Digital Printing of Textiles (Woodhead Publishing in Textiles) 0849391008, 9780849391002, 1855739518, 9781855739512, 9781845691585

Citation preview

Digital฀printing of฀textiles Edited฀by฀H.฀Ujiie

Digital printing of textiles

Related title:

Total colour management in textiles (ISBN-13: 978-1-85573-923-9; ISBN-10: 1-85573-923-2) Managing colour from the design stage to the finished product can be difficult as colour perception is subjective and can therefore be inconsistent. Total colour management in textiles covers all aspects of managing colour from the design stage to the final product ensuring that the designer's vision is fulfilled in the finished colour. There have been many new developments in the area of colour measurement and colour perception which are discussed. These include discussion of the sensory effect of colour for design and use in product development, and digital colour simulation. Details of this book and a complete list of Woodhead's titles can be obtained by: · visiting our website at www.woodheadpublishing.com · contacting Customer Services (e-mail: [email protected]; fax: +44 (0) 1223 893694; tel.: +44 (0) 1223 891358 ext. 30; address: Woodhead Publishing Limited, Abington Hall, Abington, Cambridge CB1 6AH, England) If you would like to receive information on forthcoming titles in this area, please send your address details to: Francis Dodds (address, tel. and fax as above; email: [email protected]). Please confirm which subject areas you are interested in.

Digital printing of textiles Edited by H. Ujiie

Published by Woodhead Publishing Limited in association with The Textile Institute Woodhead Publishing Limited Abington Hall, Abington Cambridge CB1 6AH England www.woodheadpublishing.com Published in North America by CRC Press LLC 6000 Broken Sound Parkway, NW Suite 300, Boca Raton FL 33487 USA First published 2006, Woodhead Publishing Limited and CRC Press LLC ß 2006, Woodhead Publishing Limited The authors have asserted their moral rights. This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. Reasonable efforts have been made to publish reliable data and information, but the authors and the publishers cannot assume responsibility for the validity of all materials. Neither the authors nor the publishers, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by this book. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming and recording, or by any information storage or retrieval system, without permission in writing from Woodhead Publishing Limited. The consent of Woodhead Publishing Limited does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from Woodhead Publishing Limited for such copying. Trademark notice: product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. Library of Congress Cataloging in Publication Data A catalog record for this book is available from the Library of Congress: Woodhead Publishing Limited ISBN-13: Woodhead Publishing Limited ISBN-10: Woodhead Publishing Limited ISBN-13: Woodhead Publishing Limited ISBN-10: CRC Press ISBN-10: 0-8493-9100-8 CRC Press order number: WP9100

978-1-85573-951-2 (book) 1-85573-951-8 (book) 978-1-84569-158-5 (e-book) 1-84569-158-X (e-book)

The publishers' policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp which is processed using acid-free and elementary chlorine-free practices. Furthermore, the publishers ensure that the text paper and cover board used have met acceptable environmental accreditation standards. Project managed by Macfarlane Production Services, Dunstable, Bedfordshire ([email protected]) Typeset by Godiva Publishing Services Ltd, Coventry, West Midlands Printed by TJ International Limited, Padstow, Cornwall, England

Contents

Contributor contact details

1

The evolution and progression of digital printing of textiles

xiii

1

V CA H I L L , VCE Solutions, USA 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11 1.12 1.13 1.14 1.15 1.16 1.17 1.18 1.19 1.20

2

Introduction The origins of digital textile printing technologies Digital carpet printing Sublimation Thermal inkjet and textile printing Seiren Digital grand format and textile printing FESPA 1996 FESPA and ITMA 1999 ITMA 1999 Drupa 2000 Heimtextil 2001 DPI 2001 ITMA 2003 Drupa 2004 SGIA 2004 FESPA 2005 Other key elements Conclusion References

1 2 3 4 4 5 5 6 6 7 7 7 8 8 9 10 10 14 14 15

A designer's perspective ± digital versus traditional

16

L NI C O L L , Consultant, Italy 2.1 2.2

Introduction What difference does digital make?

16 17

vi

Contents

2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10

How is this done using traditional methods? How do they compare? How can the designer use these twinned technologies? Freedom Thinking about creativity Resistance Transparency The new market

18 22 23 23 24 24 25 26

Part I Printer/print head 3

Ink jet printing technology (CIJ/DOD)

29

E. MA R I A N O FR E I R E , DuPont Ink Jet, USA 3.1 3.2 3.3 3.4 3.5 3.6 3.7

4

Introduction Ink jet technologies Aspects to consider and metrics to use in the print head selection process Companies currently active in print head technology Future trends Sources of further information and advice References

Drop formation and impaction

29 29 45 48 49 49 52

53

W W CA R R , H PA R K , H OK , R FU R B A N K and H DO N G , Georgia Institute of Technology, USA and J F MO R R I S , City College of New York, USA 4.1 4.2 4.3 4.4 4.5

5

Introduction Drop formation from particle-laden liquids Drop impaction Future trends References

Industrial production printers ± DuPont ArtistriTM 2020 textile printing system

53 54 57 65 66

69

M RA Y M O N D , DuPont Ink Jet, USA 5.1 5.2 5.3 5.4 5.5 5.6

Introduction Industry needs Markets and applications ArtistriTM 2020 printer Competitive environment ArtistriTM 2020 textile printing technology

69 69 70 72 73 74

Contents 5.7 5.8 5.9 5.10 5.11 5.12 5.13 5.14

6

Process color printing versus spot color printing Cost of printing Opportunities and new markets ArtistriTM Technology Center Applications support, technical service and training Future trends Sources of further information and advice Bibliography

Industrial production printers ± DReAM

vii 79 79 80 81 82 82 83 83

84

L CA C C I A and M NE S P E C A , Reggiani Macchine S.p.A., Italy 6.1 6.2

84

6.4

The DReAM project in the present textile printing scenery Goals of the project and description of the DReAM machine (technical and technological parts: Reggiani, Ciba Specialty Chemicals and Scitex Vision) New opportunities offered by the new Reggiani digital printing machine: Digital Technological Center (DTC) Bibliography

7

Industrial production printers ± Mimaki'sTx series

98

6.3

87 95 97

H KO B A Y A S H I , Mimaki Industries, Japan 7.1 7.2 7.3 7.4 7.5

8

Evolution of digital printing Marketing profile of Mimaki's Tx series Market needs for digital textile printing Technical issues and solutions The future of digital printing

Integration of fabric formation and coloration processes

98 99 101 101 120

123

B R GE O R G E , D WO O D , M GO V I N D A R A J , H UJ I I E , M FR U S C E L L O , A TR E M E R E , and S NA N D E K A R , Philadelphia University, USA 8.1 8.2 8.3 8.4 8.5

Introduction Experimental Results and discussion Conclusions References

123 126 129 142 143

viii

Contents

Part II Printer software 9

Digital image design, data encoding and formation of printed images

147

T L DA W S O N , formerly of University of Manchester, UK 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8

Introduction Computer aided design, editing and data storage systems Pixel and image formation by ink jet printers Control of the printing machine Machine performance monitoring Future trends Sources of further information and advice References

10

Digital colour management

147 148 152 159 159 160 161 162

163

T L DA W S O N , formerly of University of Manchester, UK 10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8 10.9

11

Introduction General numerical colour specifications Characterising display, input and output devices Colour gamut and rendering intent Colour communication Colour reproduction performance of equipment operated with a CMM Future trends in colour management Sources of further information and advice References

ICC color management for digital inkjet textile printing

163 166 168 173 175 175 177 177 178

180

E. LO S E R and H-P TO B L E R , ErgoSoft AG, Switzerland 11.1 11.2 11.3 11.4 11.5 11.6 11.7 11.8 11.9 11.10

Introduction Overview of textile colors and common color spaces ICC basics ICC advantages and disadvantages Requirements and problems for ICC profiling Current technologies Results Conclusion and future trends Sources of further information and advice References

180 181 182 184 185 187 194 196 197 198

Contents

ix

Part III Digital printing coloration 12

Substrate preparation for ink-jet printing

201

12.1 12.2 12.3 12.4 12.5 12.6 12.7 12.8 12.9 12.10

C HA W K Y A R D , University of Manchester, UK Introduction Ink systems Fabric pre-treatments Pre-treatments for ink-jet printing Post-treatments Jet printing machines Limitations Future trends Bibliography References

201 204 206 207 213 214 214 215 215 215

Pigmented ink formulation

218

13

Z FU , Rohm and Haas, Philadelphia, USA 13.1 13.2 13.3 13.4 13.5 13.6 13.7

Introduction Overview Pigmented ink formulation for digital textile printing Tests and test methods for pigmented textile inks Optional pre- and post-treatments for pigmented digital textile printing White ink Sources of further information and advice

14

Formulation of aqueous inkjet ink

218 219 221 227 231 231 232

233

H NO G U C H I and K SH I R O T A , Canon Inc., Japan 14.1 14.2 14.3 14.4 14.5 14.6 14.7 14.8 14.9 14.10

Dye±fiber interaction Organic solvents and surface energy of ink Time-dependent phenomena and surface-active components Additives Reliability Production process of inkjet-printed textiles Reactive dye ink Disperse dye ink Acid and direct dye ink formulation References

233 235 235 237 237 240 240 245 250 251

x

Contents

15

Effect of pretreatment on print quality and its measurement

252

Y. K. KI M , University of Massachusetts-Dartmouth, USA 15.1 15.2 15.3 15.4 15.5

Introduction Textile pretreatments for inkjet printing Effect of pretreatments on print quality Concluding remarks and future trends References

