose by roughly 12 percent between 1970 and 1976, a far cry from figures that suggested the program's cost

had doubled or even

tripled. 56

Cost estimates did go up as the 1970s

important to distinguish between increases that should be attributed to flaws in the development program and those that should be attributed to flawed economic forecasts. In the end, more than $6 billion was spent on B-i research, development, engineering, testing, and preproduction activities. 57 To be sure, the B-i's development program was not a roaring success. It had performance slippages, delays, and cost growth. But, at the same time, it was not the unmitigated disaster its detractors made it out to be. Its track record was mixed and not entirely undistinguished. In many respects, the B-i program was a qualified success. This does not mean, however, that the B-i should have been built. President Carter's decision was based on factors besides those analyzed in this case study, such as unfolded, but

it

is

—

the cost-effectiveness of the B-i in comparison to other force structure

The debate in the 1970s naturally featured these other considerwhich is why the B-i development program itself has been mis-

options. ations,

understood. AFSC/HO,

History of the Air Force Systems Command, Jan. -Dec. 1976, vol. 1, pp. 191193; B-i Development Concept Paper, 1976, pp. 13-14. 37. Four prototypes were built and flight testing of the B-i continued through the late 1970s; see Jeffrey M. Lenorovitz, "B-i Proposed as Core Aircraft," Aviation Week and Space 56.

Technology, Sept. 17, 1979, pp. 14-15; Herbert J. Coleman, "Rockwell Applies ence," Aviation Week and Space Technology, Aug. 1, 1983, pp. 34-36.

B-iA Experi-

[264]

—

Low-Altitude Penetration

Conclusions

One

of the

most

striking things about the origins of the B-i

is

the

insignificant role played

by technological and economic factors in the early years of the program. The Air Force did not start a new program in the late 1950s, for example, when important advances were made in the

development

of the variable-sweep wing.

Nor did

it

focus

its first

round

of design studies in 1961-62

the Air Force got around to

on variable-sweep technology. By the time incorporating the swing wing into the design

variable-sweep technology was old hat. Clearly, this process was not driven by the Air Force's rush to exploit a new, hot technological opportunity. In addition, there is no indication that the aerospace industry played a leading role in the early years of the B-i effort in 1963/

program. Indeed, the Air Force's first design studies for a low-altitude bomber were in-house efforts that did not involve contractors from the aerospace industry. The Air Force was responsible for initiating the B-i program, and it was driven by a combination of strategic and bureaucratic considerations.

Although

it

recognized throughout the 1950s that high-altitude

operations would become increasingly problematic, it remained fully committed to its traditional line of thinking even after the U-2 incident in i960. The Air Force did not modify its course in a major way until political circumstances compelled it to do so in the early 1960s. Even then, it did not abandon the idea of building a new, high-altitude bomber; it simply added low-altitude requirements. Similarly, the Air Force's

supersonic speed requirements persisted until they were first modified by the civilian leadership in the Pentagon in 1969 and later scaled back in a last-ditch attempt to cut costs and shore up political support for the program. The B-i was optimized for high-altitude, high-speed operations until late 1972,

when

the low-altitude penetration mission

was

given more emphasis. All of this attests to the durability of operational doctrines and organizational belief systems. Air Force thinking changed

more

—

response to external political pressure from McNamara in the early 1960s, Packard in the late 1960s, and Congress in the early 1970s than it did to intelligence data and analytic arguments about the stratein

and operational environment. The B-i program turned out to be moderately ambitious technologically. Although it did not have to contend with the B-70's high gic

Mach requirement,

dual-mission requirement involved challenges of its own. In particular, the offensive avionics system had to be sophisticated; a system designed to fly a strategic bomber at high subsonic speeds at low altitudes as well as supersonic speeds at high altitudes its

Flying Blind

The engines also had to be highly advanced, and variable-sweep wing for an aircraft as large as the B-i in-

had never been building a

built.

volved challenges as well. Because several of its major subsystems required major improvements, the B-i belongs in the fourth position on the nine-part scale of technological ambitiousness outlined in Chapter 1 (see Table

1).

Why

did the B-i have moderately ambitious development objectives, in sharp contrast to many of its predecessors? First, the B-70 experience

was counterdevelopment problems and made the program

taught the Air Force that technological adventurousness

productive if it led to vulnerable to cancellation. Second, the civilian leadership intervened in 1969 and 1970 to lower the B-Ts performance requirements and adjust its system specifications. And third, overwhelming pressure to cut the B-Ts costs led the Air Force to lower its sights in the early 1970s.

At the same time, the B-i program was guided by a mixed procurement strategy. This strategy was nominally sequential in that a discrete production decision followed the flight testing program. A great deal of emphasis was put on prototype testing, and at least the first three aircraft were legitimate developmental prototypes built on soft tooling. Paper studies were relied on for design and source selection decisions, although not for the production decision. At the same time, the B-i program featured a great deal of concurrency. The design of the system itself was set early in the development process, essentially in the early 1960s. Competition ended prior to full-scale development. Subsystem integration was emphasized throughout development. The program featured even more concurrency in its latter stages when a fourth, preproduction prototype was approved and procurement of long-lead production tooling was authorized, even though development was still under way and a formal production commitment had not yet been made; this represented a de facto commitment to production in the development phase of the program, even though this commitment was later overturned. Finally, the level of investment in the B-i program was high; over $6 billion was spent on it through the late 1970s. One might compare this to the B-47 and B-52 prototype programs, each of which cost well under $100 million. In short, the B-i program was guided by a mixed procurement strategy that exhibited five of the eight attributes of concurrency outlined in Chapter 1 (see Table 2). Therefore, the B-Ts procurement strategy was reasonably well suited to its technological challenges, which is why it did not see the cataclysmic problems faced by the B-58 and B-70. Even so, concurrency compounded the B-Ts problems in several respects. First, Rockwell would have been pushed harder to overcome its assembly and subsystem integration problems if it had faced competitive pressure; com[266]

Low-Altitude Penetration

petition at the system level

ended

however. Second, because many activities took place concurrently (system assembly, subsystem integration, and preproduction planning) Rockwell had to engage in many activities at the same time. It would have been better able to solve the program's more pressing problems, such as those involving the variable-sweep wing's carry-through structure in the fuselage, if it had not been distracted by other responsibilities. Third, the fact that the design of the B-i was frozen early in the development process came back to haunt the Air Force in 1974, when it was desperately searching for ways to reduce the program's costs. Switching to a fixed wing would have been the most effective way of simplifying the design and slashing costs, but the time for taking such a step had come and gone. In the B-47 and B-52 programs, major design changes like these including wing changes were made several times fairly late in the development proin 1970,

—

—

cess. Finally, the fact that $6 billion

was invested

in the

program before

a

production decision was made can be traced directly to the amount of concurrency in the program: four prototypes, subsystem integration, and preproduction funding were expensive. Organizational preferences for concurrency eventually overwhelmed the efforts of even one of the Pentagon's most forceful policy reformers, David Packard. His attempts to impose a more sequential strategy on the B-i program in 1970 were, at best, only partially successful. He reduced the number of prototypes in the development program from five to three, to keep costs under control, but a fourth aircraft found its way back into the program after he left the Pentagon in 1971. In addition, his campaign to ensure that the program contained a discrete production decision was undermined by the level of investment in the program and by the variety of preproduction activities that began before the production decision was formally made. Again, many of these developments took place after Packard left office. Although Packard influenced the course of the B-i program in some respects in the short run, organizational preferences reasserted themselves in the long run. 58 In the end, final

the B-i's procurement strategy did

what

a

truly sequential strategy

would not have done: it biased the production decision. The importance of organizational preferences was especially striking in the case of the B-i program. The Air Force's commitment to the B-i triumphed over Robert McNamara's outright opposition to it, David Packard's attempt to shape it, and, ultimately, Jimmy Carter's decision to cancel

it.

Packard might have been more successful if he had imposed a sequential strategy on at the beginning of the design competition in 1969. When he tried to change course in 1970, he found it hard to make a sequential strategy stick because the program was already picking up momentum. 58.

the

program

The

Politics of Stealth:

The B-iB and B-2

Modernization of the U.S. strategic bomber force has been controversial for decades. Although new bomber programs thrived in the early postwar years, they have fared less well in the missile age. The missilegap scare of the late 1950s gave a boost to American ICBM and SLBM programs, and the U-2 incident of i960 dramatized the apparent obsolescence of the manned bomber. The Kennedy administration concluded that, at the very least, the manned bomber was not particularly cost-effective, and with that in mind it canceled the B-70 in 1961 and closed the B-52 and B-58 production lines in 1962. In 1965, Secretary of Defense McNamara proposed building a small fleet of FB-ms to replace the huge numbers of B-47S, B-52S, and B-58S that were gradually being retired from the force structure. Although the B-i program moved ahead in 1970, it was canceled in 1977 by President Carter. As a result, the number of medium- and long-range bombers in SAC's active inventory decreased from a peak of 1,854 m *959 to under 500 in the 1970s. One might have concluded in the late 1970s that the manned bomber was dying out. Within just ten years, however, the situation had completely turned around. President Reagan resurrected the B-i in 1981, and within months production contracts had been signed to build a modified version of the B-i, which the Air Force called the B-iB. Reagan also decided to proceed with development of the stealth bomber, so known because it was designed to slip by enemy air defenses undetected. The stealth 1

1.

SAC/HO,

Development of the Strategic Air Command, 1946-1986, Sept. 1986,

[268]

p. 82.

The

bomber

is

Politics of Stealth

now known

formally as the B-2. By 1988, the

last of

100 B-iBs

had rolled off the assembly line, and production activities were well under way in the B-2 program. Although the American strategic bomber force was being systematically modernized for the first time in over twenty-five years, controversy still reigned over the programs themselves. The B-iB had serious problems with its avionics systems, especially its defensive avionics system, long after it entered service in 1986. These problems could be traced to the fact that the B-iB program, which featured as much if not more concurrency than any acquisition program in history, was also deceptively ambitious technologically. The procurement program for the technologically exotic B-2 also featured a great deal of concurrency.

Origins of the B-iB and B-2 Programs

Reemergence of the B-i Option After the B-i

was canceled

in 1977, President Carter

and the Air Force

continue flight testing to squeeze as much information as possible out of the enormous investment that had already been made.

agreed

to

They disagreed, however, over how many prototypes were needed

for a

reasonable flight testing program. Carter believed that the Air Force could get by with four. He recognized that building two additional prototypes would keep the B-i production line open until 1980, which

meant

that the production issue

would continue

to

loom over him.

was anxious to bring the B-i debate to a final resolution as possible. The Air Force, on the other hand, wanted six proIt argued that keeping the B-i production option open would administration more leverage in strategic arms control negotia-

Naturally, he

soon as totypes.

give the

tions with the Soviet Union.

contracts that

It

also maintained that canceling the B-i

were already signed would

ing the two additional aircraft.

cost almost as

much

as build-

hope, of course, was that Carter would eventually reverse his decision or lose the 1980 election to someone willing to breathe life back into the program. Congress continued to be equivocal about the B-i. Although it was not enthusiastic about going ahead with high-rate production, it was reluctant to close the production line altogether. It had already appropriated funds for the fifth and sixth prototypes, and it was not inclined to reverse its decision. Many in Congress believed that a B-i production option should be preserved while the arms control negotiations were Its

under way and until the ALCM, which was had proved itself. Carter consequently had to

under development, fight long and hard to get still

Flying Blind

Congress

funds for the two additional prototypes, which it did not do until February 1978. Even then, Rockwell did what it could to preserve a production option. It stored B-i parts, construction mateto rescind the

and machine

and it held together a cadre of project engineers. B-i flight testing was scheduled to continue until early 1981. 2 Although Carter opposed the B-i, he was not planning to phase out the air-breathing leg of the triad altogether. He planned to deploy large numbers of ALCMs and consequently needed a cruise missile carrier in the force structure as well. Although the B -52 would be an adequate carrier in the short run, it was not particularly cost-effective because it could carry only 12-20 ALCMs and was expensive to operate and maintain. 3 It would have to be replaced in the long run in any event because even the newest B-52S were fifteen to twenty years old. A specialized cruise missile carrier, however, would not have penetration capabilities of its own, which would make it relatively inexpensive. Moreover, a specialized carrier might carry up to 50 ALCMs, making it highly costeffective. Wide-body commercial and cargo aircraft such as the Boeing 747, McDonnell Douglas DC-10, and the Lockheed C-5A emerged as leading contenders for this mission because of their low procurement costs and substantial carrying capacities. The Air Force and Rockwell saw this mission as an opening for the B-i. They argued that commercial and cargo aircraft would be slow to get away from their bases and too soft to survive nearby nuclear explosions. Since prelaunch survivability was important, they suggested modifying the B-i, which was designed with these operational considerations in mind, for cruise missile carriage. They claimed that a stripped-down B-i would not be prohibitively expensive. The Air Force and Rockwell made rials,

tools,

several proposals of this type in the late 1970s, only to see Carter reject

them 2.

all.

4

Jim Klurfeld, "Bombers Won't Fade," Neivsday, Oct.

Attempt

to

Keep

B-i

Bomber

3,

Alive," Philadelphia Inquirer, Feb.

1977; "Senate Rejects House 2, 1978; Nick Kotz, Wild Blue

Yonder: Money, Politics, and the B-i Bomber (New York: Pantheon, 1988), pp. 175-182; Robert R. Ropelewski, "B-i Studied as Cruise Missile Carrier," Aviation Week and Space Technology, Dec. 10, 1979, pp. 48, 53-54; "B-i Tests Yield Penetrativity Data," Aviation Week and Space Technology, June 16, 1980, pp. 188-190; Stratford P. Sherman, "How Rockwell Kept the B-i

Alive," Fortune, Nov.

2,

1981, pp. 106-112. B-i flight testing

was extended when

the B-iB

program got under way in October 1981. 3. B-52GS were ultimately equipped with 12 ALCMs on external pylons. B-52HS were equipped with 8 missiles on an internal rotary launcher and 12 externally. 4. See Alton H. Quanbeck and Archie L. Wood, Modernizing the Strategic Bomber Force (Washington: The Brookings Institution, 1976), pp. 85-92; Donald E. Fink, "B-i CostEffectiveness Claimed," Aviation Week and Space Technology, Dec. 12, 1977, pp. 14-16; Ropelewski, "B-i Studied," pp. 48, 53-54; Fred Kaplan, "The Plane That Wouldn't Die," Boston Globe, May 16, 1983, pp. 10-13, 51-52, 56, 58, 68; Frank Greve, "Is the B-i a Plane Whose Time Has Come?" Philadelphia Inquirer, March 18, 1984, pp. 20-25, 29-3 o; Kotz, Wild Blue Yonder, pp. 186-194; Jeffrey M. Lenorovitz, "B-i Proposed as Core Aircraft," Aviation Week and Space Technology, Sept. 17, 1979, pp. 14-15.

[270]

The

Politics of Stealth

By the end of the decade, though, public and congressional attitudes toward Soviet-American relations and U.S. defense spending had started to shift. Detente was no longer seen as a panacea, and many were concerned about the massive Soviet buildup in offensive strategic forces. A few believed that the United States would soon be confronted by a "window of vulnerability." A growing consensus emerged for a stronger overall defense posture and, in particular, for a more robust strategic force structure. The Soviet invasion of Afghanistan in December 1979 strongly reinforced this line of thinking and made it difficult for policy makers to take a benign view of Soviet military policy. President Carter subsequently withdrew the SALT II treaty from the Senate, where ratification was still being debated, and took steps to increase defense spending. By 1980 the idea of building a new strategic bomber was being actively considered by both Congress and the White House. 5 Many policy makers had begun to share the Air Force's reservations about the

ability of the B-52 to penetrate after the late 1980s,

Soviet Union

was expected

when

the

deploy a capable look-down/shoot-down air defense system. There were said to be several good reasons for maintaining a penetrating bomber in the force structure. First, deployto

ing bombers in addition to

concentrating

all its air

ALCMs

prevented the Soviet Union from defense efforts on a single threat. Second, there

were doubts about the cruise missile's reliability; its terrain-following guidance system would have to operate over unfamiliar territory in time of war, and it would have to contend with seasonal variations in the topography of the Soviet Union caused by heavy snowfalls. In addition, the missile would have to work after being kept in storage for long periods of time, perhaps up to three years. Third, there were doubts about the operational effectiveness of cruise missiles because, although they were hard to detect, they were relatively easy to shoot down once they were detected; they could not take evasive action nor could they carry many electronic countermeasures. Fourth, the ALCM's programmed guidance system would not lend itself to attacks against the Soviet Union's growing set of mobile targets, including ballistic missiles and command centers. And fifth, deploying a new bomber might lead the Soviet Union to step up its air defense efforts, a relatively benign area of activity from the American point of view. 6 If a new bomber was needed, Congress was ready to push ahead. In July 1980 it approved an amendment to the defense authorization act 5.

"Pressure for

New Bomber

Rises in Congress," Aviation Week ami Space Technology,

Feb. 11, 1980, p. 12.

Charles Mohr, "Cruise Missile Passes Test but Its Critics Score, Too," Neiv York Times, July 17, 1983; Tony Velocci, "The Case for the Bomber," National Defense, Nov. 1981, pp. 2832, 41-42. 6.

[271]

Flying Blind

which directed the secretary of defense to begin developing a new longrange combat aircraft. This aircraft was to be capable of operating as a penetrating bomber, a cruise missile carrier, or a conventional weapons platform. A squadron was to be deployed by 1987. The options to be considered by the Department of Defense included a stretched version of the FB-111, the FB-111H; the B-i or a modified version of the B-i; and an advanced bomber that would take advantage of technological gains made since the B-i was designed. Although this appeared to be a fairly open-ended directive, it actually stacked the deck in favor of some version of the B-i: the FB-111 could not perform

all

the required missions,

an entirely new bomber could not be developed, built, and deployed by 1987. 7 Carter was caught in a bind. He could not endorse the B-i, which he was not inclined to do in any event, without reversing his 1977 decision and appearing to be indecisive. On the other hand, opposing the new project would make him appear soft on defense in the midst of a presidential campaign in which he was already being harshly criticized for his handling of U.S. defense policy in general and for his cancellation of the B-i in particular. Carter escaped this trap in August, when his secretary of defense, Harold Brown, confirmed scattered press reports indicating that significant progress had been made in developing stealth technologies. Press accounts subsequently speculated that it might be possible to build an aircraft, even a strategic bomber, that would be invisible to air defense radar. Carter's Republican opponents charged that this announcement was politically motivated. It certainly defused charges that the president had neglected national security, and it provided Carter with an implicit justification for his original B-i decision: why build a bomber based on 1960s technology when one based on revolutionary technologies was just around the corner? 8

and

in all probability

Stealth Technology

and

the Stealth

Bomber

Although the general public knew little about stealth technology in 1980, the idea of making aircraft hard to detect had a long history. Camouflage paint had been used for decades to make aircraft hard to spot visually. But once air defenses began to rely on radar-guided and heat-seeking missiles, camouflage paint was not enough. The advent of high-performance missiles, moreover, meant that aircraft speed and al7.

May 8.

New

Douglas D. Mitchell, Bomber Options 3,

1982, pp.

1,

for Replacing B-52S,

CRS

Issue Brief IB81107,

7-8; Kotz, Wild Blue Yonder, pp. 186-194.

Based on Kotz, Wild Blue Yonder, pp. 193-198; "U.S. Builds Plane That York Times,

August

21, 1980; "Carter Clarifies Position

Aviation Week and Space Technology, Sept. 15, 1980, p. 23.

on Stealth

Foils Radar,"

Aircraft Leaks,"

The

Politics of Stealth

titude could

no longer ensure

minimizing detectability became an increasingly important requirement. A few specialized aircraft, such as the U-2 and SR-71 reconnaissance planes, were designed in the 1950s and 1960s with the detection problem in mind. It was not until the 1970s, however, that minimizing an aircraft's radar reflectivity and heat emissions, measured in terms of its radar cross section and infrared signature, became major design considerations. The importance of this issue was driven home during the Vietnam War and the October War in the Middle East, where American-built aircraft equipped with advanced electronic countermeasures incurred heavy losses. It was clear that Soviet radar and weapons had become much too effective to be countered by such measures or traditional forms of self-defense. 9 A Lockheed program to build a small, experimental stealth prototype was funded soon after the 1973 war ended. The first of these prototypes flew

and

in 1977,

full-scale

survivability;

development

of a full-sized stealth fighter

began

10

This fighter, like the stealth bomber that would follow it, drew on three basic sets of stealth technologies: new designs, materials, and in 1978.

avionics.

The starting point for stealth technology was careful design of the body of the weapon to minimize its radar cross section and infrared signature. The parts of an aircraft that contribute most to its radar visibility are the sharp, angular joints between the fuselage and the wings and tail; large, flat surfaces; sharp angles exposed in the cockpit and engine inlets; and the engines themselves. Reducing an aircraft's radar reflectivity, therefore, was mainly a matter of eliminating the highly reflective corners and angles; hiding the cockpit, engine fans, and engine inlets as much as possible; and eliminating large, flat surfaces. There were two basic solutions to the latter problem. One was to break up the aircraft's large, flat surfaces into many smaller panels, or facets, none of which would generate a strong radar reflection by itself. The other approach was to round off wing and fuselage surfaces, and then to carefully blend the wings into the fuselage; the same would be done with the aircraft's tail, if it had one. Taken to an extreme, an aircraft based on this approach would be a flying wing, perhaps similar to Northrop's B-49 of the 1940s. In any event, engine exhaust would also have to be carefully screened to lower the aircraft's infrared signature. The B-52 was obviously not designed with these considerations in mind. 9. See "USAF Aircraft Destroyed in Crash Believed to Be Stealth Fighter," Aviation Week and Space Technology, July 21, 1986, pp. 22-23; EM Sweetman, Stealth Aircraft (Osceola, Wis.: Motorbooks, 1986), pp. 5-30.

The first flight of the full-sized stealth fighter, later known as the F-117A, took place 1981. The F-117A became operational in 1983, and 52 aircraft were delivered to the Air

10. in

Force by 1988.

Flying Blind

had numerous

It

flat

surfaces

and sharp angles, and

it

had eight engines

prominently spread out along its long, angular wings. 11 Radar-absorbent materials constituted a second set of stealth technologies.

Many

posites,

or glass

new materials were made of lightweight comwhich were made in a two-step process. First, carbon, graphite, fibers were woven into strips of tape or broadcloth. Then, epof these

oxy, polyester, or phenolic

together and

mold them

polymers were used

into specific shapes.

to

One

the problem of reducing an aircraft's reflectivity

bind these strips

radical

approach

was constructing

to its

airframe largely out of composite materials, rather than steel, titanium, or aluminum. The aircraft's skin could then be coated with radar-absor-

bent graphite sheets or other coatings. Honeycomb panels composed of several different layers of such material, each effective against a different radar frequency, could be used. Coating the skin of conventional aircraft with such materials was possible as well. 12 Advanced avionics systems completed the stealth package. Much was learned about the kinds of avionics low-altitude operations would require as F-16 and B-i flight testing progressed in the 1970s. As a result, a concerted effort was made to improve system performance in each of these areas. Better offensive avionics (forward-looking radar, terrainfollowing radar, terrain-avoiding radar, and other guidance and navigation systems) enabled aircraft to fly safely at lower altitudes, where they were harder to detect. Improved defensive avionics (radar-warning receivers; radar and communications jammers; chaff, flare, and decoy dispensers) were better able to confuse enemy air defense systems. Obviously, an aircraft built with a stealthy design and stealthy materials would not be as dependent on defensive avionics as a more conventional aircraft. 13

One

can readily see that it was inherently easier to build a stealth fighter than a stealth bomber: fighters were smaller than bombers, which gave them a much smaller radar cross section to begin with. But improving air defenses posed a challenge to both tactical and strategic aircraft.

American

intelligence reports in the late 1970s indicated that the

See Malcolm W. Browne, "America's Invisible Airplane: Its Principles Are Clear Enough," New York Times, Sept. 7, 1980; Drew Middleton, "Stealth Technology: Progress in Concealing Planes," New York Times, May 31, 1983; Ralph Vartabedian, "Stealth Technology,'' Los Angeles Times, May 16, 1988; Sweetman, Stealth Aircraft, pp. 31-58; Jay H. Goldberg, "The Technology of Stealth," Technology Review, May-June 1989, pp. 33-39. 12. Based on Ralph Vartabedian, "Composites Take Off," Los Angeles Times, Oct. 30, 1989; James Kitfield, "Concern over Composites," Military Forum, Jan. -Feb. 1988, pp. 3811.

41.

Richard Carroll, "Facing the Manned Bomber Dilemma," Defense Electronics, May 1981, pp. 100-104; William A. Davis, "Principles of Electronic Warfare," Microwave Journal 23 (Feb. 1980), 52-59; Philip J. Klass, "Introduction of Stealth Will Change Avionics Needs," Aviation Week and Space Technology, March 18, 1985, pp. 233-237. 13.

I274]

The

Politics of Stealth

Soviet Union had

begun

an airborne warning and control system capable of tracking low-flying targets, fighters equipped with lookdown/shoot-down radar and missile systems capable of engaging lowflying targets, and an advanced low-altitude surface-to-air missile systo test

tem, the SA-io. Deployment of defenses such as these would

make

any bomber that relied exclusively on low-altitude capabilities. The Air Force, which was not at all opposed to building a new penetrating bomber, began preliminary design studies of a stealth bomber in the late 1970s. 14 Reports about these studies and the stealth fighter program soon began to circulate in the trade press. These were the reports confirmed by the Carter administration in August 1980. In September 1980, Lockheed and Northrop were asked to submit survival difficult for

proposals for a stealth bomber. Lockheed's bomber design resembled its stealth fighter in that it relied on obliquely angled facets to keep its radar cross section to a minimum. Northrop, drawing on its experience with the B-35 and B-49 in the 1940s, designed a stealthy flying wing.

Weighing the Options

As

turned out, the fate of the Lockheed and Northrop proposals and other options in the bomber modernization competition was to be decided by the incoming Reagan administration. Policy makers in Washington had six main bomber modernization options from which to choose in it

1981.

The first option was to do nothing. Reagan, however, believed that American defense programs had just suffered from a decade of neglect, and he was a staunch advocate of a vigorous military buildup. He believed, moreover, that immediate steps were needed to compensate for the large numbers of advanced strategic systems the Soviet Union had deployed in the 1970s. Since the Trident submarine program could not be accelerated to any significant degree and the MX CBM program was floundering, the Reagan administration emphasized bomber modernization. As far as it was concerned, at least one new bomber had to be added to the American strategic force structure. Congress, for its part, had already decided that a new bomber was needed by 1987. The second bomber option was to build the FB-111H instead of Rockwell's modified version of the B-i (known as the B-iB) or the stealth bomber. This option received little support because the FB-111, no matter how many times it was stretched, was still a fighter modified to fly 1

Sweetman, Stealth Aircraft, pp. 72-73; Rick Atkinson, $70 Billion Muddle," Washington Post, Oct. 8, 1989.

14.

to

"Stealth:

From 18-Inch Model

Flying Blind

strategic missions. Its inherent range, payload,

and avionics

limitations

simply could not be overcome. The FB-nTs heavy reliance on tanker support and its limited payload made it a relatively cost-ineffective system, and its limited avionics capacity raised questions about its ability to contend with sophisticated Soviet air defenses. The FB-111H had many of the same limitations. 15 In addition. Congress had specified that the new long-range combat aircraft be a capable cruise missile carrier, which the FB-111H would not be. The Reagan administration believed, in any event, that more needed to be done to redress the strategic balance than deploying a few squadrons of modified fighters. The third bomber modernization option was to build just the B-iB. Few supported this idea even though the B-iB would be much more capable than the FB-111H; it was also less expensive and would be available sooner than the stealth bomber. Most policy makers in Washington were captivated by the lure of the stealth bomber's exotic technol-

and projected penetration capabilities. The stealth bomber would undoubtedly be a much better penetrator than the B-iB, and the Air Force was determined to buy the best penetrator available. The stealth bomber would also have a dramatic impact on the strategic balance if it worked as promised, which appealed to the Reagan administration. Finally, Democrats in Congress were desperate to be seen as strong supporters of at least one strategic system; the B-i was thought of as a "Republican bomber," and, since the stealth bomber surfaced during the Carter administration, Democrats on Capitol Hill tried to claim it for themselves. In short, the stealth bomber was the consensus choice of the American defense policy community as the bomber of the future. Therefore, many in Washington favored a fourth option: skipping over both the FB-111H and the B-iB and proceeding directly to the stealth bomber. The idea of leapfrogging the B-iB and FB-111H had several advantages. The stealth bomber would undoubtedly be the most capable penetrator of the lot, and unlike the B-iB and FB-111H it would probably be an effective penetrator for many years; since it was being ogies

designed from scratch with stealth in mind, it could take advantage of the full range of stealth technologies. The B-i and FB-m, on the other hand, could not exploit many stealth advances because their basic designs were already set. Supporters of this option argued that it made little sense to spend tens of billions of dollars on B-iBs and FB-mHs, which could not be deployed until 1986 anyway, when the stealth bomber could be fielded soon thereafter. They believed that these interim systems would become obsolete soon after the B-52 in any event. Finally,

15. John J. Kohout, III, "Post-B-i Look at the Review, July-Aug. 1979, pp. 41, 45, 48.

Manned

Strategic

Bomber," Air University

[276]

The

Politics of Stealth

they maintained that there was no need for an interim bomber because late-model B-52S would be effective penetrators until the end of the decade. 16

The idea of skipping over the B-iB and FB-111H had serious political and strategic drawbacks as far as the new administration was concerned. This option's main political liability was fairly obvious: it would not distinguish the policies of the

new

administration from those of the old

dramatic way. This option's main strategic liability was that a stealth bomber could not be deployed until 1987 at the earliest. 17 So, even if a highly compressed stealth program proceeded smoothly and achieved an initial operating capability in 1987 (a mighty assumption), it would not redress the strategic balance in the short run. Many in the adminin a

istration believed that this

was

a critical issue

because improving Soviet

defenses were beginning to make the B-52 obsolete as a penetrator; some estimates indicated that older B-52S would start to become ineffec-

air

tive in the

mid-1980s.

If

this

was

etration capability in the force

and if maintaining a robust penstructure was absolutely necessary, then true,

skipping ahead to the stealth bomber involved some risks. For those who wanted to deploy an interim bomber as well as the stealth bomber, the debate boiled down to the relative merits of the FB-111H and the B-iB. Therefore, a fifth option in the bomber modernization debate was to deploy the FB-111H as an interim system before proceeding to the stealth bomber. This option was favored by many in the Air Force, including Gen. Richard Ellis, SAC's commander. 18 At first glance, this might seem incongruous; SAC had never been enthusiastic about the original FB-111, which had been imposed on the Air Force by See William Kaufmann, “The Defense Budget," in Joseph A. Pechman, ed.. Setting The Brookings Institution, 1983), p. 60; Hodding Carter III, “MX and B-i: Two Very Expensive Molehills," Wall Street Journal, Nov. 12, 1981; Gordon Adams, “The B-i: Bomber for All Seasons?" Council on Economic Priorities Newsletter, Feb. 1982, pp. 1-6. See also Alton K. Marsh, "Bomber Faces Opposition in House Funding Group," Aviation Week and Space Technology, Oct. 12, 1981, pp. 23-25; Kotz, Wild Blue Yonder, pp. 200-218. 16.

National Priorities (Washington:

Northrop is said to have claimed that a stealth bomber could begin flight testing in 1984 or 1985, and that a squadron of bombers could be fielded by 1987. The Air Force was also aiming for deployment in 1987. Most disinterested observers estimated that a stealth bomber could be deployed in 1989 or 1990 at the earliest; see Kotz, Wild Blue Yonder, p. 206; Mitchell, Bomber Options, pp. 13; Modernization of the U.S. Strategic Deterrent, Hearings before the Senate Armed Services Committee, 97th Cong., 1st sess., Oct. -Nov. 1981, 17.

pp. 5-6, 23; Strategic Force Modernization Programs, Hearings before the Senate Armed Services Committee, 97th Cong., 1st sess., Oct. -Nov. 1981, pp. 311, 329-333; John L. McLucas, “The Stealth Bomber Won't Fly for a Decade," New York Times, July 7, 1982; “Look Who's Heading For No. 1 in Defense: Northrop," Business Week, April 19, 1982, pp. 70-79; George C. Wilson, “Pentagon Wants Stealth Date to Become Invisible," Washington Post, 18.

B-i a

March

13, 1982.

See Kotz, Wild Blue Yonder, p. 209; Sweetman, Stealth Aircraft, p. 73; Greve, "Is the Plane Whose Time Has Come?"; Senate Armed Services Committee, Strategic Force

Modernization Programs, pp. 364-390.

[277]

Flying Blind

McNamara in the 1960s. Since the FB-111H would FB-m's operational limitations, why did SAC prefer it

Secretary of Defense

have

many

to the

of the

B-iB as an interim bomber?

SAC's overriding interest was in preserving the penetration mission, so its primary objective in the bomber modernization debate was getting the stealth bomber. The stealth bomber, unlike its competitors, would almost certainly be viable as a penetrator into the next century. SAC preferred the FB-111H in the short run, not because it was the best nearterm strategic option clearly, it was not but because it posed less of a threat to the stealth bomber in the long run. In fact, SAC preferred the FB-mFi precisely because it was less capable than the B-iB: it was hard to imagine the FB-111H ever supplanting the stealth bomber in the force structure because it was not, in the final analysis, a true strategic bomber. The B-iB, on the other hand, was a legitimate strategic system, with impressive range, payload, penetration, and cruise missile carriage capabilities. The existence of the B-iB in the force structure would inevitably weaken the case for the stealth bomber. Since the B-iB would have impressive cruise missile carriage capabilities, deploying it instead of the FB-111H would put an ALCM alternative to the penetrating bomber in the force structure. SAC's worst-case scenario was that deploying the B-iB would lead to cancellation of the stealth bomber; when the B-iB could no longer penetrate, it would have to get by as a cruise missile carrier, which would leave SAC without a penetrator in the force struc-

—

—

ture.

many

headquarters in the Pentagon favored a sixth option: rushing the B-iB into production while proceeding with development and preliminary production work on the stealth bomber. 19 Supporters of this option argued that the B-iB promised to be a highly capable system in several important respects. It would certainly be a better low-altitude bomber than the B-52 and a better all-around strategic system than the FB-111H. Moreover, it would be a much better penetrator than the original B-i because its design would be modified to reduce its radar reflectivity: the wings and fuselage would be blended together more smoothly to eliminate sharp joints; the cavity around the variable-sweep wings would be shielded more thoroughly; the engines Still,

at Air Force

would be hidden more

and baffles would be designed for the engine inlets. In addition, Rockwell planned to use radar-absorbent materials on the engine inlets and on the leading and trailing edges of the wings and tail. 20 Although the B-iB was not the stealth bomber, it was a 19.

See Sweetman,

carefully;

Stealth Aircraft, p. 73; Kotz, Wild Blue Yonder, pp.

Whose Time Has Come?" "A New B-i Bomber May Take off Soon,"

200-218; Greve, "Is

the B-i a Plane 20.

Richard Halloran, "Struggle

May 25, 1981, Shaping Up on Bomber Award," New York Times, Business Week,

pp. 160, 163; Oct. 6, 1982;

[278]

The

Politics of Stealth

bomber. The B-iB's design, however, was still based on that of the original B-i, so there were limits to what these refinements could accomplish. The net effect of these relatively modest design changes was nonetheless striking: the head-on radar cross section of the B-iB was roughly 10 percent of the original B-i's and 1 percent of the B-52's. Since the B-iB could carry a large defensive avionics suite, it was expected that its advanced electronic countermeasures could achieve the jamming levels necessary to mask the aircraft's remaining reflectivity. It was expected, therefore, that the B-iB would be an effective penetrator even against the look-down/shoot-down air defenses the Soviet Union was expected to deploy in the late 1980s. Although Reagan administration officials gave conflicting testimony about the B-iB's long-term effectiveness as a penetrator, they eventually concluded that it would be a viable penetrator well into the 1990s. Some supporters of the program claimed that it would still be a capable penetrator in the first decade of the twenty-first century. Supporters of the B-iB also noted that it could carry 22 ALCMs, making it an efficient cruise missile carrier when its days as a stealth

penetrator were over. 21

Supporters of the B-iB argued, moreover, that it could be pushed into production immediately, which was critically important for two reasons. First, it meant that the B-iB could be deployed in the mid-1980s, when

some of the older B-52S would start to become obsolete as penetrators. The B-iB option could therefore compensate for the force structure's eroding penetration capability better than any of the other options. Second, the B-iB would be virtually impossible to cancel by 1984 or 1985, because so much money would have been spent on production by then. The stealth bomber, on the other hand, could not be pushed into full-

development and production immediately, which meant that important decisions on the stealth project would have to be made after 1984, perhaps by another president. B-iB supporters were determined to get a new bomber while Reagan was in office, and they were not willing to risk everything on the 1984 election. If Reagan lost in 1984, they did not want to be left empty-handed or with just the FB-111H, which they thought was almost as bad. In October 1981 Reagan announced his decision to go ahead with both scale

Sweetman,

Stealth Aircraft, p. 75; Vartabedian, "Stealth

Technology." Also, Kitfield, "Con-

cern over Composites." 21. The radar cross sections of the B-52, B-i, and B-iB are discussed in Middleton, "Stealth Technology"; Mitchell, Bomber Options,

p.