16

Ink jet printing of cationized cotton with reactive inks

252 254 258 272 274

276

P J HA U S E R , North Carolina State University, USA and M KA N I K , University of Uludag, Turkey 16.1 16.2 16.3 16.4 16.5

Introduction Experimental Results and discussion Conclusions References

276 278 280 288 288

Part IV Design and business 17

Digital printing and mass customization

293

2

M FR A L I X , [TC] , USA 17.1 17.2 17.3 17.4 17.5 17.6 17.7 17.8 17.9 17.10

Introduction From craft to mass production to mass customization Limitations of mass customization Time, technology, and connectivity Product life cycles Forecasting the opportunities Traditional supply chains Direct digital printing supply chains Future trends in the digital supply chain References and bibliography

293 295 297 298 299 300 304 307 309 310

18

Just-in-time printing

312 2

K MA G U I R E KI N G , [TC] , USA 18.1 18.2 18.3 18.4

Introduction Enabling the process Just-in-time order processing Case studies

312 314 323 327

Contents 18.5 18.6

19

Conclusion References

Design and workflow in digital inkjet printing

xi 336 336

337

H UJ I I E , Philadelphia University, USA 19.1 19.2 19.3 19.4 19.5 19.6

Introduction Evolution of textile printing workflow New design styles New definitions for the textile printing industry Future trends References

337 338 343 350 354 354

Index

356

Contributor contact details

(* = main contact) Chapter 1 Mr Vincent Cahill VCE Solutions, a division of VC Enterprises Inc. 219 North Franklin Street Waynesboro, PA 17268 USA Tel: +1 717 762 9196 Fax: +1 717 762 9683 E-mail: [email protected] [email protected]

Chapter 4 Professor W.W. Carr, H. Park, H. Ok, R. Furbank and H. Dong The School of Polymer, Textile and Fiber Engineering Georgia Institute of Technology 801 Ferst Drive NW MRDC 1 Atlanta, GA 30332-0295 USA Tel: +1 404 894 2490 Fax: +1 404 894 8780 E-mail: [email protected]

Chapter 2 Mr Lee Nicoll Printed textiles consultant Formerly of Ratti, Italy

J.F. Morris City College of New York USA

E-mail: [email protected] [email protected]

Chapter 5 Dr Mike Raymond DuPont Ink Jet Barley Mill Plaza 30/2312 4417 Lancaster Pike PO Box 80030 Wilmington, DE 19880-0030 USA

Chapter 3 Dr E. Mariano Freire DuPont Ink Jet Experiment Station E357/1138 200 Powder Mill Road Wilmington, DE 19880-0357 USA E-mail: [email protected]

E-mail: [email protected]

xiv

Contributor contact details

Chapter 6 Dr Lucia Caccia* and Dr Marco Nespeca Reggiani Macchine S.p.A. Via Zanica, 17/O PO Box 41 24050 Grassobbio (Bergamo) Italy Tel: +39 035 3844511 Fax: +39 035 526952 E-mail: [email protected] [email protected] Chapter 7 Dr Hisayuki Kobayashi Mimaki Industries 1333-3 Kazawa Tohmi-shi Nagano, 389-0514 Japan Tel/Fax: +81-268-64-2413 E-mail: [email protected] [email protected] Chapter 8 Professor Brian R. George*, Deanna Wood, Professor Muthu Govindaraj, Professor Hitoshi Ujiie, Monica Fruscello, Alexa Tremere and Swapnil Nandekar School of Engineering and Textiles School House Lane and Henry Avenue Philadelphia, PA 19144 USA E-mail: [email protected]

Chapters 9 and 10 Dr T.L. Dawson Formerly of University of Manchester UK E-mail: [email protected] Chapter 11 Edgar Loser and Hans-Peter Tobler* ErgoSoft AG Moosgrabenstr. 13 CH-8595 Altnau Switzerland E-mail: [email protected] Chapter 12 Dr Chris J. Hawkyard 10 Kempton Close Hazel Grove Stockport SK7 4SG UK Tel. +44 (0)161 4838715 E-mail: [email protected] Chapter 13 Dr Zhenwen Fu Rohm and Haas Co. 100 Independence Mall West Philadelphia, PA 19106-2399 USA E-mail: [email protected] Chapter 14 Dr Hiromichi Noguchi* and Koromo Shirota Canon Inc. Inkjet Supply Product Operation 30-2 Shirmomaruko 3-chome

Contributor contact details Otta-ku Tokyo 146-8501 Japan E-mail: [email protected] [email protected] Chapter 15 Professor Yong K. Kim Textile Sciences Department University of MassachusettsDartmouth N. Dartmouth, MA 02747 USA Tel: +1 508-999-8452 Fax: +1 508-999-9139 E-mail: [email protected] Chapter 16 Professor Peter J. Hauser* North Carolina State University 2401 Research Drive Raleigh, NC 27695 USA Tel: +1 919 513 1899 Fax: +1 919 515 6532 E-mail: [email protected] Dr Mehmet Kanik University of Uludag College of Engineering and Architecture Textile Engineering Department Gorukle-16120, Bursa Turkey Tel: +90 224 442 81 74/1052(ext.) Fax: +90 224 442 80 21 E-mail: [email protected]

xv

Chapter 17 Dr Michael Fralix [TC]2 211 Gregson Drive Cary, NC 27511 USA Tel: +1 919-380-2173 Fax: +1 919 380 2181 E-mail: [email protected] Chapter 18 Dr Kerry Maguire King [TC]2 211 Gregson Drive Cary, NC 27511 USA Tel: +1 919-380-2173 Fax: +1 919 380 2181 E-mail: [email protected] Chapter 19 Professor Hitoshi Ujiie Director The Center for Excellence of Digital Ink Jet Printing of Textiles School of Engineering and Textiles School House Lane and Henry Avenue Philadelphia, PA 19144 USA Tel: +1 215 951 2682 Fax: +1 214 951 2651 E-mail: [email protected]

1 The evolution and progression of digital printing of textiles V C A H I L L , VCE Solutions, USA

1.1

Introduction

The saga of digitally printing and dyeing of fabrics, yarns and garments involves a past of a few decades, a dynamic present and likely a bright future. This introduction accounts for the origins and evolution of textile printing to digital solutions. It identifies some of the many creators and pioneers of these technologies and assesses their impact on the textile printing industry. It discusses a few of the false starts that in turn contributed to sustained successes of digital printing technologies for textile decoration. It uses the exhibitions that have witnessed the introduction of innovative digital technologies for textile printing as road marks that indicate trends and point the way to digital textile printing's bright future. In focusing on the market demand for textile printed applications, it attempts to answer questions such as: · What are the market forces driving the adoption of digital technologies for printing textiles? · What are the characteristics and qualities of digital printing that either favor or discourage its adoption for meeting market demand for decorated fabric and fibers? · What are textile applications that digital printing can supply more cost effectively than existing analog printing methods? · How is technology evolving to address market demand? · What are the global trends that shed light on the future of digital printing of textiles? This chapter opens the door and invites the reader into the exciting world of digital textile printing. It introduces this volume's tour of the many chambers of the digital textile print edifice that subsequent sections describe.

2

Digital printing of textiles

1.2

The origins of digital textile printing technologies

An early example of textile printing is found on a block-printed tunic dated from the fourth century CE.1 Evidence suggests that carved block printing, also known as xylography, originated about the fourth century in China and initially found use in printing textiles and short Buddhist texts that believers carried as charm protection.1,2 Sui emperor Wen-ti ordered the printing of Buddhist images and scriptures in an imperial decree of 593. The British Museum houses the oldest known block-printed book, the Diamond Sutra, dated 868 CE from Dunhuang, China. Block printing of textiles began to flourish in Surat in Gujarat (India) during the twelfth century for the printing of wall hangings, canopies and floor spreads.3 The printing of textiles spread around the world along the Silk Road and through the spice trade. While we can only infer the origins of textile printing from the few artifacts that have survived, digital printing evolved in an age of record keeping. In 1686, Edme Mariotte suggested the basis for inkjet printing with the publication of his seminal work on fluid dynamics, `Traite du movement des eaux et des autres corps fluids'. It included observations on drop formation of fluids passing through a nozzle. Ebenezer Kinnersley added to this foundation when he demonstrated that electrical current could pass through water in 1748. During the following year of 1749, l'Abbe Nollet examined the effects of static electricity on the flow of drops from a capillary tube. Lord Kelvin (Sir William Thomson) received the first patent for an inkjet printing system in 1867, `Receiving or Recording Instruments for Electric Telegraphers'. Eleven years later in 1878, Lord Rayleigh (Sir John William Strutt) described the role of surface tension in drop formation. The 1920s and 1930s witnessed patent applications and issuances for inkjet recording devices, including notable inventions from Richard Howland Ranger and Francis G. Morehouse in 1928, Clarence W. Hansell4 for an electrically charged recycling device in 1929, and Kurt Gemscher in Germany in 1938. During the same year, 1938, Chester Carlson invented analog electrophotography in Astoria, Queens, New York. It took Carlson and his subsequent partner company Haloid over 20 years and a few intermediate steps along the way, such as the Haloid A1 in 1949 and Copyflo in 1955, to deliver a successful office plain paper copier with the Xerox 914 in 1959. The A1 failed for the purpose that Haloid intended it as an office copier, but succeeded as a plate maker for commercial printing. Digital laser versions of electrophotography produced transfers in the 1980s to decorate fabrics, particularly T-shirts and other sewn garments and accessories. Researchers at Georgia Tech and North Carolina State University investigated the feasibility of printing fabric with electrophotography with some success.