16;

"Look Who's Heading

for

No.

1

in

Armed

Forces Committee, Strategic Force Modernization Programs, effectiveness of the B-iB as a penetrator in the 1990s was pp. 93, assessed by various administration officials in Senate Armed Services Committee, Strategic

Defense,"

p. 72;

Senate

161, 303, 355.

The

Force Modernization Programs, pp. 36, 124, 132, 157, 264, 276, 305, 308, 318-319, 355, 359, and Modernization of the U.S. Strategic Deterrent, pp. 26-27, 43-4 7-

Flying Blind

bomber. The Air Force was directed to move ahead expeditiously with production of the B-iB, which was expected to be operational in squadron strength in late 1986. A total of 100 B-iBs were to be delivered to the Air Force by early 1988. 22 Development of the stealth bomber would continue, and production would begin when the B-iB program wound down. It was expected that the stealth bomber would be operational in the early 1990s, generally assumed to be 1991 or 1992, and that 132 B-2S would eventually be deployed. The administration's plan was to retire old B-52DS and transform most of SAC's B-52GS and B-52HS into cruise missile carriers over the course of the 1980s. B-iBs would begin to take over the B-52's penetration mission starting in 1986. When the B-iB's effectiveness as a penetrator began to erode in the 1990s, it would be used more as a cruise missile platform, and the last of the B-52S would be retired. The B-2 would be in place by then to take over the penetration mission. 23 Rockwell was selected to build the B-iB, and Northrop was tapped to develop and produce the stealth bomber. the B-iB

and the

stealth

The B-iB Acquisition Program From

program was to proceed as quickly as possible. The Reagan administration wanted to begin strengthening the American nuclear arsenal without delay, and the B-iB was the centerpiece of its strategic modernization program. The administration was also under pressure to begin deploying the B-iB by 1987 to meet the long-range combat aircraft deadline Congress had set in 1980. The Air Force was anxious to push the B-iB along as well, although for somewhat different reasons. It wanted a production program that would be virtually impossible to cancel even if Reagan lost the 1984 election, and it recognized that the B-iB had to stay far ahead of the B-2 in the procurement pipeline; its strategic rationale would collapse if its deployment data slipped while the stealth bomber's moved up. the beginning, the overriding goal of the B-iB

Concurrent Development and Production

The administration and Air Force agreed that the B-iB would be developed, produced, and deployed on a highly accelerated schedule. Once To get approval for the B-iB in 1981, the Reagan administration had to guarantee that production would be stopped at 100 aircraft, even though this was uneconomical. The stealth bomber's supporters opposed an open-ended B-iB production program because it would have posed a threat to stealth production. 23. Senate Armed Services Committee, Modernization of the U.S. Strategic Deterrent, pp. 5-6; CBO, The B-iB Bomber and Options for Enhancements, Aug. 1988, pp. 5-7. 22.

The

Politics of Stealth

Congress approved the administration's two-bomber plan in December, the Air Force wasted no time setting up a crash program to build the B-iB.

signed the necessary contracts with Rockwell in January 1982 and established a series of demanding deadlines for the program itself. The first flight of a B-iB prototype was to take place in October 1984; the first squadron of 15 B-iBs was expected to be operational by October 1986; and all 100 aircraft were supposed to be built by April 1988. 24 Thus, the B-iB was scheduled to be operational less than five years after its

It

development contracts were signed;

demanding

its

timetable

was one

of the

most

in history.

To meet this timetable, the Air Force made a commitment to produce the B-iB at the outset of the development program, almost three years before flight testing of the

first

prototype.

It is

understatement to overlapped in the B-iB

a gross

say that development and production activities program: they were, quite literally, coincident. The Air Force signed both development and production contracts for the B-iB in January

Development and production contracts for all subsystems were signed at the outset of the program as well. Rockwell began to produce long-lead items and tooling for the production program almost immediately. Even the first batch of B-iBs was to be built on production tooling; there were to be no developmental prototypes. By November 1982, 16,300 of the aircraft's 18,000 production drawings had been completed; 41.000 of the program's 61,000 manufacturing orders had been released; 54.000 of its 58,000 tool orders had been sent out; and 16,000 parts had been built for the first two production aircraft. 25 Rockwell was totally dependent on paper studies and extrapolations from the original B-i program when it made its system specifications; hard data from the B-iB flight testing program simply did not exist. An enormous financial commitment was made to the B-iB in the early years of the program to fund the wide range of production activities already under way. The B-iB's fiscal year 1982 appropriation, for example, was over $2 billion, and its 1983 appropriation was over $4.6 billion. Once the 1984 budget passed, 50 percent of the B-iB's total budget had been appropriated, and 78 percent of the program's total budget was appropriated once the 1985 budget was approved. Congress approved 1982.

funding

for the final batch of 48 B-iBs in the 1986 budget.

testing did not begin until late 24.

Carole Shiffrin, "Bomber

ington Post, Sept.

2,

B-iB flight 1984. By then, 78 percent of the program's

Facility

Construction Running ahead of Schedule," Wash-

1982.

"Rockwell Signs $2.2 Billion B-i Contract," Aviation Week and Space Technology, jan. 25, 1982, p. 19; Robert R. Ropelewski, "Rockwell Begins B-iB Production Plans," Aviation Week and Space Technology, Oct. 12, 1981, pp. 26-29; "B-iB Test Pace Quickens," Aviation Week and Space Technology, Aug. 20, 1984, pp. 55-56. See also The B-rB: A Program Review, Report of the House Armed Services Committee, 100th Cong., 1st sess., March 1987, pp. 11-13; 25.

cited hereafter as

HASC

Report.

Flying Blind

budget, for both development and production activities, had already been approved. Congress approved funding for the final batch of total

B-iBs, the

end

of the production program, during the

phase of the This was a far cry from the days when Congress would not allow a production program to begin until a prototype competition had taken place and the results of flight testing had been fully analyzed. The B-iB program was, then, one of the most concurrent programs the Air Force ever conducted. It relied exclusively on paper studies for critical design decisions. Secondary subsystems were integrated into the program at the beginning of the development process. It was assumed that the design of the aircraft and its subsystems could be frozen virtually at the outset of the development process and that production activities could proceed on that basis. Production tooling and materials were procured, not just in the development phase of the program, but at the outset of the development cycle. Development and production schedules were planned in intricate detail at the beginning of the program. A substantial financial commitment was made to the B-iB in the first few years of the program. The program relied on production prototypes for hardware testing and verification of system and subsystem designs. And finally, the development and production phases of the program were highly compressed: they overlapped, to say the least. It was a challenge to keep the B-iB's budget within strict limits. The administration certified at the outset of the program that 100 B-iBs would be built for no more than $20.5 billion (in 1981 dollars), which was projected to be $28.3 billion when inflation was factored in. 27 It had to make this commitment to get congressional approval for the program in the first place. Congress probably would not have accepted the resurrection of the B-i if its budget had posed a threat to the stealth bomber, so the administration and Air Force had to set the B-iB's budget ceiling at a low level, and they had to ensure that the program met this target. They recognized that any cost growth in the program would give the B-iB's opponents an excuse to reopen the issue of bypassing the B-iB and proceeding directly to the stealth bomber. The B-iB program, therefore, was hard-pressed to meet demanding tail

first

flight testing program. 26

26.

Dagnija Sterste-Perkins, B-iB Strategic Bomber,

1987, p.

CRS

Issue Brief IB87157,

August

17,

3.

27. The Air Force originally claimed that 100 B-iBs could be built for $19.7 billion (in 1981 dollars). Provisions for carriage pushed the program's estimated cost to $20.5

ALCM

The General Accounting Office estimated in 1989 that $3.7 billion had been spent on items that were not included in the program's $20.3 billion baseline budget. These items included simulators, continuing engine development, interim contractor support, and replenishment spare parts. See U.S. GAO, B-jB Cost and Performance Remain billion (1981 dollars).

Uncertain,

GAO/NSIAD-89-33,

Feb. 1989, p. 24.

[2821

The

Politics of Stealth

deadlines and cost targets.

would

these goals were to be met, everything have to proceed as planned for the duration of the development, If

and production phases of the program. Any technical problem would set the program back; there was, as one Rockwell executive put it, "zero slack in the schedule." 28 Any major delay would set off a chain reaction, because all the elaborate plans and schedules in this highly concurrent program were interconnected. It was critically important, therefore, that the B-iB program contain few technological puzzles. testing,

Development Challenges

At

seemed to be the kind of low-risk system that could be developed and produced concurrently. After all, the B-i's design had been worked on since the 1960s, and it went through one fullscale development program in the 1970s. Four B-i prototypes had already been built, and almost 1,900 hours of flight testing had been conducted by the time the B-iB program was given the green light in late 1981. Although changes were made to reduce the B-iB's radar reflectivity, these refinements were relatively minor. They did not involve major structural changes; the level of commonality between the B-i and B-iB airframes was 80-90 percent, and the level of commonality between the B-i and B-iB engines was 95 percent. The Air Force consequently claimed that the design of the B-iB was stable and techfirst

glance, the B-iB

nologically mature. 29

The B-iB was, however, deceptively ambitious technologically, mainly in terms of its avionics systems. The offensive avionics system was a much improved version of the one used in the B-52 and the original B-i.

new

automatic terrain-following radar system, which had to work reliably for the B-iB to fly high-speed, lowaltitude missions. The B-iB also had a new flight control system, which had to keep the aircraft stable even at low altitudes with heavy loads. This was a formidable task, given that the original B-i had a design gross weight of 395,000 pounds. The design gross weight of the B-iB, however, was significantly greater than that of the B-i. The B-iB's structure In particular,

it

included a

was strengthened

ALCM

which added 7,000 pounds to the weight of the aircraft itself; 50,000 pounds to its payload; and 25,000 pounds to its fuel capacity. A fully loaded B-iB thus weighed 82,000 pounds more than the original B-i. Since a 477,000-pound aircraft would Quoted

for

carriage,

Ropelewski, "Rockwell Begins B-iB Production Plans." 29. W. H. Shelley, Jr., "The B-i Bomber Program: A New Start," GAO Report No. B-206613, April 13, 1983, p. 5, 7-8. See also Frank C. Conahan, "Analysis of DOD Request for Multiyear Contract Authority for the B-iB Weapon System," GAO Report B-206650, 28.

June

in

16, 1983, pp.

6-7;

HASC

Report, pp. 11-13.

Flying Blind

Rockwell B-iB (U.

S.

Air Force)

have different handling characteristics than a 395,000-pound aircraft, much of the flight test data accumulated over the course of the B-i

program was not applicable to the B-iB. 30 The B-iB's defensive avionics system was also a much improved version of the one used in the original B-i. It contained 118 major components, or "black boxes," as opposed to 88 in the B-Ts system and 23 in the B-52 's. In addition, the B-iB's electronic countermeasure system was given additional radar bands to cover, and it had to integrate a new tail

radar into the system. strategic

bomber

was

defensive electronics suite for a designed as an integrated system. 31 It was, according It

the

first

Gen. Lawrence Skantze, the AFSC commander, "the most flexible, robust ECM [electronic countermeasure] system we have ever built." What compounded the development problem for this system was that, as Skantze put it, "no one had ever attempted to build such a large, complex ECM system on an accelerated production schedule." 32 Another Air Force official explained, "We asked them to build a Cadillac to

Gen. Lawrence A. Skantze, "B-iB: Timely Lesson in Risk Management," Aviation Week and Space Technology, March 23, 1987, p. 11; Shelley, "The B-i Bomber Program," p. 7; Ropelewski, "Rockwell Begins B-iB Production Plans." 31. "Tests of B-iB ECM System Uncover Basic Design Flaws," Aviation Week and Space 30.

Technology, July 18, 1988, p. 31.

Skantze quoted pp. 222-223. 32.

in

"Skantze Answers B-iB

Critics," Aerospace Daily, Feb. 11, 1987,

[284]

The

Politics of Stealth

within an impossible time frame. For any system 33

like the B-i, that's a tall

countermeasure system involved many design changes, technical advances, and procurement risks. Given that the B-iB program was extraordinarily concurrent, it would have been a risky venture even if all its major design elements were known quantities. This was not the case, however, because its offensive and defensive avionics systems were fundamentally new in important order."

In short, the B-iB's electronic

respects.

Acquiring

Momentum

The B-iB debate barely touched on procurement issues in the early years of the program, mainly because there was no way to determine if the program was proceeding smoothly until flight testing got under way. Naturally, the Air Force maintained that the B-iB was on course, and all superficial indicators seemed to support this contention. The program seemed to be meeting its formal milestones. The first flight of a prototype, for example, took place on schedule in October 1984, and the program seemed to be on track to meet its 1986 operational deadline as well as

its

cost target. 34

What confounded policy makers outside the Air Force was a Catch-22: there was no way to prove that the B-iB program was experiencing serious problems until flight testing expanded in 1985-86; by that time,

however, the program was a fait accompli because the vast majority of its funding had already been appropriated. This "buy before you fly" approach virtually immunized the program against cancellation: Congress had no real ammunition to use during the few years in the early 1980s when it could have canceled the B-iB; by the time hard evidence about the B-iB became available, the program had already reached the point of no return. The debate over the wisdom of building the B-iB was moot by 1985. This was not accidental. The terms of the debate shifted once it became clear that 100 B-iBs would be built. The question then became, would Rockwell build any more bombers after its first production run was completed? One option Quoted

James M. Dorsey, "Radar Shield Problems Hold up B-iB," Washington Times, Feb. 4, 1987. For more technical details on the B-iB's defensive avionics system, see Hans Peot, "An Electrical Umbrella: The B-iB Defensive Avionics System," Defense Systems Review, May 1984, pp. 15-18. See also the testimony of the General Accounting Office's Frank C. Conahan in National Defense Authorization Act for Fiscal Years 1988/1989, Hearings before the House Armed Services Committee, 100th Cong., 1st sess., Feb. -March 1987, 33.

in

pp. 11-14; cited hereafter as Conahan Testimony. 34. The B-iB was generally seen as a well-run

program prior to 1985; see Shiffrin, "Bomber Facility Construction"; Clarence A. Robinson, Jr., "USAF, Rockwell Conduct Critical Design Review on B-iB," Aviation Week and Space Technology, Jan. 24, 1983; "Procure-

ment Success

Story," Wall Street journal, Feb. 6, 1984.

Flying Blind

was

more

Another was

an even stealthier version of the bomber, a B-iC. Supporters of these options maintained that the unit costs would be competitive because Rockwell's production line was already in place. Rockwell estimated that it could build 48 more B-iBs for $140 million each or 100 B-iCs for $100 million each. These estimates to build

B-iBs.

compared favorably

to build

to the units costs of the first 100 B-iBs,

$205 million,

or to the estimated unit costs of the stealth bomber, $277 million. 35 Supporters of the B-iB also argued that extending Rockwell's production

run would maintain competitive pressure on Northrop, which would help keep the stealth bomber on track. In addition, it would provide another procurement option in case the stealth program collapsed altogether.

opponents argued that extending its production run would keep the stealth bomber on the back burner; it might even lead to cancellation of the stealth program. This was undesirable, it was said, because Northrop's advanced bomber would be needed as a penetrator in

The

B-iB's

the 1990s. To be effective against increasingly sophisticated air defenses,

would sooner or later have to make a great leap forward technologically. The stealth bomber was the only system under development that could take full advantage of recent advances in the

manned bomber

force

stealth technology.

Opponents

production plan would inevitably generate pressure for an extended production run. According to the production schedule, one B-iB would be built in the program's first year; 7 in the second; 10 in the third; 34 in the fourth; and 48 in the fifth and final year. The program would build to a crescendo and then, theoretically, end abruptly in 1988. Roughly 60,000 employees and 5,200 contractors in 48 states would be working on the B-iB at that point. The program's critics charged that the production schedule had been set up this way precisely because it would generate pressure on Congress to keep the assembly lines open. Although the Office of the Secretary of Defense and the Air Force insisted that there were no official or unofficial plans to extend B-iB production beyond 1988, Congress was skeptical. In 1986 it decided not to include any money in the 1987 appropriations bill for an extended B-iB production run, and it directed the Air Force not to use funds appropriated for other purposes on B-iB production of the B-iB recognized that

its

activities. 36

These figures are in 1981 dollars. See Eugene Kozicharow, "Some USAF Planners Promote Extended Production of B-iB," Aviation Week and Space Technology, Feb. 25, 1986, 35.

pp. 30-51; J. Ernest Beazley, "Rockwell Bids to Sell 48 More B-is for 35 Percent Less than Northrop's Stealth," Wall Street journal, March 21, 1986; Dagnija Sterste-Perkins, Strategic Forces: B-iB Bomber, CRS Issue Brief 1684017, Nov. 13, 1986, pp. 1-6. 36. See the sources cited in note 35. The Air Force maintained that the B-iB's schedule was set up this way so that all 100 aircraft would be built by 1988, at which point produc-

[286]

The

Politics of Stealth

This directive essentially decided the issue and set the stage for production of the stealth bomber. The fact that neither Congress nor the administration was willing to compromise its commitment to the stealth

program, even in the face of enormous pressure to extend B-iB production, was a testament to the appeal of the stealth bomber's exotic technologies.

Production Problems It

was not

appear

in

problems in the B-iB program began to 1986. Opponents of an extended production run were despera coincidence that

ately searching for chinks in the B-iB's armor.

More important,

flight

was finally starting to generate a detailed picture of how well the program had actually progressed. At first, the problems seemed relatively minor and routine. For example, the bomber's fuel tanks had a series of leaks in September and October 1986. These leaks, more a testing

were corrected once tighter manufacturing tolerances and better quality control measures were instituted. 37 It soon became clear, however, that the program was experiencing much more serious problems with its avionics systems. Congress was alerted to this fact when the Air Force earmarked some of its fiscal year 1988 budget request for work on various avionics problems, which led to a General Accounting Office investigation of the program and a series of highly publicized hearings conducted by the Ffouse Armed Services Committee in February and March 1987. The B-iB's terrain-following radar and flight control system each had problems. To keep the bomber close to the ground as it flew over hills, through valleys, and across mountain ranges, the terrain-following radar had to make an accurate profile of the terrain directly ahead of the bomber. But the system frequently generated false terrain spikes; that is, it determined that there were hills, mountains, and other obstacles in the bomber's flight path when in fact there were none. And once the system concluded that there was an obstacle in the bomber's path, it automatically pitched the aircraft up to higher altitudes. These unnecessary flyups wasted fuel and exposed the bomber to detection by groundbased radar. Under some conditions, the radar system caused the aircraft to pitch down as it approached large obstacles. Obviously, both nuisance than anything

tion of the stealth

else,

bomber was supposed to begin. But, at the same time, racing to get to a aircraft per month made the program difficult to cancel in the early

production rate of four

1980s and difficult to stop in the mid-1980s. 37. "B-iBs Developing Leaks in Fuselage, nology, Sept. 29, 1986, p. 19;

Defense Week, Oct.

6,

1986, p.

Wing Tanks,"

Rowan Scarborough, 8.

Aviation Week and Space Tech-

"B-iB's, Leaks Plugged,

Go

to

Work,"

Flying Blind

problems impaired the bomber's effectiveness and undermined crew confidence, and the Air Force had to restrict low-altitude training exercises for many months. These problems proved to be stubborn; even in mid-1988 the system was operating at only 80-90 percent of capacity. Training restrictions were, however, eased in 1989 as the terrain-following radar continued to improve. 38 The basic problem with the flight control system was that it could not fully exploit the aircraft's aerodynamic lift capacity. This limited the B-iB's maneuverability, its munitions load, and its fuel load, the latter of which sharply affected low-altitude range. 39 According to Maj. Gen. Peter Odgers, the deputy for the B-iB program at ASD, aerodynamic data from the original B-i were used to calculate the flight control speci-

He

was

done on ... a piece of paper by engineers and computer modeling." The B-iB, however, weighed 82,000 pounds more than the original B-i, giving it some unique handling characteristics. The significance of this was not driven home until flight testing began in late 1984. Odgers said that the Air Force and Rockwell did not demand changes in the specifications for the fications for the B-iB.

explained that

"it

all

system prior to that time because "we thought we knew this aircraft aerodynamically pretty well. When we went to the flight test, it was an unacceptable system." 40 The flight control problem is a classic example of how flawed assumptions about technological risk can undermine a highly concurrent program. The Air Force and Rockwell had to rely on paper studies and computer analyses of the flight control system in the first few years of the program because flight testing would not begin until later. As a result, neither the Air Force, Rockwell, nor Congress knew or could know that there were serious problems with this system until very late in the game. By then, there was not enough time to test the system thoroughly and resolve its problems before the bomber entered operational service. In short, it was a mistake to try to develop and produce a system such as this concurrently. The flight control system involved too many technological advances to be rushed into production. The Air Force developed three components to improve the flight control system, two stall inhibitor systems (SIS-I and SIS-II) and a stability flight control

—

—

Based on CBO, B-iB Bomber, pp. 19-24; HASC Report, pp. 7-8; Conahan Testimony, p. 11; GAO, B-iB Cost and Performance Remain Uncertain, pp. 14-15. See also James M. Dorsey, "Air Force Trims B-iB Radar Rules," Washington Times, March 30, 1987; Molly Moore, "Defense Dept. Restricts Low-Level Flights of B-i Bomber in Wake of Crash," Washington Post, Dec. 4, 1987; William B. Scott, "B-iB Stability Enhancement Tests Verify Low-Level Mission Capability," Aviation Week and Space Technology, June 26, 1989, pp. 34-35. 39. Technical details on the B-iB's flight control system can be found in CBO, B-iB Bomber, pp. 15-19. See also HASC Report, pp. 6-7; Conahan Testimony, pp. 10-11. 40. Odgers quoted in Molly Moore, "B-i Bomber Repair Fund Is Requested," Washington 38.

Post, Jan. 7, 1987.

[288]

The

Politics of Stealth

enhancement system. SIS-I was to be incorporated into the B-iB by the time the bomber entered operational service in 1986, but it had to be redesigned three times and then retrofitted into the entire fleet, which took until April 1988. SIS-II and the stability enhancer, a computerdirected flight control system with additional sensors and improved software, were designed to provide even more precise handling at higher operational weights. These advanced systems were retrofitted into 90 percent of the fleet by mid-1990; the other aircraft in the fleet required additional modifications, and retrofitting the SIS-II and stability enhancer into these aircraft was scheduled to take until 1994. Thus, most of the fleet's operational capabilities were constrained for many years after the B-iB entered service. In addition, retrofitting was expensive. 41 The B-iB's most serious problems plagued its defensive avionics system. The individual components had been tested in the laboratory and worked well. The software needed to integrate these components into an effective system left much to be desired, however, and this was not appreciated until the entire system was assembled and flight tested. It was not until 1986, when flight testing with the electronic countermeasure system began in earnest, that the magnitude of the problem started to become clear. There was not enough time to resolve this difficulty before the B-iB entered service, so the Air Force was forced to deploy the problem-ridden system while continuing its development. Modifications had to be introduced during production, and several different defensive avionics systems were fielded between 1986 and 1988. 42 One of the first corrective actions the Air Force had to take, therefore, was standardizing the systems in service. The standardization problem was only the tip of the iceberg, however, because none of the systems in service met the Air Force's performance requirements. Two problems eventually proved to be relatively manageable. First, the system occasionally sent out a rogue signal that acted as a beacon, illuminating the bomber instead of hiding it. The Air Force developed an inhibitor that prevented the system from beaconing on those frequencies where it tended to do so, but this limitation cut into the effectiveness of the system as a whole. The defensive avionics system also suffered electronic leakage. Signals from some active jamming frequencies

A

equipped with the stability enhancer will be able to spend much more time at low altitude than the Air Force had originally required, because the aircraft will be able to carry a heavy fuel load safely; the low-altitude leg of a typical mission will be 2,900 nautical miles instead of 1,500 nautical miles. See GAO, B-iB Cost and Performance Remain Uncertain, pp. 12-14, 20 "USAF, Rockwell Test Systems Needed to Expand B-iB's Flight Envelope," 41

.

B-i B

/

Aviation Week and Space Technology,

Enhancement

HASC

November

28, 1988, pp. 82-84; Scott,

"B-iB Stability

Tests."

Report, p. 5. See also Howard Silber, "Gen. Chain Defends SAC's B-iB 42. Bombers," Omaha World-Herald, Feb. 8, 1987; "Skantze Answers B-iB Critics."

Flying Blind

bled into and

jammed

some signals interThe Air Force conse-

passive receivers. In addition,

and

fered with the offensive avionics,

vice versa.

quently had to restrict simultaneous use of the offensive and defensive avionics systems during training. The long-term solution to this problem was development of a radio frequency signal management system that coordinated the activities of the two avionics suites; the signal management system ensured that the two suites did not operate on the same

frequency at the same azimuth at the same time. Although development of this system was tripped up by a few technical complications, most of the major problems were overcome by late 1988. 43 The B-iB's defensive avionics system also had problems that limited

and defeat air defenses. This, of course, was its raison d'etre. Three problem areas stood out. 44 First, although the system could identify and locate some threats, it could not track all threats capacity to detect

its

specified in

its

requirements. Problems with the

radar were especially pronounced

aircraft's

in this regard. 45

tail-warning

Second, the system's

numbers of radar signals at the same time. As a result, the system could be overwhelmed in a highthreat environment. Third, there were problems with the active countermeasure systems. In particular, automatic jamming capability was limited; in some situations, the crew would have to jam enemy radar and receiver/processor could not handle large

missile systems manually.

The Air Force concluded the

in 1988 that these

system's architecture and

that

the

problems could be traced

current

electronic

to

counter-

measures system would be incapable of meeting its original performance requirements. It eventually proposed spending $500 million on improvements to the system which would enable it to deal automatically with the most important but not all of the expected air defense threats. Another $500 million would be spent to add a radar warning receiver to the B-iB which would alert the crew to employ electronic countermeasures manually or take other evasive action. These expenditures would drive the B-iB program $400 million over its $20.5 billion target. Flight testing of the core system was finally completed in early

—

43.

CBO, B-iB

—

Bomber, pp. 27-28; HASC Report, pp. 5-6; Conahan Testimony, p. 14; is 'Beacon' to Soviet Radar," Los Angeles Times, Feb. 26, 1987;

Gaylord Shaw, "B-i Bomber

Fred Kaplan, "B-i Problems, It Reparable, Could Cost $3 Billion," Boston Globe, Feb. 18, 1987; GAO, B-iB Cost and Performance Remain Uncertain, p. 15. 44. Based on CBO, B-iB Bomber, pp. 12-15; HASC Report, pp. 5-6; Conahan Testimony, pp. 11-14; Eliot Marshall, "Bomber Number One," Science, Jan. 29, 1988. 45. In the case of the tail-warning radar, the Air Force assumed that flight tests over water could be used to calculate the effectiveness of the system over land, an assumption that proved to be invalid; see Tony Capaccio, "Air Force Admits to New B-i Problems," Defense Week, Oct. 1, 1990, p. 1.

[290]

The

Politics of Stealth

and other defensive avionics modifications were expected to take until 1993. Even then, there was no guarantee that the electronic countermeasures suite would be able to meet its performance requirements. The best-case scenario, therefore, was that this system would be fully capable approximately seven years after the bomber itself entered service. 4(1 Since the B-iB depended on its defensive avionics system, its penetration capabilities would be adversely affected while these problems were being sorted out. This was another case where rushing a technologically adventurous system into production proved to be disastrous. The Air Force badly underestimated (or intentionally downplayed) the technical risks associated with electronic countermeasure development. It was assumed (or 1991,

alleged) that the design of the system could be frozen early in the development process and that production activities could begin at that point. Inadequate attention was given to hardware and prototype testing in the

development phase of the program. Instead, the system was pushed into production, and many defective units were built and deployed. Extensive retrofitting was ultimately required to standardize the systems in service and make them more capable. As always, retrofitting was expensive and time-consuming. The early operational capability promised at the outset of the

cy

program never materialized.

was counterproductive because production

interfered with development.

In fact, concurren-

activities

and

retrofitting

took a long time just to undo the caused by the program's accelerated production schedule. 47

damage

It

46. Molly Moore, "New Flaws Found in B-i Bomber," Washington Post, July 10, 1988; B-iB Bomber ECM Capabilities, Hearings before the House Armed Services Committee, 100th Cong., 2d sess., July 1988; John D. Morrocco, "B-iB Defensive Avionics Meet Only Half of Intended Goals," Aviation Week and Space Technology, Aug. 1, 1988, pp. 14-15; CBO, B-iB Bomber, pp. 12-15; 8-18 Defensive Avionics System, Hearings before the Senate Armed Services Committee, looth Cong., 2d sess., Oct. 1988; GAO, B-iB Cost and Performance Remain Uncertain, pp. 11-12; "USAF Will Use Stand-Alone Radar Warning System to Shore up B-iB's EW Capabilities," Aviation Week and Space Technology, Jan. 2, 1989, p. 101; Peter Grier, "What Do Weapons Really Cost?" Military Forum 6 (Nov.-Dee. 1989), 60-68; Bruce D. Nordwall, "Air Force Completes Flight Tests of B-iB's Defensive EW Systems," Aviation Week and Space Technology, March 4, 1991, p. 67. 47. HASC Report, pp. 5-6, 11-13; Conahan Testimony, pp. 11-14, 19-20. Concurrency also compounded one of the B-iB's other main problems. Instead of relying on a prime contractor to orchestrate the project, the Air Force acted as its own program manager. Given its lack of administrative resources and management experience, this would have been challenging under the best of circumstances. In this case, the program's frenetic schedule simply overwhelmed the Air Force's ability to supervise the activities of all its contractors. In particular, the Air Force did a poor job of overseeing its electronic countermeasure contractor, Eaton-AIL. See the testimony of John E. Krings, Director of the Office of Operational Test and Evaluation, Office of the Secretary of Defense, in National Defense Authorization Act for Fiscal Years 1988-7989, pp. 120-121, 142. See also HASC Report,

pp. 13-15.

[291]

,

Flying Blind

Assessing the B-iB Program

The public debate about the

Some

B-iB's final evaluation

was vigorous but

charged that the B-iB had serious operational limitations because it was overweight, underpowered, and not maneuverable enough to fly at low altitudes. Some charged that its range was limited because it could not cruise at high, fuel-efficient altitudes. Other critics complained that the bomber's avionics would need years of expensive modifications, and even then the electronic countermeasure system would probably fall far short of its original design specifications. In any event, the bomber's avionics would have to be improved to contend with sophisticate air defenses in the 1990s. It was expected that these enhancements would cost $3-8 billion. Some of the program's critics wondered if it was worthwhile to spend money on B-iB modifications and enhancements since the stealth bomber seemed to be just around the corner. Many analysts concluded that, in the final analysis, the B-iB was "a flying Edsel" and "a dismal failure." 48 In response, the Air Force correctly pointed out that the B-iB weighed exactly what it was supposed to weigh; it met its design specification. Although it was true that a fully loaded B-iB weighed 82,000 pounds more than the B-i, this was not due to a breakdown of the procurement not rigorous.

process; a decision

critics

was made

ALCM

in 1981 to

make

the necessary weight

The Air Force also argued, again correctly, that the stall inhibitor and stability enhancement systems would improve the B-iB's lift, maneuverability, and cruising capabilities substantially. Finally, the Air Force was correct when it pointed out that one had to distinguish between modifications needed to bring the B-iB up to its original performance requirements and enhancements that would provide capabilities beyond the original baseline; only true remedial efforts were legitimate charges to the B-iB account. 49 At the same time, the Air Force did not have a good response to the provisions for

carriage.

David Evans, "The B-i: A Flying Edsel for America's Defense?" Washington Post Jan. 4, 1987; Molly Moore, "Travails of the Centerpiece Weapon," Washington Post, Aug. 10, 1987. See also Molly Moore, "Air Force Management of B-i 'Screwed It Up,' Aspin Says," Washington Post, March 31, 1987; "Text of Aspin Memo to HASC Members on B-iB Problems," Aerospace Daily, Feb. 18, 1987, pp. 253-254. Critics also charged that the Air Force used bait-and-switch tactics to get the B-iB program funded in the first place. They maintained that the Air Force excluded some costly items from the program in 1981, such as a forward-looking infrared radar, only to reintroduce them as "system enhancements" after 48.

program was well established; see HASC Report, pp. 15-17. 49. Chief of Staff Gen. Larry D. Welch and Secretary of the Air Force E. C. Aldridge, Jr., "The B-i is Fulfilling Its Mission," Washington Post, Jan. 12, 1987; Gen. John T. Chain, SAC, "To the Critics of the B-iB: Bah!" Washington Post, May 2, 1987; Benjamin F. Schemmer, "Is the B-iB Bomber an Edsel or Does It Deserve a Grade of B + ?" Armed Forces Journal Internathe

tional,

Feb. 1987, pp. 60-64;

CBO, B-iB

Bomber, pp. 32-34.

The

Politics of Stealth

main charge leveled

—

B-iB that its avionics systems were not as capable as they should have been in the late 1980s. This meant that the B-iB was not as survivable as it should have been and not as effective as the Air Force itself had originally expected. The Air Force's weak reat the

sponse was that the B-iB could still perform its basic mission because it had other performance capabilities on which it could draw. As far as Air Force leaders were concerned, the B-iB was "a fantastic flying machine" and "the best warplane in the world today, on anybody's side." 50 This might have been true, but it was beside the point. There was simply no denying that a fully capable avionics suite would have made the B-iB

more effective. The B-iBs track record was

mixed one. Although

was not the unmitigated disaster its detractors made it out to be, it was not the strikingly successful enterprise the Air Force claimed it was. The program's biggest success was in coming close to its cost target. This was one of the program's most prominent benchmarks, and it was unusual for a program this complex to do reasonably well on cost grounds. At the same time, at least $1 billion had to be spent on avionics modifications to bring systems up to, or close to, their design specifications. a

These modifications were expected lion

over

its

$20.5 billion cost

to drive the

it

B-iB program $400 mil-

target. 51

The B-iB program clearly did a much poorer job of meeting its performance targets and its timetable. Although the B-iB ultimately demonstrated impressive low-altitude penetration capabilities, its operational capabilities would have been even more impressive if its avionics sys-

tems had been

on schedule. Some doubted that the B-iB's defensive avionics system, which was vitally important, would ever meet its design specifications. The program had many delays, although it did meet most of its formal deadlines: the first prototype flew in October 1984; the first squadron became operational in October 1986; and the last aircraft rolled off the assembly line in April 1988. Still, the program failed miserably in terms of providing full operational capabilities on schedule. The Air Force's estimate of when the electronic countermeasure system would be fully capable slipped repeatedly, from 1988 to 1990 to 1993, which is just one indication of how stubborn these fully capable

Quoted

in Molly Moore, "Air Force Defends B-i as World's Best Bomber," Washington Post, Feb. 24, 1987. See also Gen. Larry D. Welch, "Narrow Focus Unfair to B-iB Bomber," Chicago Tribune July 23, 1988; Richard Halloran, "General Defends B-i Bomber but Concedes Flaw," New York Times, July 13, 1988.

50.

,

51.

Grier,

Savings elsewhere

in the

"What Do Weapons

Uncertain, pp. 24-26.

program kept the

Really Cost?"

p.

61;

overrun to $400 million; see B-rB Cost and Performance Remain

total cost

GAO,

Flying Blind

defensive avionics problems were. 52 Both the offensive and defensive avionics systems required extensive retrofitting. Although compressing the B-iB's development and production enabled the Air Force to in 1981,

it

was

sell

a recipe for trouble in the

strategy did not leave

House and Congress long run. The highly concurrent engineering development and

the program to the White

much

time for

and in the avionics area especially a great deal engineering development and hardware testing was sorely needed. was not a coincidence that the B-iB's most serious problems were hardware

testing,

of It

in

these technologically ambitious areas.

The

B-2 Acquisition

Program

Redefining the State of the Art

The B-2

bomber program

an extraordinarily ambitious technological undertaking. Revolutionary advances are expected to lower the radar cross section and infrared signature of the aircraft dramatically, which should make it much more difficult to detect than the B-iB. This, in turn, should enable the B-2 to penetrate effectively even against highly advanced air defenses. 53 To assess the magnitude of the technological challenges the B-2 faces, we have to consider how the B-2 stealth

is

exploits the three basic sets of stealth technologies.

The

B-2, unlike the B-iB,

was designed from

scratch with the detec-

problem in mind. Northrop's engineers did not have to limit themselves to making minor modifications on an existing design; instead, they could go to great lengths to minimize the aircraft's detectability. As a result, the B-2's design is radically different from those of more conventional aircraft, such as the B-iB. Instead of having a conventional fuselage with traditional wings and a large vertical tail, the B-2 is a flying wing. Its wings have been completely blended into the fuselage, and a tail has been eliminated altogether. The B-2 therefore does not have highly reflective joints between the fuselage and the wings and tail; these problem areas have simply been designed away. In addition, wing surfaces have been rounded, and the cockpit has been tability

52.

David M. North, “Development Problems Delay

Full

B-iB Operational Capability,"

Aviation Week ami Space Technology, Nov. 3, 1986, pp. 34-35; "USAF Expects Fully Operational B-iB by 1988 within Spending Limits," Aviation Week ami Space Technology, Jan. 26, 1987, p. 25.

Although authoritative estimates remain

generally accepted that the B-2's radar cross section will be much lower than the B-iB's. Even so, the B -2 will not be radar-invisible. Low-frequency early warning radar systems will probably be able to detect its presence even if they cannot track it well enough to coordinate an attack against it. 53.

classified,

it

is

1

294]

The

Politics of Stealth

carefully blended into the wing.