The evolution and progression of digital printing of textiles

3

In 1959, the Research Labs of Australia began exploiting its invention of developing electrostatic images with liquid toners. Xerox and others developed a similar liquid toner variation on electrophotography for wide format printing, electrostatic printing. In 1979, Xerox introduced its 2080 engineering copier. Xeroxcolorgrafx, Raster Graphics, 3M, Nippon Steel-Synergy Computer Graphics, Calcomp, Silvereed, Phoenix Precision Graphics and others advanced this technology with the development of E-stat printers during the late 1980s and 1990s. Almost all of these companies have discontinued production of their electrostatic printers. Only 3M of St Paul, Minnesota, USA, currently supplies and supports an electrostatic printer, the Scotchprint 2000. Hilord supplies both pigment resin and sublimation dye toner for electrostatic printers. Beta Color of Ontario, California, and others have developed processes for using Scotchprint 2000 for cost-printing polyester and Nylon 6.6 fabrics with sublimation toners. In 1949, Elmquist applied for a patent for `Measuring Instrument of the Recording Type'. Two years later in 1951, Siemens released the first commercially produced inkjet printer based on the Elmquist patent, the Elema Oscilomink. Carl Helmuth Hertz and Sven Eric Simmonsson applied for patent on high-resolution continuous inkjet in 1965. This invention and other inventions of Dr Hertz and his colleagues at Sweden's Lund Institute of Technology led to the development of the Stork and Scitex Iris proofing systems. This type of continuous inkjet technology features mutual charged droplet repulsion that produces very fine ink droplets at a very high frequency. It can produce high apparent resolution images with many gray or halftone levels. It suffers, however, from slow print production speed, a slight background from stray droplets, and the complications inherent with recirculation of unprinted toner drops. This process uses dye-based colorants that lack the level of permanence that pigments provide. The textile and fashion design industries have used these systems since their introduction. Stork also developed a successful proofing system for the commercial print industry in conjunction with DuPont. In 1967, Professors Sweet and Cummings of Stanford University in California applied for a patent on a binary continuous inkjet array. In 1968, printer manufacturer A.B. Dick commercialized Sweet's invention with the Videojet 9600. This device launched the marking and coding industry on its digital path. While early applications of this technology were primarily for coding cans, containers and other packaging, it was capable of marking fabric as well.

1.3

Digital carpet printing

In the early 1970s, Milliken of Spartanburg, South Carolina, USA, developed a digital carpet printer, which it launched in 1975 as the Milliken Millitron. This device fires continuous streams of dye from an array of nozzles along the full

4

Digital printing of textiles

print width. Targeted streams of air deflect drops that do not contribute to the image are recycled. Undeflected drops continue on to strike a web of white carpet. Milliken advanced this technology from its early 10 dpi resolution to over 70 dpi. In 1976, Zimmer announced its carpet printer. Today, most printed commercial carpeting is digitally printed.

1.4

Sublimation

In 1973, RPL Supplies Inc., a company now in Saddle Brook, New Jersey, USA, developed a process for transfer printing digitally generated video images to fabric. This company and others developed this process with impact and thermal sublimation dye ribbon for use in customizing and personalizing gifts and promotional products.

1.5

Thermal inkjet and textile printing

In 1977, Canon's Endo discovered the principle of thermal inkjet when placing a flame on the side of a pipette containing liquid that then emitted a drop of that liquid. Soon after, researchers at Hewlett-Packard encountered a similar phenomenon. Canon and HP applied these discoveries to the development of thermal inkjet print heads. Canon called its version `Bubble Jet'. In 1984, HP introduced the first commercial desktop inkjet, the HP Thinkjet. Canon's Bubble Jet office printer followed in 1985 with the introduction of the BJ-80. Canon and HP licensed their inventions to each other and to other manufacturers, including IBM, Siemens, and others. Lexmark took over the IBM license when it purchased IBM's printer division. Canon developed a Bubble Jet textile printer in the mid-1990s that printed fabric up to 1.6 meters in width at a throughput speed of a square meter per minute. The unit did not gain market acceptance due to its high sticker price and limited production capability, but it demonstrated a model for designing, printing, and processing textiles digitally that others have followed. Canon used a material transport system from Ichinose, which later introduced its own twelve-colour inkjet textile printer using HP thermal inkjet print heads that it exhibited at ITMA 1999 in Paris. This device also did not gain market adoption. Ichinose later partnered with DuPont to produce the Artistri 2020 printer using modified Seiko Instruments piezoelectric print heads. This device has won significant market adoption with about 160 printers installed by February 2006. Perfecta, in conjunction with Zund, debuted a textile flatbed printer using Hewlett-Packard thermal inkjet print heads at FESPA 1996 at Lyon, France. Encad offered an inkjet textile printing system using Lexmark thermal inkjet print heads for proofing and short-run production in 1997. Despite a strong marketing effort, market adoption did not match company expectations and Encad eliminated its textile division.

The evolution and progression of digital printing of textiles

5

In 1984, Canon introduced a digital laser copying system, the NP-9030, following its 1979 development of its LBP-10 laser beam printer. Canon continued to develop laser technology, resulting in the release of its CLC1 colour laser copier in 1987. This technology provided a means for producing four-colour process heat transfers for garment, accessory, and promotional product printing.

1.6

Seiren

In the early 1980s, the largest textile printer in Japan, Seiren of Fukui, began developing the possibility of inkjet printing of fabric directly. In 1989, it undertook to build a manufacturing facility for printing fabric digitally. By 1991, Seiren had added inkjet printing to complement its analog operations. It had a few hundred piezo inkjet printing devices constructed for its digital printing operations. It brought its considerable expertise with fabric inks to build a digital textile printing business with an annual gross sales volume in excess of $100 million by 2000. Seiren digitally prints textiles for automotive upholstery, active and swimwear, banners, and apparel. It has also developed the process of digital dyeing. Seiren opened ViscotecsTM stores where customers could order fabrics tailored to their needs. It has also developed information technology (IT) to supply online response to consumer and industry demand for printed and dyed products. Seiren created a model with its ViscotecsTM digital system for agile manufacturing that connects its mass customization production operations directly to the market. It has extended its digital printing of fabric around the world with production facilities in Japan, the United States, China, Thailand, Italy, Belgium, and Brazil.

1.7

Digital grand format and textile printing

During the 1980s, a number of companies developed digital methods to print billboards, building wraps, and large banners. In 1987, Gerber Scientific built large-drum digital grand format printing systems for billboard maker Metro Media Technologies (MMT), which has since become the largest supplier of digitally printed large and grand format graphics worldwide, with digital production locations in North and South America, Europe, Asia, and Australia. MMT prints on textiles in addition to paper and plastic substrates. In 2002, MMT unveiled two of the world's largest inkjet billboard printers with their MegaDrums that measure 63 feet in circumference and 32 feet wide. MMT's competitors in the advertising and billboard markets quickly followed with their thrust into digital printing. In 1989, Vutek introduced its 801 digitally controlled airbrush billboard printer and in 1990 offered its 16-foot wide 1630 billboard printer. Other equipment manufacturers, such as Belcom, Data Mate Company Ltd, LAC Corporation, Matan/Scitex, Nur Macroprinters, and Signtech/Salsa also developed grand format printers for MMT's

6

Digital printing of textiles

competitors. These manufacturers have employed a variety of digital printing technologies including airbrush/valve jet, continuous inkjet, and piezo inkjet. Fabrics and fabric-reinforced vinyl have provided the primary substrate for grand format digital graphics banner and building wrap applications. Geoff McCue filed a patent in 1990 for an inkjet computer to screen mask printer. Gerber Scientific acquired the rights to the McCue patent and produced a device to print a photo mask on photo emulsion coated screens. Stork of the Netherlands and Luescher of Switzerland combined to acquire the patent from Gerber. The inkjet masking systems based on the McCue patent provided the advantages of digital imaging to improve the cost and speed of analog print prepress. In the fall of 1993, a group of engineers led by Patrice Girard formed Embleme that developed a continuous inkjet garment-printing device that used water-based UV-cure inks, Imaje CIJ print heads, and Fusion Systems curing lamps. Embleme established the feasibility of printing garments and operated a shop that offered customers digital printing of customer generated designs on sportswear.

1.8

FESPA 1996

As previously noted, Perfecta introduced a TIJ textile printer at FESPA 1996. Idanit exhibited its high-speed 162 Ad that demonstrated the advantages of large arrays of print heads for production printing, albeit targeting paper and vinyl sheet printing. Around the same time, Matthew Rhome of Bradenton, Florida, applied for a patent for an inkjet printer (on 19 July 1996) and the US Patent office awarded him patent number 6,095,628 on 1 August 2000. The original Rhome printer used thermal inkjet print heads. More recently, Mr Rhome developed a T-shirt printer using Brother PIJ print heads, which the company exhibited at the ISS Exhibition in Atlantic City, New Jersey, during March 2005. In the early 1990s, Sawgrass of Mount Pleasant, South Carolina, USA, won a number of patents for thermal transfer and inkjet sublimation printing. In the late 1990s, it developed an indirect process called Natura for printing garments using electrophotography for use on white and pastel colored cotton and cotton± polyester blend garments. This process produces lighter hand and more vibrant color than resin-based toners. Other manufacturers have developed electrophotographic printers to produce sublimation transfers for receptive garments, accessories, and fabrics.