The B-2's engines have been completely buried in the body of the aircraft, and its engine inlets have been carefully hidden; baffles shield the engine fans. The trailing edge of the wing helps to screen engine exhaust, which will probably be mixed with and cooled by ambient

air.

54

The B-2 s exotic design is a mixed blessing, however; any aircraft based on an unproven design faces a long and potentially tumultuous flight testing program. At least two of Lockheed's smaller stealth aircraft crashed during flight testing, for example, and both the B-49 and B-58 had serious stability problems when they finally took to the air. Even the B-iB, a relatively conventional aircraft that had been thoroughly tested in another incarnation, had problems in flight testing. The B-2 involves

many more aerodynamic advances

than the B-iB. All-wing

aircraft,

moreover, do not have an established aerodynamic track record at high or low altitude, and the B-2 is required to operate in both flight regimes. In short, the B-2's adventurous design involves development risks. 55 Moreover, the B-2 relies heavily on new materials. Although the aircraft's central frame is made of titanium, its large wings are made of graphite-based, radar-absorbent composite materials. It has been estimated that composites make up 80 percent of the B-2's structure. In addition, honeycomb panels are used along the leading edges of the

wings

provide structural strength and to diffuse radar waves. In addition, special radar-absorbent materials will be used around the aircraft's engine inlets, to cut down on radar reflectivity in this critical area, to

and the

aircraft will

be coated with special paints designed to minimize

and the aircraft's visibility. 56 Although exotic radar-absorbent materials might make the B-2 more difficult to detect, seemingly mundane materials problems could badly disrupt the overall program, as they did to the B-70. Composite materials are more difficult to work with than metal, and they are not well suited to high-rate production. New tools and manufacturing techreflection

54. Richard Halloran, "Stealth Bomber Takes Shape," New York Times, May 16, 1988; "USAF, Northrop Unveil B-2 Next-Generation Bomber," Aviation Week and Space Technology, Nov. 28, 1988, pp. 20-23; Michael A. Dornheim, "Initial Test Flight Reveals New Details

about Design Characteristics of B-2," Aviation Week and Space Technology, July 31, 1989. Th e B-2's altitude requirements are discussed in Michael A. Dornheim, "Air Force ^ 55 Cites 1984 B-2 Redesign as Major Reason for Schedule Lag," Aviation Week and Space Technology, Nov. 7, 1988, p. 20; Rick Atkinson, "Unraveling Stealth's Black World," Washington -

Post,

Oct.

9,

1989.

56. This discussion is based on published accounts of Northrop's incorporation of advanced materials into the B-2's design; precise details remain classified. See Atkinson, "Stealth"; Vartabedian, "Composites Take Off"; Bill Sweetman, "B-2 Comes out of the Black," International Defense Review, July 1988, pp. 787-789; James B. Schultz, "Air Force

Presses for Maiden Flight of B-2 Bomber before Year's End," Defense Electronics, Sept. 1988, pp. 87-90; Sweetman, Stealth Aircraft, pp. 71-82; Malcolm Browne, "Will the Stealth Bomber Work?";

New

I

2 95 ]

York Times, July 17, 1988.

Flying Blind

Northrop B-2 (U.

S.

Department of Defense)

niques have to be devised for even relatively simple procedures, such as drilling and fitting, and some conventional mass production methods do not work at all. Bonding these materials to the aircraft's metal parts

composites is an especially complicated process. As a result, Northrop and its team of subcontractors have been forced to invent 900 new manufacturing processes over the course of the B-2 program. New composite materials consequently involve some risks, and they n/ could cause major problems in the B-2 production program. Finally, the B-2 needs an advanced offensive avionics system because Developing flight it is to penetrate at low as well as high altitudes.

and

to other

and terrain-following radar systems for a low-altitude bomber is challenging under the best of circumstances; high-speed aircraft are escontrol

Composites also have high operating and maintenance costs. They are sensitive to temperature changes and moisture, so a stealth bomber fleet will have to be kept in hangers. Composite materials also have special maintenance requirements that are both complicated and expensive; David C. Morrison, "Sites Unseen," National Journal, June 4, 57.

Aviation Week, Dec. 1988, pp. 1468-1472; "B-2 to Be Delivered to Whiteman in Fiscal 1991," "How Stealth's Consensus Crumbled," Washington Post, Oct. 5, 1988, p. 22; Rick Atkinson, 10, 1989.

[296]

The

Politics of Stealth

Northrop B-2

(U. S.

Department

of Defense)

pecially difficult to control at

low altitudes because atmospheric pressures and turbulence are great and the margin for error is virtually nonexistent. The B-iB faced enormous development problems in this area even though four B-i prototypes had flown for many years; the addition of 82,000 pounds to the design was enough to throw development of the B-iB's flight control system off course. The B-2, on the other hand, is a fundamentally new aircraft, and flying wings have not been tested extensively at low altitudes. Developing offensive avionics for the B-2

is

therefore a tricky proposition.

Progress in this area has been complicated by the fact that the B-2 was originally designed to fly at high altitudes only. The Air Force decided in 1984 that the new bomber also needed a low-altitude penetration capability, so the offensive avionics had to be completely redesigned late in

development process. The airframe also had to be redesigned at this point to withstand the wear and tear of low-altitude operations. In addition, electrical and hydraulic systems had to be reconfigured. These design changes added $1 billion to the cost of the program and delayed the development effort by one year. The scheduled date for the B-2's first flight was consequently pushed back to December 58 1987. the

See Dornheim, “Air Force Cites 1984 B-2 Redesign"; Atkinson, “Unraveling Stealth's Black World.' It is not clear why the B-2 was originally designed for high-altitude opera58.

tions only.

Although Air Force doctrine

was generally recognized by

traditionally

emphasized high-altitude operations, low-altitude bombers were harder to conclude that the emergence of stealth

the late 1970s that detect than high-altitude aircraft. One is tempted to technology in the late 1970s allowed Air Force traditionalists to rescue high-altitude missions from the operational ashcan. The B-2 requirements formation process is, however, it

shrouded

it is impossible to be sure how these deliberations unfolded! not clear why the B-2 program was reoriented in 1984. It is possible that Soviet air defenses were improving more rapidly than expected, that stealth gains were more modest than expected, or both.

still

Similarly,

it

in secrecy, so

is

(297]

Flying Blind

The B-2 is less adventurous than the B-iB in one respect: it does not need a robust defensive avionics system because it will be hard to detect in the first place. It will consequently be able to sidestep one of the B-iB's thorniest problem areas. It would be a mistake to conclude, however, that the B-2

is

safely within the state of the art simply because

it

does not need impressive electronic countermeasure capabilities. In short, the B-2 program is one of the most ambitious weapon development projects undertaken in the United States since the end of World has to overcome truly spectacular technological challenges in several critical areas. It is said, moreover, that the manufacturing processes that have to be devised for the B-2 may be as revolutionary as the 59 For example, Northrop built most of its production tools aircraft itself. on the basis of computer models, bypassing the step of building master

War

II.

It

tools.

Embracing Concurrency

Gen. Larry Welch, the Air Force chief of staff, maintained that “it would have been a disaster" if the B-2 program had been pushed into production in the early 1980s, because of the many technological unknowns that confronted the program at that time. According to Welch, people who advocated accelerating the B-2 earlier in the decade “tre-

mendously underestimated the technical challenge in the B-2.“ Even in 1987, Welch maintained, the Air Force was far from understanding “all 60 The B-2's costs, of the costs and risks associated with the program." risks, and performance capabilities could not be accurately assessed in 1987 because construction of the first B-2 had just begun and flight testing had not yet gotten under way. In fact, the first B-2 did not roll out of the Northrop hanger until November 1988, and the first flight did not take place until July 1989. Moreover, according to the General Accounting Office, the first twelve months of flight tests did not provide “a

rigorous demonstration of the B-2's full performance capabilities" be61 cause they did not simulate “realistic operational flight profiles."

The Air

Force, however,

made

a

de

facto

the B-2 early in the development process.

production commitment to

The Air Force and Northrop

committed themselves to a specific aircraft design long before flight testing even began. Final decisions about the system's design were conWilliam B. Scott, “New Design, Production Tools Will Play Key Role in B-2 Cost," Aviation Week ami Space Technology, Dec. 5, 1988, pp. 18-21. 60. Welch quoted in Dornheim, “Air Force Cites 1984 B-2 Redesign," p. 20; Ralph Vartabedian, "Northrop Delays Initial Flight of Stealth Bomber for Four Months," Los Angeles 59.

Times, jan. 5, 1988. 61. GAO, B-2 Bomber:

Initial Flight Tests,

GAO/NSIAD-90-284,

Sept. 1990,

p. 5.

The

sequently

Politics of Stealth

made on

the basis of paper studies and computer modeling. In addition, subsystems were completely integrated into the aircraft at an early stage. As Northrop chairman Thomas Jones pointed out, one of the challenges Northrop faced was "figuring out how to integrate all of these systems up front." 62 Moreover, a full array of production tools

was

purchased during development because all of the 132 B-2S in the original procurement plan were to be built on production tooling. The first B-2S off the assembly line were therefore not true developmental prototypes but full-fledged production articles. Finally, a substantial financial investment was made in the B-2 early in the development process. The Air Force awarded a $2 billion production contract to Northrop in November 1987, twenty months before the first B-2 took to the air. A total of $22 billion

was appropriated

for the B-2 (approximately $13 billion of

which

was

spent) before flight testing of the new bomber began. The Air Force expected that B-2 appropriations would exceed $48 billion by 1993, when critical performance capabilities would finally be tested. As the General Accounting Office noted, this is "the point in testing when problems are typically discovered." 63

According

to the Air Force's original

of the B-2 fleet

would be

built

of the aircraft concluded. 64

As

procurement plan, roughly half

and delivered

to

SAC

before flight testing

a result, a great deal of

production would take place while development testing was still under way. One clear sign of the Air Force commitment to producing and deploying the B-2 regardless of the outcome of its flight testing program was the fact that construction of base facilities for the first squadron of B-2S began prior to

Congress appropriated $144.3 million for fiscal year 1988 and 1989. 65 There is no doubt that the B-2 program has featured its first flight.

this

purpose

in

a great deal of

62. Jones quoted in Bruce A. Smith, "Northrop Chief Is Confident in B-2 despite Tight Budgets," Aviation Week and Space Technology, Dec. 5, 1988, pp. 21-22. 63. GAO, Strategic Bombers: B-2 Program Status and Current Issues, GAO/NSI AD-90-1 20, Feb. 1990, p. 3. See also John D. Morrocco, "Management Problems Delay First Flight of Northrop B-2," Aviation Week and Space Technology, Jan. 11, 1988, pp. 16-17; "Technical, Manufacturing Problems Will Delay First Flight of B-2 Bomber," Aviation Week and Space

Technology, April 18, 1988, p. 23; "Employment Fluctuations Point to Changes in B-2 Program," Aviation Week and Space Technology, March 21, 1988, p. 21; "Air Force Adds ATB Tooling to Development Program," Aerospace Daily, May 4, 1987, p. 187; Vartabedian, "Northrop Delays Initial Flight"; Eileen White Read^ "Northrop Gets Stealth Award," Wall Street Journal, Jan. 25, 1988, p. 7; "Northrop Awarded $2 Billion for B-2 Bomber Production," Aviation Week and Space Technology, Feb. 1, 1988; Smith, "Northrop Chief Is Confi-

dent." 64. See Barbara Amouyal, "Air Force Will Buy before B-2 Bomber Is Ready to Fly," Defense News, May 8, 1989, pp. 4, 34; John D. Morrocco, "Bipartisan Opposition to B-2 Grows despite Release of New Information," Aviation Week and Space Technology, July 17, 1989, pp. 22-23.

65.

"Congress Backs Funding

11, 1988, p.

16.

[299]

for First B-2 Base," Aviation

Week and Space Technology, Jan.

Flying Blind

was already in first flight. Employing this much concurrency in a technologically exotic program was obviously risky. Even supporters of the B-2 admitted that there was a lot of concurrency in the program and that it consequently involved more concurrency. Northrop's Jones maintained that the B-2 production when it made its rollout, months before its

risk

than the typical program. 66

Emerging Problems

The B-2 program experienced a variety of problems as its development and production activities unfolded. The original design of the centersection structure created excessive loads in some areas; it was redesigned when the aircraft was reconfigured for low-altitude operations in 1984. The fuselage, engine inlets, and engines were not as aerodynamically compatible as they needed to be; the engine inlets had to be redesigned three times as a result. The design team working on the B-2's engines failed to notify the structural engineering team of changes that affected the layout of the airframe. Northrop had trouble working with the aircraft's composite materials; relatively simple matters, such as drilling and fitting, proved to be troublesome. At the time of the rollout, the radar-absorbent materials used on the leading edges of the wings still presented significant production problems. These material problems were expected to become even more serious when the program moved beyond simple prototype construction because building the aircraft involved "cut-and-fit" techniques that were incompatible with high-rate production. Published reports suggest that offensive avionics development for the B-2 fell at least two years behind schedule and that the offensive avionics suite might not be fully developed until the mid-1990s, after the B-2 was supposed to be in service. Provisions were made in the B-2's two-man cockpit for a third crew member, in case avionics automation failed to proceed as hoped. There were also problems with the aircraft's electrical system and ejection seats. Finally, the

General Accounting Office reported that there were 30,000 more defects in the first two B-2S than the Air Force had expected and that changes in the B-2's design were taking place at a rate of 2,000 per month in early 67 1989, when production efforts were supposedly well under way. Confident," pp. 21-22; AuCoin in "Despite Reports of Trouble, ATB Boosted," Defense Week March 14, 1988, p. 14. 67. Again, this discussion is based on published reports about the B-2's progress; details remain classified in many respects. See John D. Morrocco, "Air Force Unveils Design of Northrop B-2 Bomber," Aviation Week and Space Technology, April 25, 1988, pp. 16-17; Richard Halloran, "U.S. Delays Debut of Stealth Craft," New York Times, Aug. 5, 1988; Schultz, "Air Force Presses"; Morrocco, "Management Problems," p. 16; Dornheim, "Air Force Cites 1984 B-2 Redesign"; John D. Morrocco, "Budget Constraints Force One-Year Delay in 66. Jones in Smith,

"Northrop Chief

Is

,

(300J

The

Politics of Stealth

Whereas the B-iB's schedule did not begin to slip until its flight testing program was well under way, the B-2's slipped dramatically before flight testing

even began.

In fact, the first flight of the B-2

was delayed from

late

1987 to April 1988, to August 1988, to late 1988, and finally to July 1989, a delay of nineteen months. The program's technical problems

were surprisingly stubborn. And because the B-2 was on a tight schedule, as concurrent programs always are, every delay in the development phase disrupted the elaborate production plans that were in the process of being implemented. This, in turn, drove up the program's costs dramatically; large numbers of production workers had to be paid while they were idle. 68 Although 132 B-2S were supposed to cost $36.6 billion, or million $277

each

(in

1981 dollars), the program's cost

problems and The cost of the B-2's troubled avionics suite, for between 1988 and 1989. By the time the B-2 began flight

delays took their

example, tripled

grew

as technical

toll.

testing in July 1989, the

Pentagon estimated that the program would cost $43.8 billion (in 1981 dollars), an increase of about 20 percent. The B-iB, on the other hand, had a cost overrun of around $400 million, about 2 percent of

program costs. Factoring inflation into the equation, B-2 program costs were expected to reach $70.2 billion. By early 1990, estimated program costs had grown to $75.4 billion, or $371 million per total

aircraft. 69

Though alarming, this was the Pentagon's best-case estimate. Every technical snag would drive costs up more, as would congressional decisions to fund the program at lower rates than Air Force plans called It

seemed almost

for.

however, that Congress would cut budget given that the Air Force planned to ask for around

inevitable,

requests for the B-2,

Peak Production of B-2," Aviation Weekend Space Technology, Dec. 19, 1988, pp. 24-25; Molly Moore, "New Electronics Bug the Pentagon," International Herald Tribune, November 8, 1989; "USAF, Northrop Unveil B-2," p. 22; GAO, Strategic Bombers, 21-25; Bruce a! pp. Smith, "B-2 Production Delays Drive up Program Costs," Aviation Week and Space Technology, July 24, 1989, pp. 26-27; Molly Moore, "Air Force Report Sharply Criticizes Stealth Bomber Maker," International Herald Tribune, Oct. 11, 1990. 68. David Fulgham, "Lower Classification Seen for Black B-2 Bomber," Air Force Times,

Bomber to Fly in August?" International Defense Review, Feb. 1988, pp. 107-108; Vartabedian, "Northrop Delays Initial Flight"; "B-2 First Flight Set for Mid-December," Aviation Weekend Space Technology, Aug. 8, 1988, p. 16; "USAF, Northrop

Jan. 18, 1988, p. 12; "Stealth

or

F

Unveil B-2." 69. Barbara Amouyal, "B-2 Bomber Poses a Quandry for Budget-Conscious Lawmakers," Defense News March 5, 1990, p. 11; John D. Morrocco, "Cheney Proposes ,

Stretchouts, Cuts in B-2, C-17 Programs," Aviation Weekend Space Technology, April 30, 1990, pp. 1 8—20. The B-2's cost problems led Congress to consider bringing Boeing in as a second-source producer in 1987. It ultimately decided that it would be too expensive to set

up

second production line, given the relatively small number of aircraft to be built. The growth led the under secretary of defense for acquisition, Robert Costello, to try cancel the program in 1988. a

B-2's cost

to

Flying Blind

$8 billion per year for 1992-94. Key legislators pointed out that Congress was unlikely to fund the B-2 at this rate; the defense budget as a whole

and many important programs were already underfunded. It was estimated that if annual expenditures on the the amount spent on the Strategic DeB-2 were reduced to $4 billion B-2 unit fense Initiative, the largest program in the defense budget costs would climb to $725 million. And if annual expenditures were limited to $3 billion, the cost of each B-2 would exceed $1 billion. As Les Aspin, the chairman of the House Armed Services Committee, explained to the Air Force, the B-2 program was caught on the horns of a dilemma. According to Aspin, "If you make the annual costs acceptable, you make the sticker price unacceptable. And if you make the sticker 0 price acceptable, you make the annual cost unacceptable." With this cost problem in mind. Secretary of Defense Richard Cheney

was shrinking

in real terms,

—

—

decided in April 1990 to build only 75 B-2S. This brought the Defense Department's official cost estimate for the program down to $61.1 billion. By early 1991 this estimate had grown to $64.7 billion, or $863 million per aircraft. 71 By the time policy makers in Washington came to apprreciate the magnitude of the B-2's technical and financial problems, sunk costs in the program had to be measured in tens of billions of dollars. This investment created enormous pressure to continue to spend money on the program, regardless of

Cheney noted

its

strategic value

and

cost-effectiveness. 72

As

over the B-2, one of the reasons the Bush administration wanted to go ahead with the program was to "gain the payoff on the enormous investment" that had already

been made

at

one stage

in his deliberations

73

Congress also found it difficult to turn its back on this investment; attempts to cancel the program outright were defeated in 1989 and 1990. With each step forward, though, the momentum in the B-2 S concurrent development and production program became harder in

it.

to resist.