1.9

FESPA and ITMA 1999

FESPA (for screen and digital printing) in Munich, Germany, and ITMA (for textile production and decoration) in Paris overlapped during June 1999.

The evolution and progression of digital printing of textiles

7

Perfecta exhibited its flatbed textile printer using XAAR XJ-500 print heads at FESPA.

1.10 ITMA 1999 Stork displayed its full line of digital printers at ITMA 1999 in Paris. It exhibited its Amethyst, a seven-color continuous inkjet that Stork developed for use with reactive and acid dyes for printing cellulosic and protein fibers. The Amethyst generated a 254-dpi matrix with gray levels for very high apparent resolution. It could print at a maximum throughput speed of 17.5 m2/hr. Stork also exhibited its Zircon drop-on-demand piezo inkjet based on the Konica eight PIJ print head printer. Stork configured this device to print disperse dyes. It produces 360 dpi and throughput of 6.9 m2/hr. Stork also exhibited its Amber PIJ printer based on the Mimaki seven-color 360±720 dpi TX device. Stork offered the Amber for printing reactive dyes for cellulosic fabric printing. It also exhibited a dual chamber steamer for fixing digitally printed dyes. Stork ran into technical hurdles with the Amethyst, resulting in its discontinuation. It continues to rebrand Mimaki and Konica printers enhanced with Stork software. Encad showed its four-color TIJ 300 dpi textile printer at ITMA. In addition to Stork's versions, Mimaki exhibited its seven-color TX PIJ inkjet and Konica exhibited its eight-color 360 dpi PIJ printer configured for either disperse or reactive dyes. A number of other value-added manufacturers, including DGS of Como, Italy, developed improved material handling and software for the Mimaki printer. As mentioned previously, Ichinose Toshin Kogyo Co. Ltd demonstrated its 12-color 300 dpi TIJ Image Proofer printer outputting 4±12 m2/hr. Perfecta Print AG exhibited its Print Master four-color inkjet. Salsa (formerly Signtech, now part of Nur Macroprinters) exhibited one of its solvent printers as a digital textile banner printer.

1.11 Drupa 2000 At Drupa 2000, DPS and Aprion unveiled their Magic PIJ print head system that will subsequently drive the Reggiani DReAM printer.

1.12 Heimtextil 2001 DuPont introduced the Artistri 3210 at Heimtextil in Germany in January 2001. It used Spectra's water tolerant Nova Q PIJ print heads shooting DuPont Artistri inks on a Vutek 3.2 meter wide media platform. Its print width is 3.05 m. DuPont promoted these devices for the printing of home furnishings. In 2002, DuPont unveiled its Artistri 2020 printer using DuPont modified Seiko Instruments PIJ print heads on an Ichinose sticky belt textile transportation system.

8

Digital printing of textiles

1.13 DPI 2001 In the spring of 2001 at the Atlanta DPI Exhibition, Leggett and Platt (L&P) introduced its Virtu series of UV-cure printers that it built for printing mattress ticking and for other applications.

1.14 ITMA 2003 ITMA 2003 at Birmingham, UK, during October 2003 marked the introduction of a number of breakthrough developments for digital printing on textiles and the beginning of production digital textile printing of yard goods. The key developments involved companies with considerable experience building conventional textile printing equipment, including Ichinose, Reggiani, Robustelli, and Zimmer. · DuPont Ink Jet exhibited two of its Artistri 2020 printers with its Japanese partner and machine builder, Ichinose Toshin Kogyo Co. Ltd. The Artistri 2020 uses 16 Seiko Instruments PIJ print heads arranged with eight heads on each of two gantries. This configuration enables the use of two different ink types or using the same ink on both gantries for greater production speed. The device prints rolls of fabric up to 1.8 m wide at print resolution of 600 dpi at a production rate of about 30 m2/hr. Its roll-to-roll adhesive print blanket system enables printing on woven, and knits, including elastomeric fabrics. The Artistri system and inks can print on nylon, silk, cotton, polyester and blend materials including DuPont Lycra blends. The 2020 stood out from other production inkjet printers because its 2020 devices were printing pigmented and disperse dye ink sets while competitive devices were printing less challenging acid and reactive dyes. DuPont also offered fiber reactive and acid ink sets for its inkjet printers. DuPontTM ArtistriTM inks are available in cyan, magenta, yellow, black, light cyan, light magenta, orange and blue for reactive dye; cyan, magenta, yellow, black, light cyan, light magenta, red and blue for disperse dye and pigment; and cyan, magenta, yellow, black, red, blue, green, orange, fluorescent yellow and fluorescent red for acid dye. The DuPont dyes require post-print fixing typically used for conventional fabric printing. The Artistri system also includes the DuPontTM ArtistriTM Color Control and Management System (CCMS) and RIP software. ITMA marked the emergence from beta testing and the commercial launch of the Artistri 2020. By February 2006, DuPont had placed about 160 Artistri 2020 printers in locations around the world. Sign and banner manufacturers in addition to other product samples and small to medium production fabric printers have acquired this device. · L&P introduced a UV-curable dye for use with its Virtu system. Subsequent versions of the Virtu print line have achieved print speeds of about 200 m2/hr.

The evolution and progression of digital printing of textiles

9

· Mimaki unveiled its TX3 fabric printer with more robust material handling capability than the Mimaki TX2 to tension and print difficult-to-handle fabrics such as elastomeric fabrics. The TX3 provides advanced material handling like that provided for the TX2 on the DUA Graphic Systems Srl (DGS) Cromos textile printer. · DGS of Como, Italy, also exhibited its other software and textile printer offerings at ITMA. DUA Graphic Systems expanded from a supplier of software solutions to the screen engraving industry, to a digital printing software provider and value-added manufacturer of advanced textile handling devices for existing digital printing systems. In addition to the Cromos printer based on the Mimaki TX2, DGS offers the Star G8, based on the Encad Novastar 850, the Colorspan Fabrijet XII, and the Luxor 7 based on the Mimaki TX printer. DGS supplies these systems with its Match Print II software. In February 2005, DGS and DuPont announced a marketing partnership for the sale of the Artistri 2020 and Match Print II software. · Reggiani, in cooperation with Scitex Vision and Ciba, introduced its DReAM textile printer using 42 Aprion 512-nozzle PIJ print heads for this six-color digital textile-printing device. Reggiani reports that the DReAM can print 600 dpi throughput at a rate of 150 m2/hr. Reggiani had installed over a dozen of these devices by May 2005, most of which are located in Italy. · Robustelli, in partnership with Epson, launched its Monna Lisa textile printer. This device is capable of printing very high-resolution images for which Epson print heads are known. The Monna Lisa features half again as many print heads as the Mimaki TX2, along with improved material handling. Robustelli has placed about a dozen of these printers in Italy. · Zimmer exhibited its Chromotex printer using arrays of `Flatjet' piezoelectric stimulated spray nozzles. It also showed its Chromojet carpet mat inkjet printer.

During 2003, Mimaki exhibited its GP 604 garment-printing device with 60 cm (24 inch) print width. Mimaki also offers the GP 1810 inkjet garment printer with 1.8 m (73.2 inch) print width.

1.15 Drupa 2004 One digital printing device at this quadrennial exhibition in Duesseldorf presaged potential developments for digital textile printing. Sun Chemical and Inca Digital introduced a high-speed single pass inkjet printer for decorating corrugated cardboard for the packaging industry. It employed a full-width array of Spectra PIJ print heads. This half-meter wide demonstration model points the way to digital high-speed production printing for a wide range of applications including textile printing.

10

Digital printing of textiles

1.16 SGIA 2004 At SGIA 2004, held from 6 to 9 October 2004 in Minneapolis, Minnesota, USA, Kornit Digital of Moshav Magshimim, Israel, exhibited two robust inkjet garment printers using Spectra AAA print heads, the single platen 930 and double platen 931. Kornit configures its Spectra AAA print heads to shoot 77 picoliter droplets, thus enabling fast printing and greater ink deposit and color saturation. Kornit uses mild solvent-based inks with its 930 and 931 printers, but its versatile print heads can also fire water-based and UV-cure inks. Kornit has configured its printers to yield 450  450 dpi, 540  540 dpi, and 630  630 dpi. Kornit has focused its production efforts on its 931 printer in response to market demand for a high production T-shirt printer. It also increased the frequency of drop generation from its print heads, thereby increasing print throughput to about 300 printed garments per hour for its 931. Also at SGIA 2004, the US Screen Printing Institute (USSPI) introduced its Fast T-JetTM garment printing device based on the Epson 2200, and Jumbo Fast T-JetTM garment printer based on the Epson 7600. These devices print at 360 and 720 dpi resolutions. The Fast T-Jet is the lowest cost inkjet garment printing device available. It can print a double-hit 12 inch by 12 inch print at 360 dpi in about two minutes and the same size image at 720 dpi in about four minutes. USSPI also offers a number of platens for its printers including a cap platen.

1.17 FESPA 2005 Equipment manufacturers introduced a number of new garment and textile printing systems at the FESPA 2005 Conference in Munich, Germany. The US Screen Printing Institute of Tempe, Arizona, USA offered four T-shirt printing devices. USSPI sold about 300 of its Fast T-Jet printers between their introduction at SGIA in October 2004 and the beginning of FESPA on 31 May 2005. USSPI listed this device at US$10,995 and reportedly uses about $0.40 worth of ink per image. It can print about 15±20 12 inch by 12 inch images per hour. USSPI offers its Fast T-Jet LF-2000 Jumbo for US$24,995. This device can print up to 23.5 inches by 36 inches for oversized T-shirts and beach towels. The company exhibited its Fast T-Jet XL-600 Giant with a price tag of $84,995. This eight-color printer reportedly can print 60±120 T-shirt images per hour with ink that costs less than $0.40 per image. USSPI also presented a video depicting its adaptation of the DuPont Artistri printer for T-shirt printing. USSPI claims that this new device can print from 300 to 400 T-shirts per hour. It utilizes carts with 10 mounted platens (five per side) for holding shirts. Operators place the shirts on the platens and remove them away from the printer after printing. USSPI claims ink costs per print range from $0.10 to $0.30 and lists the printer for US$240,000. Kornit Digital Ltd exhibited its Kornit 931 dual platen inkjet T-shirt printer and introduced white ink for printing dark-colored garments. Kornit indicates

The evolution and progression of digital printing of textiles

11

that it has enhanced the production capability of its 931 printer to produce from 320 to 400 T-shirts per hour. Its light solvent ink and fabric coating permits printing a wide range of fabrics. Kornit recently added 360  360 dpi capability to its list of higher print resolution capabilities. This lower resolution enables faster production with image quality that suffices for most T-shirt printing. The 931 with white lists for about ¨200,000. Textile printing and processing equipment manufacturer, MS s.r.l. of Caronno Pertusella, Italy, has introduced its MS-One T-shirt printer. MS reports that it can print an A4-sized image in 30 seconds and an A3-sized image in 60 seconds. The MS-One prints resolutions from 360  360 dpi to 1440  1440 dpi, lists for ¨14,000 and comes with a two-year warranty. MS offers its JetPrint material handling system for use with wide format plotters currently on the market. This permits material transport adjustments for improved image quality. MS includes a blanket-washing module with blanket drying, a print drying module, motor-driven fabric winding and unwinding, a pressing cylinder, an anti-static bar, and a material spreading and uncurling device. MS also offers two inkjet coating and printing devices: the MS-Coat & Print and the MS-Coat & Print SG Plus. These devices pretreat and print simultaneously inline. The MS Coat & Print SG Plus adds fixation and steaming. Colorprint snc of Gallarate, Italy, exhibited its Twister hybrid T-shirt printer that can print images up to 40 cm wide. This carousel device combines a multistation screen printing press and a multi-color piezoelectric inkjet printer. It can screen print a white as a base for the inkjet printing process when decorating colored garments. It can also add screen printed effects, such as glitter, puff, and metallic colors, to enhance and add dimension to digital garment printing. Colorprint's Twister inkjet printing device offers eight pigmented colors ± yellow, magenta, cyan, black, red, dark blue, green, and gray ± for its waterbased device and a maximum resolution of 1440  1440 dpi. Colorprint also offers the Twister as a solvent-based inkjet system. It claims a throughput speed of 100 T-shirts per hour and lists the Twister for ¨60,000. ATP Color of Senago (Milano), Italy, has developed its M-series three-platen T-shirt printer that the company reports can yield up to 50 T-shirts per hour with 600  600 dpi resolution with four-color process inks. It uses Epson printheads and lists for ¨50,000. ATP Color offers its T-series dual gantry Epson-based sticky belt textile printing system capable of printing resolutions up to 1440  1440 dpi. It provides both the T- and F-series printers in versions that can handle media widths from 162 cm to 320 cm. The double gantry T-series lists from ¨120,000 and the single gantry from ¨74,000. The F-series lists from ¨54,000. Algotex s.r.l. of Crevalcore (Bologna), Italy, introduced its Rainbow Jet fourcolor process inkjet printer series. Algotex offers three devices each with XAAR piezo inkjet print heads and solvent based inks for printing textiles, flags, and banners in addition to vinyl and paper. The RB 250 uses eight XJ 128 PIJ print heads and can print 185±370 dpi images on materials as wide as 2.5 meters. At

12

Digital printing of textiles

its lowest resolution the RB 250 can print 27 m2/hr, and at its highest image quality about 15 m2/hr. The RB 325 also uses eight XJ 128 PIJ print heads and can print 185±370 dpi images on materials as wide as 3.2 meters. At its lowest resolution the RB 325 can print 32 m2/hr, and at its highest image quality about 18 m2/hr. The RB 325 TOP uses 12 XJ 126 PIJ print heads and can print 200± 400 dpi images on materials as wide as 3.2 meters. At its lowest resolution the RB 325 TOP can print 42 m2/hr, and at its highest image quality about 25 m2/hr. As mentioned earlier, at ITMA in Paris in 1999, Stork of Boxmeer, the Netherlands, exhibited a number of continuous inkjet printing systems that it had developed and two drop-on-demand piezoelectic printers that it had rebranded and enhanced with Stork software. While Stork has since discontinued its efforts to develop a continuous inkjet short-run production printer, it has refocused its efforts on enhancing digital printing systems that other manufacturers have built through its Stork Digital Imaging BV division. Stork continues its partnership with Lectra of Paris, the world leader in textile and apparel software. Mimaki supplies its TX series of printers to Stork, which has branded them as the sevencolor Amba and eight-color Sapphire and Sapphire II. It also continues to offer the Konica, now Konica-Minolta, PIJ wide-format textile printer under its Zircon brand name for disperse dye printing of polyester and other receptive polymeric fabrics. Stork Digital Imaging BV exhibited its Sapphire II at FESPA 2005. Stork also promoted its Digital Print Asia (DPA) joint venture with the Yeh Group that has its production facility located in Samutsakorn, Thailand. Stork has developed a certification system with DPA called Stork U SeeÕ that guarantees its customers that design samples produced in one of Stork's sampling service offices can be reproduced accurately at its bulk production location in Thailand. Stork has located its sampling service offices at Boxmeer in the Netherlands, New York City and Giridara Kapugoda in Sri Lanka. Stork offers its sampling production up to 50 meters long at its service offices and production over 50 meters long from its Thai production center. This business system combined with inkjet printing offers customers the possibility of shorter print runs, less inventory risk, production to match shorter fashion cycles, unlimited colorways, and no repeat length limitation. Hollanders Printing Systems BV of Eindhoven and Boxmeer, the Netherlands, introduced its ColorBooster textile production inkjet printer. It reports 90% production uptime based on its beta experience. The Hollander value proposition for its customers is to offer the flexible advantages of digital printing and processing in a high image quality system that can operate around the clock with a minimum of operator intervention with low operation cost. It installed 14 of these printers between June 2004 and May 2005 as beta tests and reports customer satisfaction running production operations with the ColorBooster. It employs 16 piezo drop-on-demand print heads with 180 nozzles each to produce 360  360 dpi to 2880  2880 dpi images with eight print colors. Hollanders

The evolution and progression of digital printing of textiles

13

claims the ColorBooster can print at 80 m2/hr printing four-color 360  360 dpi prints (2  4 colors) at 25±50% coverage and 39 m2/hr at 100% coverage. It claims the ColorBooster can print at 50 m2/hr eight-color 720  360 dpi at 25± 50% coverage and 22 m2/hr at 100% coverage. It prints fabrics up to 2.3 m wide with images up to 2.23 m wide. Hollanders ColorBooster employs an inline print head arrangement that maintains print order during bidirectional printhead scanning. This eliminates certain types of banding and contributes to color consistency. Its open ink system carries a five-liter reservoir and ink buffer for each of its eight print colors. It also includes an anti-sedimentation system that continuously circulates ink to keep colorants from settling out of solution, and users can replenish ink without interrupting operation. The Hollanders inkjet print heads can shoot pigmented inks, acid, reactive, disperse and disperse±sublimation transfer dyes. Hollanders Printing Systems indicates that its system with a combination of techniques can achieve a high level of print-through penetration that manufacturers of flags, banners, and silk scarves require. The ColorBooster system includes color management that Hollanders says can match colors precisely. Hollanders Printing Systems offers an open ink system with the end user selecting its ink supplier. The ColorBooster also includes a newly developed material transport system that can adjust cloth tension for each substrate and maintain tension during printing. The ColorBooster automatically step-corrects to compensate for material thickness. It includes a computer climate controlled system for the printing process. The company claims the ColorBooster can print as many as 80,000 m2 of fabric per year. The ColorBooster lists for ¨145,000. d.gen International, Inc., of Seoul, Korea, offers textile inkjet printing models based on Roland Epson-based printing systems. These include the Artrix d.gen 740 TX/Be with a maximum print width of 1.879 m and the d.gen 1000 TX/Be with a maximum print width of 2.6 m. Both use 12 Epson PIJ print heads that can generate textile prints from 450  360 dpi two-pass six-color prints at 28 m2/hr to 1440  1440 dpi 16-pass prints at 3.5 m2/hr. These systems employ a one-liter continuous ink feeding system for each color. Textile printers can use reactive, acid, or disperse dye inks or pigment inks with this print system. d.gen offers the 740 TX/C with a cylinder material handling system for thin fabrics such as silk chiffon for ¨43,000. It also offers the Teleios for direct disperse±sublimation dye printing built on the same printer bases as the d.gen 740 TX/Be and 1000 TX/Be. It offers disperse dye in cyan, magenta, yellow, black, light cyan, light magenta, orange, green, gray, and deeper black. The Teleios d.gen 1377TX/74 lists for ¨50,000 while the d.gen 1377TX/100 lists for ¨90,000. d.gen unveiled its 7474 TX Heracle dual gantry inkjet printing systems. It employs a sticky belt and can print a maximum width of 1.879 m. It carries 24 print heads, 12 on each gantry, and can print at a maximum resolution of 1440

14

Digital printing of textiles

dpi. This device was still in beta testing but d.gen reports that it will be available by the end of 2005. It prints reactive, acid, and disperse dye or pigment ink. In four-pass, 360 dpi mode, the Heracle will print 36.5 m2/hr, and in the four-pass 720 dpi mode it will print at a rate of 21.4 m2/hr. d.gen has yet to announce a price for the Heracle. This sticky belt device will likely compete with the sticky belt machine from DuPont. Kimoto Ltd of Rumlang, Switzerland, introduced four inkjet printers, which it calls the Philyasystem. At the core of each of these devices is a Roland printer with Epson print head technology producing resolution up to 1440 dpi. Kimoto designed one of these devices, the TBS-1600, with an adhesive belt transport system for controlling textile during printing. Kimoto offers the printer for use with four- or six-color water-based ink sets. It lists for ¨82,500. Kimoto reports having one of its Philyasystems beta-testing in Italy.