Aspin quoted in Morrocco, “Bipartisan Opposition," p. 23. See also David F. Bond, “Latest B-2 Data Spur Mounting Hill Opposition," Aviation Week and Space Technology, July 70.

3,

1989, p. 20.

Morrocco, "Cheney Proposes Stretchouts," p. 19; David F. Bond, “Budget Squeeze, Reduced Threat Cut U.S. Strategic Program Upgrades," Aviation Week and Space Technology, 71.

March

48-49. analysis of the strategic value and cost-effectiveness of the B-2 is beyond the scope of this book. The case against the B-2 is developed in Michael E. Brown, “The Manned Bomber and Strategic Deterrence in the 1990s," International Security 14 (Fall 1989), 72.

18, 1991, pp.

An

3-46. See also the rebuttal by Secretary of the Air Force Donald Rice, “The Manned Bomber and Strategic Deterrence: The U.S. Air Force Perspective," International Security 15 (Summer 1990), 100-128; and Brown's response, "The Case against the B-2," International

(Summer Cheny quoted

Security 15 73.

1990), 129-153. in

Morrocco, “Bipartisan Opposition,"

p. 22.

[302]

The

Politics of Stealth

Conclusions

The origins of the B-iB and B-2 programs were idiosyncratic. The B-iB grew out of the program President Carter canceled in 1977, and as such it was driven by an unusual array of strategic, bureaucratic, economic, and even technological forces. The Air Force argued that the strategic case for a new bomber became even more compelling in the late 1970s when more became known about Soviet progress in developing lookdown/shoot-down air defense systems. The Air Force's bureaucratic interest in deploying a new bomber became even more intense in the aftermath of Carter

1977 B-i decision. Rockwell had a vested interest in the B-i program that contractors rarely have at such an early stage of the acquisition process, and it lobbied aggressively and effectively on its s

behalf. Finally,

developments in stealth technology made it easier for advocates of the program to argue that the B-iB was very different from the bomber Carter had canceled earlier on. It is therefore difficult to come to any simple judgment about the forces responsible for triggering the B-iB program. Several factors seem to have been significant in this case.

The B-2 program is difficult to assess because the origins of the program are still shrouded in secrecy. That said, it seems clear that three factors played

an important role in the early stages of the program. First, the Air Force was spurred into action by the operational experiences of the wars in Vietnam and the Middle East, and by its forecasts about Soviet air defense capabilities. These strategic considerations lead it to begin developing stealth aircraft in the mid-1970s and a stealth bomber toward the end of the decade. Second, the Air Force was especially concerned about preserving the manned bomber in general and the penetrating

bomber

in particular after

Carter canceled the B-i.

The Air Force had a strong bureaucratic interest in coming up with a bomber proposal that would appeal to Carter. Finally, advances in stealth technology

in the late 1970s indicated that

it

might be possible

to build a

large, stealthy aircraft. Therefore, the origins of the B-2

have been influenced by

a

combination of

strategic,

program seem to bureaucratic, and

technological factors.

The Air Force followed a familiar course in setting up the B-iB and B-2 acquisition programs. It was allowed to go as quickly as it wanted with the B-iB and it was inclined to proceed very quickly indeed. In fact,

—

the Reagan administration encouraged the Air Force to deploy the new bomber as soon as possible. The result was one of the most concurrent

weapon

acquisition

programs

in history;

attributes of concurrency outlined in

it

exhibited

Chapter

program was, however, deceptively ambitious [303]

1

all

(see Table

eight of the 2).

The B-iB

technologically. In partic-

Flying Blind

defensive avionics system involved major advances. Because one of its major subsystems required major improvement, the B-iB belongs in the fifth position on the nine-part scale of technological amular, its

bitiousness outlined in Chapter 1 (see Table 1). It is not surprising that the program had its most serious cost, schedule, and performance problems where technological ambitiousness intersected with unbridled concurrency.

one of the most technologically adventurous weapon development programs of the past several decades. It is trying to consolidate truly revolutionary advances in aircraft design and materials, and it also involves major advances in offensive avionics. Because a new and radically different system design was needed to meet the program's requirements, the B-2 belongs in the first position on the nine-part scale of technological ambitiousness outlined in Chapter 1. At the same time, the B-2 program is very concurrent. Competition in the program was cut off before full-scale development began. Northrop's bomber design was frozen fairly early in the development process. SubIt is

no

secret that the B-2

is

systems were integrated into the design early in the greater scheme of things. Production aircraft were used for all flight testing, so production tooling had to be purchased in the midst of the development process. Northrop consequently had to rely on paper studies and computer modeling for many design and production decisions. In addition, several billion dollars

was invested

in the

program long before

flight testing

began. The Air Force even began to build bases for the B-2 before any aircraft had taken to the air, which was an indication of how much impact the B-2 flight testing program was expected to have on its production decision. In short, the Air Force made its production decision early in the development process, and the B-2 program exhibited seven of the eight attributes of concurrency outlined in Chapter

1.

Both programs, therefore, were risky from a procurement standpoint. The B-iB program involved too many technological unknowns to be pushed as quickly as it was. The same can be said of the B-2 program, which is based on a different but equally inappropriate combination

—

—

of technological ambitiousness

and concurrency-

The Origins and Outcomes of Weapon Acquisition Programs

have examined the origins of fifteen U.S. strategic bomber programs, focusing on the roles played by technological, economic, bureaucratic, I

and strategic factors in decisions to initiate new development efforts. I found no significant support for the widely held view that technological and economic factors are responsible for triggering most weapon development programs. Instead, I found some support for bureaucratic arguments and a great deal of support for strategic explanations of these

weapons

decisions.

have also examined the development objectives the Air Force pursued and the procurement strategies it employed in these programs One of my main arguments is that the success of a program in terms of cost, schedule, and performance has been largely a function of the development objectives and procurement strategy on which it was based. Technologically ambitious programs worked out well when guided by sequential procurement strategies and poorly when guided by concurrent strategies, for example. argue, moreover, that a powerful set of I

1

.

I

and bureaucratic forces pushed these programs toward extremely ambitious development objectives and highly concurrent procurement strategies at the same time. As a result, many U.S. bomber programs ultimately experienced serious problems. strategic

occasionally use "the Air Force" as shorthand for "the Air Force" in this chapter. i

.

I

AAC,

the

AAF, and

the U.S.

Flying Blind

Program Origins Technological Factors

most prominent explanations of the origins of weapon development programs emphasizes the role played by technological factors in the early stages of the acquisition process. According to this line of argument, the original impetus for a program is the emergence of a technological discovery from the laboratory. The military, it is said, reacts to developments such as these by devising strategic rationales and

One

of the

operational requirements to justify exploiting them. We saw no clear sign of such motivation in the U.S. strategic

bomber

program. The B-35 and B-36 programs were not triggered by technological developments. Given the state of the art in 1941, airplane manufacturers were hard pressed to build bombers that could fly 5,000 miles, let alone 10,000 miles.

The

B-45, B-46, B-47,

and B-48 programs

were bedeviled by technological challenges, not blessed with technological opportunities. The AAC started several jet engine projects in 1941 precisely because the United States was making little progress in this area. In 1943, the AAF pushed its leading jet engine manufacturer to build an engine more than twice as powerful as those currently under development. Large engines such as these would be needed if jet bombers were to be built. Also in 1943, the AAF began to look into specific jet bomber designs, even though high-speed aircraft had to contend with still-mysterious compressibility effects. The AAF's 1944 requirements for what ultimately became the B-52 were set far beyond the state of the art. In what was fast becoming a familiar pattern, they called for capabilities that were simply unattainable given the technologies of the day. The AAF began to work on supersonic bombers in a general way in 1944 and on specific design studies in 1946, even though the sound barrier had not yet been broken and many believed it could not be breached. If the B-58 program had been triggered by technological developments, it would not have been launched in 1946, but in 1947, when the sound barrier was finally broken. Alternatively, it might have been launched as late as 1953 or 1954, when the transonic and supersonic area rules came to the Air Force's attention. Similarly, the B-70 program was up and running in 1954, even though the discovery of the compression lift principle that

made high-Mach,

long-range

flight possible

did not

come

to

the Air Force's attention until late 1956. Although important advances in variable-sweep wings were made in the late 1950s, these developments

did not lead the Air Force to initiate B-i requirements studies in 1961. In fact, this first round of studies did not even consider variable-sweep designs. The variable-sweep wing was not folded into the program until [306]

The Origins and Outcomes of Weapon Acquisition Programs

1963,

when

the Air Force took steps to ensure that

its new bomber would measure up in terms of prelaunch survivability and high-altitude capabilities. The variable-sweep wing was belatedly incorporated into the B-i program for strategic and bureaucratic reasons; it was not responsible for initiating the program in the first place. 2 The Air Force was not guided by technological opportunism when it

initiated

these programs.

performance requirements and system specifications were not designed to exploit recent technological gains. In

many

cases, the opposite

Its

was

true: the Air Force set its

performance requirements far beyond the state of the art; they could not be met unless unforeseen technological developments took place. The sequence of events is important here; the Air Force's requirements were

set well before the technological breakthroughs in question took place. The fact that these breakthroughs were totally unforeseen by Air Force requirements planners is also significant.

Technological explanations would also predict that the Air Force's research and development commands would play an important role in

the early stages of

mands would

weapon

acquisition programs. After all, these comhear of exciting technological breakthroughs before the

other branches of the service. Presumably, they would take the lead in alerting the Air Force hierarchy to opportunities waiting to be exploited.

One would

expect, therefore, that the technical

commands would

generally initiate requirements discussions and dominate the proceedings, making sure that the right strategic rationales were devised for their new projects. Under these circumstances, one would expect technical considerations to take precedence over operational ones, because strategic and operational arguments would only be after-the-fact rationales for

ongoing research and development activities. Again, we see no sign of this in the U.S. strategic bomber program. In the B-35 and B-36 programs, for example, preliminary requirements were issued before technical assessments of the situation had been made. Later, the Materiel Division's reservations about raising these requirements and building a bomber capable of flying 10,000 miles were simply overridden by operational considerations. In the fall of 1941, the Materiel Division was given only a few weeks to assess the proposals submitted

Command

in the

(as

it

long-range bomber competition. Similarly, the Materiel was then known) was given only a few weeks in the fall

of 1943 to assess the prospects for jet

bombers before it was forced to press on with preliminary design efforts. Warnings from the Air ReTechnological factors seem to have played a supporting role in the early phases of the nuclear-powered bomber, B-iB, and B-2 programs. Since other factors were influential in these cases as well, it would be inaccurate to say that technological developments were responsible for triggering these development efforts. 2.

Flying Blind

search and Development Command and Wright Air Development Center about the risks inherent in the B-70's demanding requirements were ignored by the Air Force hierarchy. The Air Force issued demanding

performance requirements in other cases as well. Clearly, these requirements were not driven by the current state of the art, which is what one would expect if requirements were simply rationales for exploiting new technologies.

commands were not responsible for development efforts. In many cases, little time was allo-

In short, the Air Force's technical initiating these

cated for feasibility assessments, which suggests that these kinds of assessments were not particularly influential in requirements deliberations. When the technical commands did have reservations about a project's feasibility, which they rarely did, their objections were frequently

overridden.

As

a result, the technical

commands were

not successful in

dissuading the Air Force from setting demanding performance requirements. This is not to say that technological factors were unimportant at every stage of the acquisition process. Technological developments made after production was already under way frequently led to improved compo-

nents (such as new navigation systems and engines), which in turn led the Air Force to build new aircraft models. In the case of the B-52, for example, the appearance of highly efficient turbofan engines led the Air Force to replace the B-52G with the B-52H. This is not the same thing, of course, as initiating the program in the

first

place.

Economic Factors

Another prominent argument about the origins of weapon development programs holds that the parochial economic interests of defense contractors are behind these efforts. Since the key to corporate prof-

many

securing a steady stream of contracts, it is said that contractors are compelled to come up with ideas for new weapons, ideas that are then transmitted to the military. The military is

itability for

contractors

is

said to react to these ideas, subsequently devising military rationales for

pursuing them. Again, we see little evidence of such a process in the U.S. strategic bomber program. As a general rule, the Air Force took the initiative, established development programs, and then contacted defense contractors about participating in them. The Air Force contacted the contractors, not the other way around. This was clearly the case in the B-35 and B-36 programs; the AAC established preliminary performance requirements contacted manufacturers about participating in the new venture. In the jet bomber program, the AAC took the lead in pushing jet engine

before

it

The Origins and Outcomes of Weapon Acquisition Programs

development

in 1941,

and

ers about designing a jet

in 1943 the

bomber.

AAF contacted aircraft manufactur-

In the B-52

program, the AAF's perfor-

mance requirements were so demanding that no contractor could be persuaded to work on the project, even though the new long-range bomber was bound to be one of the AAF's biggest programs. Clearly, defense contractors did not play a leading role in getting the B-52 program off the ground. The Air Force also took the lead in the B-58, B-70, and B-i programs. Several contractors approached about the B-70

project declined to participate. In the case of the B-i, the Air Force folded contractors into the in-house study effort that was already under way. Detense contractors played an active role in starting only two programs, the B-60 and B-iB, but neither case conforms to the pattern

described by economic determinists. 3 The B-60 was not based on new technologies. It was just a modified version of the B-36 and Convair's last-ditch attempt to keep its production line open. The only reason the Air Force built two B-6os was to keep competitive pressure on Boeing, which was building the bomber the Air Force really wanted. The case of the B-iB is also idiosyncratic. Because the B-iB was based on a

—

system

that

was moving

into production

when

it

was canceled,

the contractor,

Rockwell, had a strong vested interest in reviving the program. In any event, strategic, bureaucratic, and technological factors were also in-

volved

much

B-iB decision, so one should be careful not emphasis on the role played by economic factors in in the

to place too this case.

Another prominent argument in the literature maintains that, even if economic factors are not responsible for triggering programs, they influence military decisions about full-scale development and production contracts. According to political scientist James Kurth, a "follow-on imperative" leads the military to award contracts to whichever contractor needs the business the most; as one contract expires, a new one follows. 4

found no evidence to support Kurth's follow-on argument in these programs and some clear cut evidence against it. In every case studied, the Air Force based source selection on aircraft performance. In three cases, it awarded contracts to the manufacturers with the most promising bombers even though other contractors needed the business more. Boeing beat Martin in the competition between the B-47 and B-48 ev en though Boeing was deeply involved in the B-52 program and MarI

—

3. The origins of the B-49 program are murky. It is not clear whether the AAF approached Northrop about putting jet engines on the B-35 or Northrop approached the AAF. Both explanations are plausible, given that the AAF was looking for a high-speed bomber in 1944 and Northrop was anxious to keep the flying wing program alive. 4. James R. Kurth, "A Widening Gyre: The Logic of American Weapons Procurement,"

(Summer 1971), 385-397; "Why We Buy (Summer 1973), 35-36.

Public Policy 19

no. 11

the

Weapons We Do,"

Foreign Policy,

Flying Blind

The Air Force preferred Boeing's B-52 to Northrop's B-49 in the late 1940s even though Boeing had already won a production contract for the B-47 and the flying wing was one of Northrop's main sources of income. Boeing's B-52 won the fly-off against Convair's B-60 even though Boeing's huge B-47 production run was just gearing up and Convair's B-36 production program was winding down. The Air Force put more emphasis on aircraft performance than economic considerations. In the B-58, B-70, and B-i cases, the contractor most in need of a new contract also came up with the best performance estimates. This is not tin

was on

the brink of financial disaster.

surprising, since contractors recognized the Air Force's priorities. In

no way to determine whether operational or economic considerations were more influential in Air Force thinking; we have an overdetermined outcome. To be sure, economic factors were often important in the weapon acquisition process. Defense contractors were frequently active in suggesting component and model improvements after production lines were up and running, but this should not be confused with initiating programs in the first place. these cases, there

is

Bureaucratic Factors

Analysts of bureaucratic politics emphasize that the military services have strong vested interests in expanding organizational autonomy, preserving military missions, extending mission capabilities, and increasing departmental budgets. These parochial interests lead them to place a high priority on having a wide array of projects in the acquisition pipeline.

phases of the B-35, B-36, B-45, B-46, B-47, B-48, B-52, B-58, B-70, and B-i programs. 5 It does not follow, however, that these programs were driven by bureaucratic considerations, because the Air Force would also be expected to play the leading role in programs driven by strategic considerations. These programs have to be looked at individually to determine which factors were especially important in each case. The AAC had a deep and abiding interest in building long-range bombers. It could not fly long-range bombardment missions without

The Air Force

clearly played the leading role in the early

Other programs were more complicated. As noted earlier, the origins of the B-49 are murky, and the B-60 is an idiosyncratic case. A combination of strategic, bureaucratic, and technological factors seems to have been influential in the early phases of the nuclearpowered bomber and B-2 cases, and a full array of strategic, bureaucratic, economic, and 5.

technological factors to

was

at

work

in the case of the B-iB.

any simple conclusions about the origins

of the

therefore impossible to B-iB, and B-2 programs.

It is

ANP,

come

The Origins ami Outcomes of Weapon Acquisition Programs

them, and strategic bombardment was the key to the AAC's case for bureaucratic autonomy from the Army. Nevertheless, bureaucratic arguments do not provide sufficient explanations for the intercontinental B-35 and B-36 programs in 1941. In these cases, the catalyst was the changing strategic situation. As the Battle of Britain raged on, the AAC took steps to reduce its dependence on British forward bases, which it might not be able to count on if the United States entered the war against Germany. It consequently began to build bombers capable of attacking Germany from bases in the United States. So, although underlying bureaucratic forces

were

at

work

in these cases, strategic influ-

ences were more decisive.