1.18 Other key elements Many disciplines and competencies contribute to producing digital textile printing. In addition to print head design and manufacture, material handling engineering, and ink chemistry, are textile manufacture and pre-treatment, postprint finishing, design, raster image processing (RIP), and color management software. Monti Antonio S.p.A. of Thiene, Italy, has developed vacuum heat presses that can produce print-through with digital printed images for flags, banners, and scarves. Other manufacturers of post-processing equipment are also expanding the capabilities and applications for digital textile printing. Color matching and management software and equipment are helping digital textile designers and printers distribute and print at multiple locations around the world with colors that match exactly. Printer±ink±media profiling and climatecontrolled environments are enabling digital textile printers to reproduce print images repeatedly. Digital textile printing is entering an era of greatly improved reliability that replaces personal skill with scientific and numeric precision.

1.19 Conclusion The textile printing and equipment manufacturing industry in Italy has provided significant leadership in applying digital printing for textile applications. Key equipment developments at Italian manufacturing companies, such as DGS, Reggiani, Robustelli, MS, Algotex, ATP Color, Colorprint snc, and Monti Antonio among others, underscore the major Italian contribution to the adoption of digital printing for textile printing. Japan has contributed much of the key print head and printer technology that has driven textile printing. Japanese-manufactured Epson, Sharp, Seiko Instruments, and Konica Minolta print heads print most of the fabric that the

The evolution and progression of digital printing of textiles

15

world prints digitally. US and UK manufacturers have also contributed with primary technology and business development. Textile print producers India, China, and Turkey are also developing digital solutions. The pace of development has begun to advance. Inkjet textile printing is growing while growth in analog textile printing remains stagnant. Digital printing is beginning to account for an increasing share of textile print production. As the length of textile print runs decreases, and the demand for short-run production and just-in-time delivery increases, digital printing is providing the cost-effective solutions. As digital print technologies improve, offering faster production and larger cost-effective print runs, digital printing will grow to become the technology that provides the majority of the world's printed textiles.

1.20 References 1. Susan Meller and Joost Elffers, Textile Designs: Two Hundred Years of European and American Patterns for Printed Fabrics Organized by Motif, Style, Colour, Layout, and Period, Abrams, 2002. 2. http://www.silk-road.com/artl/printing.shtml 3. http://www.india-crafts.com/textile/printing_tradition/block_printed/index.html 4. US Patent Number 1,941,001, applied for 19 January 1929 and issued 26 December 1933.

2 A designer's perspective ± digital versus traditional L N I C O L L , Consultant, Italy

2.1

Introduction

A very important transformation is taking place in the textile industry: the digital revolution. That sounds very twenty-first century, but it started 30 years ago, and although it is essential to understand the developments and progress made in that time regarding hardware and software, I believe that we must now consider the most important part of the equation ± the human element. Visiting the major European textile museums such as those at Mulhouse, Macclesfield, Como and Lyon, one can see the progress made through the centuries and the notable influence of new machinery and new techniques on the fabrics and designs produced. So, for example, we see the progression from the primitive Coptic looms to the Jacquard loom (based on the reading of perforated cards, the forerunner of the binary system used by today's computers), and from the first simple wood block printing to precise metal engraving, then the photographic techniques used to engrave flatbed screens, rotary cylinders and copper rollers. Now we're only at the beginning of the digital age and this new technology will be just as important, if not more so, as the arrival of the Jacquard loom so long ago. It will change the face of textile production and distribution, and it will empower the designer in a way never seen before. All developments, whether mechanical, technical or chemical, have always brought about important changes in the characteristics of the textiles produced. Not only have they improved the quality of the product, visually and in many other ways, but above all they have offered ever greater levels of creativity. So clearly progress cannot be completely attributed just to the tools available ± it has always been the people involved in the process who have used the new developments with creativity, enthusiasm and perception, thus establishing the most important centres of textile production. Como in Northern Italy is a good example. For the last 50 years it has been the most important centre of textile production excellence. Post-war Como took over from the French textile industry in Lyon, starting with beautiful silk

A designer's perspective ± digital versus traditional

17

Jacquard fabrics, then progressing to flatbed screen and rotary printing, to be recognised as the best in the world. So how did they do it? In the whole textile sector there in the mid-1970s, everyone involved used the new technologies with creativity, enthusiasm and specialist knowledge ± even inventing newer things in order to pursue the excellence the world expected of them. That's why, although it is essential to consider the progress made with hardware (or machinery) and software (or working technique), the most important element of all must be considered ± the human element. It was this and the character, the dedication and the specialist knowledge of everyone involved ± the stylists, the designers, the colour separators, the sales people, the creative entrepreneur ± which made the Italian difference. But now the world is rapidly changing and it is truly global. The same technology is available to everyone, from Korea to China, from Hungary to Turkey. The production centres are moving from the traditional areas, so how will Italy survive? Well, it certainly will continue to buy British creative textile designs as it always has. It's going to have to adapt to a new productive reality, and that could be a big problem for the industry in Italy, because the new digital technologies mean that textile designers no longer have to rely on large-scale industrial processes in order to reach the market ± any market, from one-off pieces, limited edition bespoke runs, to hundreds of thousands of metres if necessary. Now inkjet-printed textiles and computer-aided design are capable of production speed. All that is needed now is to bring in creativity, enthusiasm and specialist knowledge, and then things will change. New markets will open up, distribution methods will change. It's happening in many different industries thanks to the digital revolution. There are huge implications in what this digital technology has opened up to everyone in the creative field, from the artist to the craftsperson and the designer. But it is necessary to look at the 30-year history of computer-aided design and how it developed in order to avoid the mistakes of the past. The first CAD system arrived in 1970. Up till then the method of production was similar to that for woven Jacquard. Designs were translated to graph paper, a very complex technical task, then cards were punched to control the machines. The difference computers made was notable. Designers had complete control ± they could sample the fabric at the touch of a button, change it, keep it, produce 1 metre or 100,000 whenever needed. CAD for printed textiles arrived in 1976, but it was more for computer-aided manufacturing than computer-aided design. Films for engraving screens were produced on laser plotters permitting precise register and the most difficult challenge of all for hand-drawn colour separations ± a straight line.

2.2

What difference does digital make?

The best way to explain the difference that digital will make is to show how a textile collection is produced using firstly traditional methods, then digital technology.

18

Digital printing of textiles

Any textile manufacturer producing a twice-yearly collection of designs works in the same way as a freelance designer. In order to sell designs for eventual production, they must show samples of their work at the major world trade fairs, where mainly clothing manufacturers choose the styles they want the manufacturer to produce for them. The success rate for samples being turned into mass production is about 15± 20%. This means, of course, that for the manufacturer's factory to have a full order book for their production schedule, they must present hundreds of samples twice yearly.

2.3

How is this done using traditional methods?

A sample length can vary from two metres to five metres depending on the type of design. Many screen printers have very large print shops dedicated to sample production. To produce a sample using traditional screen engraving methods involved tooling up to full production standards. That is to say, the screens used to make the successful samples could immediately be used in the mass production process. The fully production-ready screens of unsold designs would be destroyed and the costs written off as a part of the sales process. Sample preparation is a long, complex and expensive process. The screen printing process starts with a high quality piece of artwork. This art is then separated out into the composite colours and a film positive is produced. This film is then placed onto a screen coated with emulsion. The screen is then exposed to UV light in an exposure unit. After the screen has been exposed for the proper amount of time, the screen is `washed out' with a pressure washer. This removes any emulsion that was not in contact with the UV light (the positive area). The screen is placed in a drying cabinet, and once dry taken to the press for registration. Ink is `squeegeed' through the open areas of the screen onto the fabric below. The printed fabric is then put through the conveyor oven and cured to a temperature of 350ëC. A collection starts with a group of design stylists deciding which designs are most likely to sell, according to trend predictions and their experience of the current fashion marketplace. Once this is decided they will use the textile design artist available within their organisations or buy what they need from external studios or freelance textile designers. This is the first phase, the high quality piece of artwork. The next phase is to separate this artwork into its composite colours, which is done by skilled colour separation artists. Colour separation using traditional methods is a highly skilled and very difficult job, and as time progresses it is becoming a problem finding people with the right skills to do it. Colour separation is the first reinterpretation of the design, which is divided up into its composite colours in order to engrave a screen for each of these colours. A comparison could be made to the spot colours used in computer

A designer's perspective ± digital versus traditional

19

graphics and paper printing. As opposed to CMYK four-colour printing, this permits each screen colour to be changed independently, which is essential in a textile collection in order to offer various colour variations or choices for each sample offered. The design will be put into repeat, the colours separated and at this point the most complex part of the process takes place, which is ensuring a precise register between each screen. This is essential because as each colour is printed there must be a slight overlap between each one as no fabric background should show. Making this work using traditional hand techniques such as brushes and pens with opaque inks is an extremely difficult and lengthy process and results can be inconsistent, varying between different studios and individual artists, which makes consistent quality at the production fabric printing phase difficult to achieve. Of course, it can all be done successfully, as, for example, in the Como district in Northern Italy, where an expert artisan tradition and longestablished infrastructure ensure quality recognised worldwide. Preparing well-made colour separation films comes at a cost, both financially, in the time it takes to produce them, and in the logistics involved. It takes a minimum of 10 days to get from artwork to colour-separated film ready for screen engraving. With hundreds of designs being made ready for sampling in time for a deadline such as the biannual trade fair, all happening at the same time with all the complex processes involved in the fashion business ± priorities, changing of ideas at the last minute ± it's a logistics nightmare. So after a minimum of 10 days the screen is ready to be engraved. The following is a summary of technical instructions provided by various screen engravers. Use polyester or other suitable synthetic fabric or screen material. This is one of the most exciting methods of screen printing because it offers the widest range of possibilities. It makes possible the printing of fine line drawings, large consistent colour areas, various hand and commercial lettering techniques, as well as photographic half-tone positives. All methods of photographic screen printing require three things: (1) a screen prepared with a light-sensitive coating, (2) a film positive, or equal, and (3) a light source that will enable you to transfer the opaque images on your positive to the light-sensitive screen you have prepared.

Step A Mixing the photo emulsion The photo emulsion is made by mixing two different liquids. Follow the mixing instructions given on both containers. Store the sensitised emulsion in a cool and dark place. Shelf life for the sensitised emulsion is four weeks at 90ëF, eight weeks at 70ëF, and four months when refrigerated.

20

Digital printing of textiles

Step B Coating the screen Coat the screen by first pouring a bead of the solution on one end of the bottom side of the screen. Spread it evenly and thinly with the squeegee or the plastic spreader. Use more solution where necessary. Pour a bead of the solution on one end of the inside of the screen and spread it evenly with the squeegee or the plastic spreader. Work to achieve an even continuous coating on both sides of the screen fabric. Perform the final spreading on the inside of the screen. Return any excess solution to your mixing container.

Step C Drying the coated screen In an area away from light and heat, set the screen to dry horizontally, bottom side down. This will provide the most even, flat `film' on the underside of the screen. Most commercial engravers use specialised ovens specifically developed for this process. Allow the screen to dry thoroughly. If more than 300 prints are to be run, it is best to apply a second coating of the sensitised photo emulsion to the bottom of the screen after the first coat is dry. Remember, work for a smooth, even thin coating. Repeat the drying process away from heat and light. Once the sensitised screen is dry, it must remain in a darkened area until it is ready to be exposed. A fan in the dark area will greatly speed up the drying of the emulsion on the screen.

Step D Preparing a positive With dichromate systems, the maximum allowable time between application of the sensitised emulsion to the screen and the exposure is six hours at room temperature. With diazo systems, the maximum allowable time is eight weeks at room temperature. Before removing the sensitised screen from the dark drying area, make sure everything you need to print with is on hand. At this point the colour-separated positive is attached to the photo emulsion coated screen. Register marks have been added to each positive to ensure precise register between each printed colour. The exposure lamp overhead emits controlled light intensities for specifically defined amounts of time. These intensities and times vary greatly and depend on design type and different suppliers' products. Successful engraving depends on highly skilled and experienced technicians. After exposure, remove the positive and take the screen to the wash-out process.

Step E Washing out Apply a forceful spray of water (at body temperature) to both sides of the screen. Do not use hot water. Concentrate this spray on the light images on the top side

A designer's perspective ± digital versus traditional

21

of the screen. After a few minutes, these areas will become `open'. Continue spraying until all unwanted emulsion is gone. Once you have completely washed the screen, let it dry thoroughly in a level flat position. Hold the dry frame to the light and check for pin-holes. These can be covered with screen filler or pieces of masking tape stuck to the bottom of the screen. If screen filler is used, let the screen dry again. Photo emulsion should not be left on the screen indefinitely unless a permanent stencil is wanted. It should be washed out as soon as the run is completed. The screens are now ready for printing samples, and if the samples sell, the design can be used immediately in production. Here again it can be seen that screen engraving requires highly skilled technicians and, although the turnaround times from positive colour-separated film can be as short as one working day, the fact that hundreds of designs, each with several screens, are being processed at the same time for the same deadline can cause complex logistics problems. Before samples can be printed on fabric, two full prints in reference colours in their predetermined order will be made on paper to check the accuracy of the design produced and to act as a master copy. One copy is sent to the production archive while the other copy will move between the various departments involved in sampling and production. The colourists will use this master to prepare colour variations in fashion colours suitable for the current collection. These colours will then be given their recipe mix for each individual and unique colour, as there are no standard colour ranges in stock. Here again this is the master recipe which can later be mixed in larger quantities for the eventual production process. The next step is actually printing on fabric. As mentioned, most medium to large producers have sampling departments the size of small factories. The printers are highly skilled artisans, printing by hand onto a large variety of fabrics using many different dyestuffs and techniques. The fabric sample is now submitted to the stylist for approval of colour, fabric handle and appearance. If the sample is approved it is ready for presentation at the trade fairs; if not it has to go through colouration, colour mixing and sample printing. The whole process is prolonged and expensive, can be very inflexible, and requires highly skilled and unfortunately, as time goes on, less available personnel. It's difficult to estimate times and costings involved in such a varied and complex industry, but digital makes the difference. Preparation techniques changed with the arrival of the first computers to make the colour-separated positives for engraving, but a more radical change could be seen in the precision and quality of the laser-engraved Mylar film. It enabled consistent and controllable overlap between screens and unforeseen and precise register. It really caused a quantum leap in quality and an even bigger one in creative output, as the CAD was let loose on the design community when

22

Digital printing of textiles

they saw they could go beyond what was possible by hand preparation, using the software available on the systems. Though designers were desperate for immediate hard copy of their designs, in order to see results they still had to go through the whole screen engraving and printing process. The quality of the colour-separated positive had changed but the whole process was still long, expensive, unwieldy and complex. Preparation of the films by computer in fact added to the complexity. The computercontrolled plotters were very slow, and just two plotters could be working on hundreds of separations, which caused a bottleneck compared to the hundreds of separators and engravers working by hand. It took many years to change the system radically and that happened when inkjet technology for fabric printing arrived around 1995. CAD systems and software and specialised designers were already in place; all that was missing was this long-awaited technology. In the Northern Italian area of Como, manufacturers quickly saw the advantages and savings that textile inkjet printing offered them over the traditional system described previously. In a two-year period in the late 1990s over 100 printers were installed there, while in the same period in the UK only around two or three were installed.

2.4

How do they compare?

Direct inkjet sampling cut out long, expensive and complex industrial processes. Full preparation of colour-separated positives or screens, and printing of fabrics in huge print rooms, was no longer necessary. Register between screens and colour control were no longer a problem. Inkjet sampling gave the designer direct control over the appearance of the design on fabric, without reinterpretation by separation artists. Last-minute decisions could be made as little as a few hours before presentation, or even during presentation, and there was no waste of screens prepared for unsold designs. A few problems remain. Colorant type is still limited compared to screen printing. Inkjet printers cannot use flocks, glitters or devore techniques yet. Fabric variety is still quite limited. Final production and markets must be carefully considered. Clothing manufacturers need colour variations: CMYK four-colour processing will not do. Unfortunately many salespeople are commissioning many more samples, because it's cheap, fast and easy and improves their success rate enormously. However, there is one very big problem emerging here, since when the successful samples arrive at the production part of the process, in 99.99% of cases they will still be produced on traditional printing machinery. Unlike in the traditional process, they are not ready for immediate production, as they have to go through the screen engraving process. Since ITMA 2003, there has been production machinery capable of 100 linear metres per hour, but integration is still a way off.

A designer's perspective ± digital versus traditional

23

Bad computer-aided design for inkjet-printed samples can cause massive problems at the factory production stage. Often the computer-produced designs are impossible to prepare for production machinery, or the data files given to the computer-aided manufacture separation artists are so complex and mixed up that they have to rescan the printed sample in order to produce the separated colour film properly. It's so tempting for salespeople and stylists to rush out thousands of samples made available by this new, fast, flexible technology, but caution must be taken. Ultimately the responsibility lies with the designer. Always start with and remember the final means of production, then design accordingly. The argument isn't necessarily digital versus traditional: it should be digital with traditional ± at least for the near future.

2.5

How can the designer use these twinned technologies?

It is necessary to understand the potential in these systems. Reproduction of a design is fine but you need to go beyond the normal hand-drawn images if using the programs on the system. The Ratti group set up a centre for study and research, which aided in the group's success; the keywords being innovation, sophisticated product, unique product and lead, never follow. Although these early CAD programs were not at the levels of the first knit programs, they were able to produce paper colour proofs which could be transferred to production screens when successful. Although it all worked very well it wasn't easy, because in the 1970s people were working with relatively primitive and very expensive computers compared to now. Only big industry could afford them. The computers ran at a fiftieth (or less) of the speed of any PC available on the high street today, and the RAM was only 32 MB with a hard disk of 200 MB. It was a unique studio experience and not available to many. But things have changed ± finally with inkjet technology, a designer can produce fabric at the touch of a button, just like in 1970.