American airmen

had a deep and abiding interest in building high-speed bombers, and not just because they wanted to maximize their chances in combat. They also wanted to maximize their chances of breaking away from the Army, and to do this they had to prove that strategic bombers were survivable. Even so, bureaucratic arguments do not provide sufficient explanations for the AAF's decisions to develop the high-speed B-45, B-46, 6-47, B-48, B-52, and B-58 bombers in the mid-i94os. Again, the changing strategic situation was the catalyst. Inalso

telligence reports about

German

jet

fighters triggered

all

these pro-

grams. Again, although underlying bureaucratic influences were present in these cases, strategic influences should be credited with sparking these weapon development programs. Bureaucratic and strategic forces also teamed

up

B-70 and B-i programs; bureaucratic influences were very important in the case of the B-70 and dominant in the case of the B-i. Starting with the 6-70, it was certainly true that the Air Force's underlying concern about the longin the

term effectiveness of Soviet air defenses led it to emphasize supersonic capabilities, which the nuclear-powered bomber might lack. In addition,

was becoming increasingly risky to rely on high-speed, medium-range bombers such as the B-47 and B-58 because of the growing vulnerability of American overseas bases, on which medium-range bombers were fairly dependent. The limitations of the B-47, B-58, and nuclear-powered bomber therefore led the Air Force to begin developing the B-70 in 1954. But several bureaucratic considerations pushed the Air Force in the same direction at the same time. The nuclear-powered bomber program was it

faltering,

and the Air Force needed

a fallback option to

throw into the

increasingly intense competition with naval and ballistic missile pro-

grams over missions and budgets. It was not a coincidence that the Air Force initiated what eventually became the B-70 program less than three weeks after the civilian leadership in the Pentagon established a dangerously autonomous organization for ballistic missile development in the Air Force.

Flying Blind

As

for the B-i,

building a

it

is

bomber with

true that the Air Force toyed with the idea of

low-altitude capabilities throughout the 1950s.

It

did not, however, get around to conducting preliminary design studies of such a bomber until late 1961, after the B-70 program had been effectively canceled by the Kennedy administration. Although the Air Force

continued to maintain that high-altitude operations were viable and that the B-70 should be built, it began to hedge its bets when it became increasingly clear that the high-altitude B-70

was no longer

viable politi-

cally.

Strategic Factors

This book supports the broad contention that strategic developments play an important role in sparking weapon development programs. We

have seen clear signs of this in the B-35, B-36, B-45, B-46, B-47, B-48, B-52, B-58, and B-70 programs and some indications in the nuclearpowered bomber, B-iB, and B-2 programs, where the situation was

more complicated. The Air Force's actions were responses to clear and present

not,

however, always straightforward

threats to national security. In

some

vague and distant threats to the future of bomber operations. In these cases, intelligence reports were usually too vague to provide dependable guidance about performance requirements. This is not surprising, given that it took ten years to build and deploy a new bomber, which might be in service for another twenty years or more. It was simply impossible to predict what kinds of operational capabilities bombers would need twenty or thirty years in the future. The ambiguity and uncertainty that surrounded these programs therefore permitted doctrinal predispositions to play an important role cases, the Air Force anticipated

requirements decisions. For example, the intelligence projections that guided the B-58 program in the early 1950s called for a bomber with low-altitude capabilities; it was expected that high-altitude operations would become problematic by the end of the decade. Although these projections were fairly explicit, they inevitably involved guesswork. The Air Force consequently discounted them, which enabled it to continue to build bombers more in

suitable for traditional, high-altitude missions.

The same thing happened in the B-70 program. Therefore, developments in the strategic arena triggered many development efforts, but organizational predispositions were important in shaping the requirements decisions that followed.

Given the uncertainties gic

bomber programs,

that

surrounded the early stages of

the Air Force

was

its

strate-

flying blind in several respects.

The Origins and Outcomes of Weapon Acquisition Programs In the first place, the exact nature of the operational threat was not always clear. In the second place, the technological possibilities for

bomber development were

often vague, since the Air Force generally did not thoroughly assess the technological landscape before initiating

development efforts. In many cases pessimistic technological assessments (either from the technical commands or the contractors themselves) were disregarded, and in every case the Air Force compounded the unknowns its programs faced by setting its performance requirements beyond the state of the art. its

The Origins of Traditional Programs

Although we have cases,

I

believe

sion areas tiated

it is

be careful about generalizing from this set of safe to say that major programs in established misto

— what we might

by the military

call traditional

programs

— are generally

ini-

services. Military organizations are highly sensitive

to threats, strategic as well as bureaucratic, to their core missions,

these external

jolts

are usually responsible for

efforts in established

and triggering development

mission areas.

Naturally, strategic trends are monitored closely by military intelligence agencies, and these agencies are especially sensitive to strategic

or operational developments that might have an impact on the effectiveness of existing forces. When ominous developments appear, operational doctrines provide a

This

is

ready guide to performance requirements.

when intelligence forecasts are too vague what kinds of new capabilities are needed, which

especially important

indicate precisely

to is

frequently the case. Military organizations are therefore quick to respond to strategic threats, however vague, to existing forces and tradi-

programs. One would expect strategic catalysts to be especially important when the operational environment is changing rapidly during wars and intense arms races. It is not surprising, therefore, that tional

—

developments were especially important in many U.S. bomber programs in the 1940s and 1950s. Military organizations also have a vested interest in perpetuating esstrategic

they are sensitive to bureaucratic threats to important missions, programs, and budgets. One would expect bureaucratic catalysts for new programs to be especially important tablished

activities.

Accordingly,

when

mission assignments defense budgets are tight. It ic

among

the services are in flux

and when

not surprising, therefore, that bureaucratcatalysts were especially important in the U.S. strategic bomber pro-

gram

is

in the latter half of the

When new programs

1950s and early 1960s.

are begun, military requirements are set in the

early stages of the acquisition process,

and they are frequently

set far

Flying Blind

beyond the

Technology is rarely the driving force in established mission areas, because military demand usually outpaces technological supply by a wide margin. At the same time, defense contractors rarely play the leading role in the early stages of major programs, because military requirements usually demand much more than contractors are capable of delivering. It does not follow, however, that military organizations dominate the entire acquisition process. Technological developments and industrial initiatives seem to play a greater role in suggesting component and model improvements once producstate of the art.

tion lines are

up and running.

Although traditional programs have not received a great deal of attention from students of weapon acquisition issues, work in this area shows that this pattern is not limited to the U.S. strategic bomber program. For example, the F-102 program was started in the late 1940s because existing American fighters would be ineffective against high-altitude jet bombers, which the Soviet Union was expected to deploy in the mid-1950s. Fighters such as the F-86D could fly only subsonically, and they were limited to altitudes under 50,000 feet. To deal with the emerging Soviet threat, the F-102 had to be able to intercept a bomber capable of flying Mach 1.3 at 60,000 feet. In short, the F-102 was triggered by strategic developments, and its requirements called for a dramatic improvement in aircraft capabilities. 6 In the late 1950s, the head of the Air Force's Tactical Air Command (TAC) was anxious to replace existing F-105 fighters for a combination of strategic and bureaucratic reasons. First, the F-105 needed long (2-mile) concrete runways for taking off and landing, and it was dependent on overseas bases for most operations. Large overseas bases were becoming increasingly vulnerable, however, due to improving Soviet missile capabilities. In addition, TAC wanted to expand its nuclear role because doing so would enable it to stake a claim to larger budgets. Defense budgets were tight at the time, and nuclear-capable systems were more likely to be funded than their conventional counterparts. TAC thus formulated a set of requirements for a fighter-bomber capable of flying across the Atlantic without refueling, operating from rough fields, carrying nuclear weapons internally, and engaging enemy fighters at high altitudes and high speeds. It was fortuitous that advances in variablesweep wing design enabled the Air Force to reconcile these aerodynamically contradictory requirements and move forward with what eventually became the F-111. At the same time, the U.S. Navy saw a pressing need to improve fleet air defense against Soviet attack. The 6. T. A. Marschak, The Role of Project Histories Paper, P-2850, Jan. 1964, pp. 66-81.

in the

Study of

R&D, Rand Corporation

[314]

The Origins and Outcomes of Weapon Acquisition Programs

Navy was concerned about

bombers armed with long-range

Soviet Tu-22

With

the growing vulnerability of

mind,

its fleet to

new

air-to-surface missiles.

formulated requirements for an aircraft capable of cruising for long periods while armed with long-range air-to-air missiles. this in

The Navy

s

goal

it

was

to build a fighter capable of intercepting Soviet

they could launch their missiles against American ships. As it turned out, Secretary of Defense Robert McNamara forced the Navy to consider a modified version of the F-m for this mission. 7 Similar kinds of decisions were made in several U.S. Army programs.

aircraft before

As one

analyst has observed, “in the 1920s and 1930s, the requirements

process in tank development tended to be unidirectional: doctrine was enunciated as a theoretical exercise; specifications were drawn from doc-

technology was asked to respond." 8 Although information on Soviet cases is sketchy, similar patterns of behavior seem to operate there as well. One study of Soviet decision trine;

making in traditional weapon acquisition programs observed that, although requirements for new weapons can originate in many places, the most usual is in the using service, probably in the operations directorate of the service staff." 9

A

study of Soviet tank programs concluded: "Much of the pressure for innovation comes from two sources: the first is the need to overcome defects in tank design as these are made clear in combat and in exercises; and the second is the need to match foreign developments, whether these are encountered in battle or merely observed in the foreign press. In this way new requirements are generated, for example, for greater firepower or better armor." 10 Similarly, an analysis of Soviet missile systems argues that "the historical record of Soviet technology allows us largely to dispose of technological determinism in accounting for the development of Soviet missile accuracy. Strategic goals have shaped the technology, rather than vice versa." 11 Finally, according to a study of Soviet fighter procurement, "there is a standard J. Art, The TFX Decision: McNamara and the Military (Boston: Little, Robert F. Coulam, Illusions of Choice: The F-111 and the Problem of 15-27; 1968), pp. Acquisition Reform (Princeton: Princeton University Press, 1977), pp. 36-56; and Vincent Davis, The Politics of Innovation: Patterns in Navy Cases (Denver: University of

7.

Based on Robert

Brown, Weapon

Denver, Graduate School of International Studies,

1967), pp. 4-16. In addition, see the in D. Douglas Dalgleish and Larry

discussion of the origins of the Trident program Schweikart, Trident (Carbondale: Southern Illinois University Press, 1984), pp. 15, 41-44. 8. Arthur J. Alexander, Armor Development in the Soviet Union and the United States, Rand Corporation Report, R-1860-NA, Sept. 1976, p. 55. 9. Arthur J. Alexander, Decision-Making in Soviet Weapons Procurement, Adelphi Papers

147-148 (London: International Institute for Strategic Studies, 1978), p. 31. 10. David Holloway, "Innovation in the Defense Sector," in Ronald Amann and Julian Cooper, eds.. Industrial Innovation in the Soviet Union (New Haven: Yale University Press, 1982), p. 386. 11. Donald

MacKenzie, "The Soviet Union and Strategic Missile Guidance,"

tional Security 13 (Fall 1988), 7.

[315I

Interna-

Flying Blind

sequence of steps posal for a

new

for the

development process of Soviet

aircraft.

A

pro-

within the aviation ministry.

aircraft usually originates

The proposal is submitted to the Council of Ministers for approval. The aviation institutes study the proposal and give broad technical direction. The proposal then goes before a scientific-technical commission composed of members of the customer and production ministries. This commission develops detailed specifications and technical parameters .

for the

new

.

.

design." 12

Although programs

within established mission areas do not constitute the entire universe of weapon acquisition cases, they are critically important. They have profound effects on the stability of strategic that

fall

balances, the volatility of arms competitions,

and the

credibility of na-

and bureaucratic influences play important roles in the early stage of such programs, contrary to what many accounts of the weapon acquisition process would have us believe. tional security policies. Strategic

Program Outcomes Development Objectives

The cases examined

book suggest several propositions about the timing, precision, rigidity, and extravagance of the Air Force's performance requirements, as well as its bias toward technologically ambitious development objectives. Many non-bomber programs have also featured extremely ambitious development objectives, which suggests that this bias is widespread in the American military. in this

The Timing of Performance Requirements. The Air Force set the performance requirements for its systems early in the weapon development process. This was true, for example, in the case of the B-35, B-36, B-47, B-52, B-70, B-i,

and B-iB programs. This was the stage

of the process

when

ambiguities about the long-term operational threats with which bombers would have to contend and about the technologies on which systems would be based were greatest.

Why

were performance requirements set so early? First, many of these programs were triggered by strategic developments, and in these cases the Air Force

was understandably anxious

threat as soon as possible, difficult

12.

9

A. Slomovic, Mig-21 Fishbed,

though

Rand Corporation

that

to define the

may have

emerging

been. 13 Sec-

Paper, P-7359-RGS, July 1987, pp.

8-

-

One would

expect requirements statements to appear later in the acquisition process if they were after-the-fact rationalizations for programs inspired by technological breakthroughs or industrial brainstorms. That was not the case here. 13.

The Origins ami Outcomes of Weapon Acquisition Programs

ond, the Air Force had to demonstrate a legitimate operational need and a system requirement that existing bombers could not fulfill; otherwise, it could not justify a new program. Third, requirements had to be defined (at least in draft form) before contractors could conduct specific design studies. Fourth, requirements statements were important for or-

ganizational reasons.

They

specified performance objectives

and timebegin breaking the development

which enabled the Air Force to problem down and acting; that is, they provided direction to the various branches and commands within the Air Force. Fifth, performance requirements were set quickly because it was possible to do so; the Air Force's operational doctrine provided a ready framework and established guidelines for making requirements decisions. tables,

The Precision of Performance Requirements. The Air Force tended to be specific about the speed, altitude, range, payload, defensive armament, and accuracy capabilities its new bombers were to have. The fact that precise requirements

process

is

were defined

at the

beginning of the acquisition

many

operational and tech-

surrounded development

efforts at this stage of

striking because, as noted earlier,

nological questions

still

the game.

The Air Force outlined

performance requirements in great detail for several reasons. Precise requirements gave contractors a clear set of design objectives. It was also important for political reasons for the Air Force to appear authoritative; the existence of poorly defined requirements would have implied that the Air Force was uncertain about the nature of the operational threat, which would have undermined a program's chances of getting funded by the civilian leadership in the executive branch and Congress. Precise requirements statements also provided the Air Force with detailed internal planning documents, which were important from an organizational standpoint. They gave the Air Force specific criteria for judging contractor performance, making source selection decisions, and determining contract payments. its

The Rigidity of Performance Requirements. Contrary to what one might think, the Air Force was not opposed in principle to modifying its performance requirements. It was perfectly willing to raise its requirements

over the course of a program, which it did on many occasions. The range requirement for the B-35 and B-36 was raised from 8,000 to 10,000 miles

The B-52's top speed requirement was raised from 450 to over 500 mph, and its cruising speed requirement was raised from 300 to 300 mph over the course of the late 1940s. The B-58's requirements were raised in the midst of the design competition itself. The B-yo's requirements became more demanding when a low-altitude requirement was inserted into the program in 1958. The requirements for the B-i became in 1941.

Flying Blind

more complicated in the early 1960s when the Air Force decided that the new bomber must penetrate at both high and low altitudes. The B-2 was redesigned in the mid-1980s to provide it with both high- and lowaltitude capabilities.

Why did

performance requirements on so many place, the operational environment tended to

the Air Force raise

occasions? In the

first

become more challenging

its

went by. As a result, the baseline for the Air Force's worst-case projections became gloomier as its programs progressed. The Air Force was understandably interested in hedging against as many of these new threats as possible. It was also in the Air Force's interest to point out to civilian decision makers that the air defense threat was worsening. This stance weakened the hand of civilians who preferred to get by with bombers already in the inventory. The Air Force had to be careful to stress, however, that the air defense problem was nonetheless manageable; otherwise, it would undermine the case for bombers in general. Finally, performance requirements went up whenever a contractor came through with an unexpected performance bonus. The Air Force was quick to capitalize on these bonuses by issuing

new performance

as time

requirements. This happened, for example, in both

and B-70 programs. At the same time, the Air Force was reluctant to lower its performance requirements, even when its programs faced serious technical problems, extended delays, or massive cost overruns. For example, the speed requirement for the B-45, B-46, B-47, and B-48 stayed high even though the contractors running these programs were confounded by the compressibility problem. The Air Force sent Boeing and North American back to the drawing board in the early stages of the B-70 program rather than compromise its performance objectives. Nor were the B-70's performance requirements lowered in the late 1950s when it became clear that they created a host of technical problems which slowed the program down and drove its costs up. The B-36's 10,000-mile range requirement persisted for years even though strategic circumstances changed fundamentally on several occasions. The intercontinental range requirement for the B-36 was set in 1941, when it appeared that Britain might fall to Germany; this requirement was not lowered when it became clear that Britain was going to withstand the German onslaught. Later, it was expected that the B-36 would be used in the Pacific theater, but the bomber's range requirement was not modified to take this into account. After the war, the B-36 was seen as a bomber that could be used to attack the Soviet Union with atomic bombs, but, again, its range requirement was not changed even though both its target set and its payload requirements had been revised. Nor were the program's formal requirements changed when the the B-52

The Origins and Outcomes of Weapon Acquisition Programs

Air Force decided to employ aerial refueling in late 1947. By the late 1940s, the B-36's 10,000-mile range requirement had achieved a stature that bore little relation to strategic reality. It was the standard by which

most people judged the program, even though an analysis of current strategic needs. 14

it

was not derived from

Similarly, the Air Force insisted for years that the B-i had to fly at Mach 1.2 at very low altitudes, even though flying the aircraft this way cut into its range significantly and added little to its ability to elude air

defenses. The Air Force, however, had an implicit faith in the operational value of supersonic capabilities. The civilian leadership in the Pentagon ultimately eliminated this requirement from the program. In looking back over these programs, we see only a few instances

when

the Air Force lowered performance requirements on its own or looked the other way when requirements could not be met. Special circumstances influenced each of these decisions. In the B-52 program,

speed and range requirements were temporarily lowered as first one and then the other was emphasized by the Air Force. In the end, though, both requirements were set at a high level. In the B-58 and B-70 programs, the Air Force established low-altitude requirements and then ignored them when it became clear that these unorthodox capabilities could only be obtained at the expense of high-altitude performance,

which

had emphasized

The Air Force permanently lowered the B-i's high-altitude speed requirement on two occasions in the early 1970s, but these were acts of desperation designed to cut the bomber's cost and prevent it from being canceled altogether.

Why

it

for decades.

did the Air Force rarely lower

its

performance requirements?

defenses generally improved as time went by, so strategic conditions tended to push requirements up, not down. Second, even if it became clear that a specified capability would not be needed, the Air Force did not object to having excess capabilities; they served as a hedge against unanticipated improvements in enemy air defenses. Third, it would have been politically inexpedient to lower performance requirements, which would have been an admission that the magnitude of the First, air

had been overestimated in the first place. Not only would this have undermined the Air Force's case for the bomber in question, it would have undermined the credibility of future threat projections or requirements statements. In the long run, it might have led to more civilian intervention in the requirements formation process, which the Air Force adamantly opposed as an infringement on its professional prerogatives and organizational autonomy. Fourth, lowering perforstrategic threat

14.

Singer

The persistence calls

of the B-36's range requirement is a classic example of what Max "the vitality of mythical numbers"; see "The Vitality of Mythical Numbers,"

Public Policy 23 (1971), 3-9.

Flying Blind

mance requirements on

might have led contractors to seriously. Finally, it would have taken

a regular basis

take requirements statements less

evidence about the strategic situation to get the Air Force to lower its requirements, given that it was inclined to make pessimistic assumptions about the future and had a vested interest in not backing down. The distant future was inherently ambiguous, however, so it be let alone should rarely became obvious that requirements could fairly clear

—

—

lowered. The Ambitious Nature of Performance Requirements. The Air Force routinely set its performance requirements far beyond the state of the art.

was

examined in this book. The 10,000mile range requirement for the B-35 and B-36 was certainly not attainable in 1941, for example. The B-47's speed requirement could not be met

This

when The

it

true in nearly every case

was

first set

first

issued because of the mysterious compressibility

effect.

of requirements for the B-52 was, as the Air Force later

admitted, so completely out of line with the state of the art that no contractor could be persuaded to participate in the program. The range

and supersonic speed requirements for the B-58 and B-70 were unattainable when they were first issued in the 1950s. In many cases, surprisingly little effort was made to determine whether Air Force requirements could be met given the prospects for technological progress. The Air Force, moreover, routinely pushed several requirements simultaneously, even though this aggravated inherent design trade-offs. The B-35 and B-36, for example, were confronted by challenging payload and speed requirements in addition to an extraordinarily demanding range requirement. These bombers were also to carry heavy defensive armament, which would impinge on range, speed, and payload capabilities. The B-47 was not optimized for speed alone; it also had demanding range, altitude, and payload requirements. The B-52 was to have the range of the B-36 and the speed of the B-47; addition, it was expected to carry a heavy payload. The B-58 was to have respectable range, payload, and avionics capabilities in addition to supersonic speed. The B-70 was to have intercontinental range as well as supersonic speed, even though the supersonic B-58 was not capable of meeting much less demanding range requirements. The B-70, moreover, had to bomb targets accurately while flying at 70,000 feet at Mach 3. The B-i was expected to have supersonic speed at both low and high altitudes, in addition to intercontinental range. Finally, the B-2 is supposed to have impressive range capabilities at both low and high altitudes, in addition to

being hard to detect.

Many

programs were unable to meet important performance requirements until major technological advances were made. The B-36 of these

[320]

— The Origins and Outcomes of Weapon Acquisition Programs

came

to rely heavily

on

engines and aerial refueling for its speed and range capabilities. Neither the B-47 nor the B-52 would have met their speed requirements without the swept wing. The B-58 program faced serious aerodynamic puzzles until the discovery of the transonic and supersonic area rules paved the way for efficient high-speed flight. The B-70 program was completely stalled until the discovery of the compression

jet

principle revolutionized designs for this class of aircraft. It is highly unlikely that any of these programs would have measured up in the long run if these breakthroughs had not been made. lift

important to note that many of these technological advances were totally unforeseen by Air Force planners and aircraft designers when performance requirements were first established for these programs. No It is

one in the AAF anticipated powered with jet engines or

would ultimately be that it would one day rely on aerial refueling in long-range operations. The discovery of the swept wing in 1945 solved what was beginning to look like an intractable aerodynamic problem. The discovery of the area rules and the compression lift principle in the 1950s were advances in basic aeronautical research, to which Air Force requirements planners seem to have paid no attention. In many cases, it was simply fortuitous that contractors stumbled in 1941 that the B-36

across the basic research that

made

necessary performance gains possible. This was certainly true in the B-47, B-58, and B-70 programs. Basic scientific research, moreover, was not influenced by the Air Force's immediate operational demands. Swept-wing aircraft were pioneered by

German

aerodynamicists; one can safely say that their research was not particularly sensitive to the AAF's performance requirements. The area rules and the compression lift principle were discovered by scientists

engaged

in basic research at

NACA;

their research

was unconnected

to

the exigencies of the Air Force's military requirements.

many

programs were ambitious from a developmental standpoint. The nine-part framework outlined in Chapter 1 enables us to categorize each of the fifteen studied programs according to the degree of technological advance needed to meet the Air Force's performance requirements (see Table 12). These programs cluster toward the ambitious end of the spectrum. The B-35, B-70, and B-2 programs involved (or involve) new and radically different system designs. They clearly belong in category 1. Although the B-47 and B-58 involved important advances at the system level the B-47's use of swept-wing technology and the B-58's use of the transonic and supersonic area rules their designs were based in part on breakthroughs made in other programs. The B-47 drew on German aeronautical research, and the B-58 drew on other Convair work with high-speed, delta-wing aircraft. The B-47 ar, d B-58 therefore belong in category 2. The nuclear-powered In short,

of these

—

[3 21

I

Flying Blind

Table 12. Technological ambitiousness in the U.S. strategic

Program

Level of technological ambitiousness 1.

2. 3.

4. 5.

6. 7.

8

.

9.

bomber program

Radically different system

B-35, B-70, B-2

New New

ANP

system technology subsystem technology Several subsystems need major improvement One subsystem needs major improvement Several subsystems need some improvement One subsystem needs some improvement

Contemporary technologies integrated No new technology or hardware

bomber needed

a

fundamentally

B-47, B-58

B-36, B-52,

B-iA

B-45, B-46, B-48, B-49, B-iB

B-60

new

propulsion system. It conseThe 8-36, 8-52, and B-iA involved ad-

quently belongs in category 3. vances in aerodynamics and at the subsystem level. The B-36 used large pusher-propeller engines, and it needed new electrical and hydraulic subsystems, for example. The B-52 needed new engines as well as advanced navigation and defensive armament subsystems. The B-iA also needed new engines, as well as new engine inlets and advanced avionics.

The

B-36, B-52,

B-46, B-48, B-49,

advances

in

and

B-i

A

therefore belong in category

4.

The

B-45,

aR d B-iB belong

one main

in category 5 because they required area: propulsion in the case of the B-45, B-46, and

an d avionics in the case of the B-1B. 15 Finally, the B-60 belongs in category 8 because Convair had only to integrate contemporary technologies to build a swept-wing, all-jet B-48; flight control in the case of the B-49;

version of the B-36.

The tendency to push the state of the art is not unique to Air Force bomber projects. In the late 1950s, as discussed earlier, the Air Force and Navy began to formulate requirements for new fighters that would perform, respectively, fighter-bomber and fleet air defense missions. Each service established demanding requirements. The Air Force wanted an aircraft capable of carrying tactical nuclear weapons internally; cruising across the Atlantic without refueling; flying at high subsonic speeds at

low altitudes over long distances; and engaging enemy fighters altitudes at speeds well in excess of Mach 2. The Navy wanted an

at

high

aircraft

capable of cruising at high altitudes for long periods of time while armed with long-range air-to-air missiles under its wings. To intercept enemy aircraft before they could launch missiles against American surface Chapter 3, the AAF considered the B-46 and B-48 more adventurous than the B-45 because the former featured novel wing and fuselage designs while the latter was quite conventional in these respects. Even so, it would be going too far to say that the B-46 and B-48 were radically different from contemporary aircraft aerodynamically or that they featured new technology at the system level. All three programs belong in category 5. 15.

As discussed

in

,

The Origins and Outcomes of Weapon Acquisition Programs

would need an extremely sophisticated panoramic search radar. It would also have to be rugged enough to operate from aircraft carriers. Both sets of requirements were demanding as they stood, but the development problem became even more challenging when Secretary of Defense McNamara forced the two services to work together in the early 1960s and build one aircraft, the F-111, that both ships, this fighter

could use. 16 To cite just one more example, the Army's MBT-70 tank program in the mid-1960s was designed to push the state of the art in many areas. It

was

to feature,

among

other things, an automatic

gun

loader, a fully

and a counter-rotating turret capsule for the driver, a laser range finder and a new ballistic computer, a guided missile launcher, a variable-height hydropneumatic suspension that enabled the tank to "squat” and reduce its visibility, and a variable compression ratio engine that could use a wide variety of fuels. 17 The tendency to push the state of the art in established mission areas is a strong one. 18 stabilized turret

The Bias tozvard Technologically Ambitious Development Objectives. A powerful set of strategic and bureaucratic forces led Air Force decision makers to adopt ambitious development objectives in most of their

bomber First,

acquisition programs. Three strategic considerations stand out.

Air Force leaders believed that technology

cisive in warfare. This conviction

ences in World War

grew out

was frequently de-

of the Air Service's experi-

which technologically obsolete aircraft were often fatally ineffective, and it was strongly reinforced by the AAF's experiences in World War II, in which technological breakthroughs in the development of radar, jet engines, rocketry, and atomic weaponry changed the course of the conflict. Air Force leaders concluded, not unreasonably, that advanced weapons would be more effective in combat, hedge against a wider range of threats, and become obsolete slower than technologically primitive systems. This view came to be shared by the rest of the American military establishment as well. As one student I,

in

Based on Art, TFX Decision pp. 15-27; Coulam, Illusions of Choice, pp. 36-56. Based on Alexander, Armor Development p. 108. 18. Other notable examples include the Navy's Trident submarine and D-5 SLBM programs; the Army's Cheyenne helicopter, Sergeant York antiaircraft gun, and Bradley infantry fighting vehicle programs; and the Air Force's F-102, C-5A, F-15, and MX programs, to name but a few. The F-102 program was discussed earlier. For more details on the other programs, see Dalgleish and Schweikart, Trident, pp. 17-44; G. K. Smith et al.. The Use of Prototypes in Weapon System Development Rand Corporation Report, R-2345-AF, March 1981, pp. 155-158; Bill Keller, "Demise of the Sgt. York Gun," Nezv York Times, Dec. 2, 1985; Hugh Lucas, "U.S. Army under Fire Again over Bradley," jane's Defense Weekly, Nov. 22, 1986, p. 1209; Berkeley Rice, The C-5A Scandal (Boston: Houghton Mifflin, 1971); James Fallows, National Defense (New York: Random House, 1981), pp. 100-106; Lauren H. Holland and Robert A. Hoover, The Decision (Boulder, Colo.: Westview, 1985), pp. 124-136. 16.

17.

,

,

MX

[323]

.

Flying Blind

of U.S. defense policy observed,

American

military planners

became

"enamored

of the promise of threshold technology" after 1945. 19 Second, Air Force leaders believed again, not unreasonably

—

technological innovation

was

— that

the United States' area of comparative ad-

vantage with respect to the Soviet Union and that qualitative superiority

was needed

in the U.S. force structure to offset Soviet quantitative ad-

made great technological strides American weapon development programs had to

vantages. But since the Soviet Union after

World War

push the

II,

they were to maintain an edge over their Soviet counterparts. The Soviet threat generally seemed to be quite formidable. state of the art aggressively

if

Third, Air Force leaders believed that they had to

make

pessimistic

assumptions about future operational threats in order to protect American national security. Given that the lead times for American weapon development programs could be five, ten, even fifteen years or more, requirements planners had to peer into the distant future when they established performance standards. The distant future, however, was inherently ambiguous, so worst-case assumptions about future operational conditions could be grim indeed. 20 It is important to note that these sentiments were widely shared by both military and civilian policy makers in the United States throughout most of the postwar period, although civilians rarely became directly involved in requirements formation. The Air Force's predisposition to favor technologically ambitious weapon development projects was strongly reinforced by several bureaucratic considerations. First, the Air Force wanted to safeguard its most cherished military mission, strategic bombardment, and one way of doing this was fielding long-lived bombers. Technologically advanced bombers were more likely to endure than their less advanced counterparts, if everything else was equal, because they were less likely to become obsolete in the near term. Building technologically advanced bombers was therefore seen as a

way

of preserving the Air Force's organizational essence.

this line of

reasoning was that technologically

The flaw in adventurous programs

occasionally self-destructed

Second, Air Force doctrine placed a premium on exceptional range, speed, altitude, defensive armament, and accuracy capabilities, which also led

it

the 1930s,

to

push the

when

the

state of the art.

AAC

intensified

Range has been emphasized since its

efforts to build

an

aircraft capa-

Robert L. Perry, "The Interaction of Technology and Doctrine in the U.S. Air Force," Alfred F. Hurley and Robert C. Ehrhart, eds.. Air Power and Warfare (Washington: Government Printing Office, 1979), pp. 388-389. 20. In fact, the future was usually too ambiguous to provide clear-cut guidance on specific requirements. 19.

in

[324]

The Origins and Outcomes of Weapon Acquisition Programs ble of conducting legitimate strategic operations. Speed, altitude,

defensive

armament

capabilities

and

have also been emphasized since the

American strategic bombardment doctrine held that it was possible to build bombers that could transcend existing air defenses and effectively sweep them aside. The B-17 was the prototype of the transcendent bomber. The B-70 capitalized on a technological finesse of dramatic proportions and held out the promise of exceptional range, speed, and altitude all at the same time. The Air Force believes that the B-2 has the same kind of transcendent capabilities that it can effectively neutralize existing air defenses, and that hostile powers will be unable to afford defenses sophisticated enough to threaten its new bomber. Finally, strategic bombardment doctrine emphasized that it was important to 1930s.

—

be able

with a high degree of accuracy. Third, American defense spending has gone through several "feastor-famine" cycles since 1900. This gave the services a strong incentive to build extremely capable systems during budgetary feasts because they needed weapons that would not become obsolete during the famines that inevitably followed. In other words, the services had a strong incentive to advance technology as much as possible whenever they got a chance to build a new weapon, because they had no idea when they to attack targets

would be

able to build a successor to

it.

It

was not

a coincidence, there-

and ANP programs took off in 1951 and the B-2 design competition began in 1980, when defense budgets were on the rise. Fourth, although the Air Force could start small research and development projects on its own, full-scale bomber programs required civilian approval early in the acquisition process. The Air Force therefore had to sell its programs to the civilian leadership in the executive branch and Congress. It was difficult to sell bombers that were less advanced technologically because they were hard to distinguish from existing forces and because they would probably have to be replaced sooner rather than later. The B-36 was criticized on these grounds in the late 1940s, as was the B-iB in the early 1980s. Programs with more ambitious development objectives were much easier to sell. The existence of demanding performance requirements reinforced the idea that the strategic threat was formidable and that existing forces faced operational obsolescence. In addition, state-of-the-art bombers were easy to distinguish from existing forces because they could do things existing forces could not do. Finally, it was advantageous from the Air Force's standpoint to be able to argue that its new bombers would not have to be replaced in the near term. The B-2, for example, has enjoyed strong civilian support for these reasons. The fact that the services have a near monopoly of information on developmental and operational matters obviously works to their advanfore, that the B-58

tage in this context.

Flying Blind

Although pursuing ambitious development objectives relieved some of the Air Force's strategic and bureaucratic anxieties, it made life more difficult from a technological standpoint. The Air Force dealt with this trade-off by sidestepping it; it simply assumed that the technological challenges its programs faced would eventually be overcome. Similarly, the Air Force avoided the implicit trade-offs between range, speed, altitude, payload, defensive armament, and bombing accuracy by assuming that technological developments would ultimately ease, if not eliminate, them. The Air Force was consequently able to avoid the painful problem of prioritizing its values and choosing among them 21 The Air Force avoided these trade-offs by downplaying the technological risks to programs involved, a final bureaucratic consideration that led it to pursue ambitious development objectives. In many cases, technological problems were not assessed carefully when performance requirements were established. As we have seen, technological assessments were rushed in many cases, if they were made at all. Requirements planners, therefore, were not always fully aware of the technological implications of their demands. But even when they were, they had an implicit faith in the ability of program managers to come up with the requisite technological miracles. As one study of American weapon acquisition observed, American military leaders believed "that almost any conceivable weapon could be built if they were willing to make a sufficient investment of ingenuity, resources, and time ." 22 It was certainly true that defense contractors produced apparent miracles and real .

breakthroughs over the years. This track record helped to nurture the belief that "holding the engineers' feet to the fire" with demanding performance requirements paid off in the long run.

The Air

program managers to make technological advances when needed was reinforced by the "can do" attitude that permeated its technical commands and the defense industry. The technical commands had a vested interest in working on new, interesting weapon projects, and it was not in anyone's best career interests to try to hold back a strategic bomber program. The technical Force's faith in the ability of

commands therefore rarely uttered a discouraging word when asked to comment on the feasibility of new projects and they were often ignored when they tried to inject a note of realism into the proceedings.

—

defense contractors had a vested interest in winning potenlucrative development contracts and, as we have seen, they were

Similarly, tially

Trade-off avoidance

discussed in detail in Robert Jervis, Perception and Mispercep(Princeton: Princeton University Press, 1976), chap. 4; John D. Steinbruner, The Cybernetic Theory of Decision (Princeton: Princeton University Press, 1974), 21.

is

tion, in International Politics

chaps. 3-4. 22.

Perry, “Interaction of

Technology and Doctrine," pp. 388-389.

The Origins and Outcomes of Weapon Acquisition Programs

occasionally chastised when they told the Air Force that faced difficult technical hurdles. Finally, the Air Force

was

able to sidestep

its

projects

problems because it was always possible that the technological breakthroughs it needed would appear sometime during the course of a program. There was always enough ambiguity in the situation for this to be conceivable. The Air Force tended to see the possible as probable because it was necessary to do so if it was to avoid the trade-offs that loomed over its acquisition programs. It is important to note that the strategic and bureaucratic forces that pushed U.S. bomber programs toward ambitious development objectives are not unique to this set of cases. The belief that it is important to push technology for strategic reasons is widely held in the United States and elsewhere. Every military organization is interested in perpetuating core activities

and

its

trade-off

traditional mission areas.

Other operational doctrines also place a premium on weapons with superior performance capabilities. All of the American military services face the fundamental problem of selling their programs to the civilian leadership. Therefore, it is not surprising that other military services push the state of the art when they set up acquisition programs within established mission areas.

Procurement Strategies

A

wide range of procurement

were used in the cases studied in this book, although concurrent strategies were employed more often than not. The framework outlined in Chapter 1 enables us to categorize these programs in terms of how sequential or how concurrent they were (see Tables 13 and 14). As we can see, with the exception of the bomber programs of the late 1940s, these cases cluster toward the concurrent end of the spectrum. Moreover, the B-45, B-47, and B-52, which started as highly sequential endeavors, were eventually accelerated into production and thus ultimately employed a fair amount of concurrency as well. The only other program to feature a moderate amount of concurrency was the B-i A, which was influenced by Deputy Secretary of Defense David Packard's acquisition reforms in the early strategies

1970s.

Historical Patterns.

What we now

call

highly sequential strategies were

almost universally employed in American weapon acquisition programs in the interwar period. Procurement activities in the 1920s and 1930s were constrained by elaborate federal rules and regulations, and formal competitions were required in almost every procurement action. A great

o

A I

CQ

xxxx

^

-

i

XX X X *1 XC _C

oxxx

13

and

oo

(33

3 ("before"

CQ rl

X l

XXXXXXXX

00

b'X 03

> *«c

X (33

v

y y

< X rH I

0 1

CQ

XXOOOXXX

I

IA

> X c 3 (33

00

LA rA

X 00 LA i

CQ

assessments

XXXXXXXX

different

Vi 03

03

£ c

*£

c/5

Two

r-

00

£ 2 0

£