2.6

Freedom

Resistance from the established designers was quite high at first. Some saw computers as restricting creativity, although the screen engravers embraced it and in fact remain world leaders due to their early involvement with the technology. But the results produced were what changed things. It didn't take words to convince them ± just innovative design and quality. It's the same position today; it's time to demystify the myths surrounding digital technology. It doesn't limit creativity, it doesn't make the process less human ± in fact it's exactly the opposite. Finally there is the freedom. New generations don't have as many phobias about these technologies and techniques. They understand that they don't have to lose time photocopying,

24

Digital printing of textiles

cutting, pasting, and struggling with repeats. They have more time for the truly creative part of a designer's work. Now designers are in direct control. With today's technology there is no longer a barrier between creativity and the production of the design on fabric or any surface. This new technology allows designers to take control of the end result, to prevent it from being changed down the production line. All the industrial preparation processes can be eliminated; it will be faster, more competitive, more creative. Designers such as garment and interior designers can also work with customers in a much more efficient and integrated way. Digital methods give the advantages of speed, communication of ideas, last-minute ideas, and beating the competition on the catwalk with a unique product. The process can start with words, then a list of ideas, then the designer can develop them on the computer, trying new variations, secure in the knowledge that the final result on the market, whether 1 metre or 100,000, will be exactly as originally wanted ± and instantly. Designers and customers have total control over the final product.

2.7

Thinking about creativity

Working by hand can make many beautiful and innovative things, but with a computer creativity can go beyond that, even to places where unimaginable ideas and images become available. Before this technology the human touch was stifled. Layers and layers of industrial process have now vanished ± it is no longer necessary to rely on a far distant factory, and the full range of possibilities is now directly in the designer's hands. So where does that leave designers? They have the power of the new technology ± will they use it in a creative way or will they stay with old ideas and attitudes? The new technology has to be embraced, as the competition doesn't hesitate. For example, in Korea they really are up to speed on this technology; they recognise the potential. A design can be completed on the other side of the world and 36 hours later a disc with the design on it can be in a factory in South Korea, slotted into an engraving system, and in production a day later. That's serious mass production, from limited edition boutiques to high streets all over the world if required. It is no longer necessary to rely on cumbersome industrial processes ± it's time for mass customisation.

2.8

Resistance

Unfortunately there is still resistance, but it will surely change. More people need to adopt the technology and learn how to use it to increase efficiency and avoid one person doing the work and four others telling them where to cut and what to move. By getting hands-on experience, people have more control; the programs are getting easier and easier to work with. New professional and craft roles are being defined. More and more unique creative skills are being

A designer's perspective ± digital versus traditional

25

transferred onto this new technology. Now the technology can be used to make whatever we want to ± it is becoming completely transparent and global.

2.9

Transparency

This technology allows designers so many opportunities. It's so important that skilled designers transfer their skills to the excellence that the technology allows. The final means of production must always be taken into account although inkjet printers are now up to production speed, the majority of products will still be produced on a large scale by low-technology methods such as screen printing. It is also necessary to take into account how the market works at present. The market expects a collection of fabrics to have prints with colourways, groups or families of colours, clearly separated backgrounds, etc. This sums up the present situation, with inkjet printers being used mostly for design sampling as they save expensive screen-making time and costs, but this can cause problems. A design made on the computer can create more problems than it solves. It is necessary to start from zero and think like a textile designer, deciding from the start that the design will have, for example, three greens, four reds and a background, and that it should be adaptable to different colourways. The future will be very different. The reason why the Italian textile industry was so successful over the last 50 years was that it always used the latest technology available in the best creative way. That's why it's so important that the creative process advances alongside, if not ahead of, the technology. This way, new markets will evolve like, for example, mass customisation ± limited editions produced in small batches in mass quantities, with the possibility of changing the design every 10 metres, for example ± all thanks to computer-aided design and inkjet technology. History tells us that the connection between art and industry is very important. Andy Warhol didn't have the technology of today, but with what he had ± screen printing ± he started to make limited editions, maybe even invented the concept of mass customisation. Or maybe it started with the lithographers. An artist could now create a design on a computer, print 10 copies and sign each copy of the print to make it unique, with the classic 1/10 numbered inscription of the lithographer. Once the print run is finished, the original file will be destroyed ± that's added value, for sure. We can also now produce limited edition garments as we don't have to engrave screens or keep a stock or a warehouse full, or manufacture thousands of garments to sell only a percentage of them. The new technologies allow true production on demand. There are also new and exciting developments that link the creation of a textile design and the pattern, cut and design of the garment. This opens up the possibility of textile designs that follow, or decide, the form of the garment. Not only the most advanced European design research projects, but major players like Levi's are now using body scanners to produce individual

26

Digital printing of textiles

patterns for garments based on each person's measurements. Link this to CAD textile design and inkjet printing and things will really change. New opportunities appear every month.

2.10 The new market Obviously it is necessary to take advantage of all that this technology offers, but there are many dangers. Take the Internet, for instance ± designs can be shown on the web and the customer can order 1 to 100,000 metres of printed fabric. It is necessary to scramble it so it can't be downloaded directly, but these ciphers can be overcome. So rather than relying on this kind of protection, it's more important that the designer makes sure that a copy obtained in this way will only ever be an obviously bad copy, which means working like a textile designer ± not a graphic designer. Work methods will surely change ± interaction between garment designers, even graphic designers, is now completely possible; software graphics programs cost nothing; powerful systems are getting cheaper and cheaper. So there's nothing stopping anyone, working from anywhere, from producing high-quality unique work, available to the mass market or the limited edition market. Or to go even further, why not supply the production system with a disc ready to plug in with each colour separated, in repeat ± ready to go? And finally, in the fashion business, when most people hear the words digital design, the first thing that pops into their minds is designs made by robots. Digital design is not just this ± how many such designs in a collection today would you sell? Many people do not even realise that designs have been produced digitally, thanks to the human touch.

Part I Printer/print head

3 Ink jet printing technology (CIJ/DOD) E M A R I A N O F R E I R E , DuPont Ink Jet, USA

3.1

Introduction

Ink jet is a technology that enables the delivery of liquid ink to a medium whereby only the ink drops make contact with the medium. It is therefore a nonimpact printing method. Much of the fundamental theory behind the ink jet technology was developed at the end of the nineteenth century by Lord Rayleigh (Rayleigh, 1878) but the development of the technology itself did not start until the late 1950s and 1960s. Ink jet has three basic components, all of which need to work well in order to produce an acceptable output. These pieces are the print head, the ink, and the medium. The objective of this chapter is to review the ink jet state of the art from the print head standpoint. We start in the next section by describing the ink jet technology classes and explaining the advantages and disadvantages of two of the most prevalent technologies. In Section 3.3 we discuss aspects to consider when selecting a print head technology. Section 3.4 provides a list of the companies that presently have active print head development programs. Finally, in Section 3.5, we speculate on what developments one might expect to see in the near future.

3.2

Ink jet technologies

Ink jet technologies are typically classified in two large classes: Continuous Ink Jet (CIJ) and Drop-on-Demand Ink Jet (DOD). In CIJ, ink is squirted through nozzles at a constant speed by applying a constant pressure. The jet of ink is naturally unstable and breaks up into droplets shortly after leaving the nozzle. The drops are left to go to the medium or deflected to a gutter for recirculation depending on the image being printed. The deflection is usually achieved by electrically charging the drops and applying an electric field to control the trajectory. The name `continuous' originates in the fact that drops are ejected at all times. In DOD ink jet, drops are ejected only when needed to form the image. The two main drop ejector mechanisms used to generate drops are piezoelectric ink

30

Digital printing of textiles

3.1 Main classes of ink jet technologies.

jet (PIJ) and thermal ink jet (TIJ). In PIJ, the volume of an ink chamber inside the nozzle is quickly reduced by means of a piezoelectric actuator, which squeezes the ink droplet out of the nozzle. In TIJ, an electrical heater located inside each nozzle is used to raise the temperature of the ink to the point of bubble nucleation. The explosive expansion of the vapor bubble propels the ink outside the nozzle. Other less common drop generator technologies disclosed in the patent literature will be briefly described below but most of our focus will be on CIJ, PIJ and TIJ. Figure 3.1 shows the classification of the various ink jet technologies.

3.2.1 Continuous ink jet In CIJ the jet of ink generated by each nozzle breaks up into droplets shortly after exiting the nozzle. Without any other intervention, the breakup would occur randomly and would result in droplets of variable sizes. This is usually corrected by providing a periodic excitation to the nozzle in the time domain that translates into a spatial perturbation in the jet of fluid. The combination of the jet velocity and frequency of the excitation determines the droplet size, which can be controlled to very high accuracy. In the traditional CIJ approach, a piezoelectric transducer is coupled to the print head to provide the periodic excitation. The oscillations are therefore mechanical in nature. After leaving the nozzle, the drops are electrically charged by an amount that depends on the image to be printed. The drops then pass

Ink jet printing technology (CIJ/DOD)

31

3.2 (left) Continuous ink jet ± binary deflection. 3.3 (middle) Continuous ink jet ± multiple deflection. 3.4 (right) Continuous ink jet ± Hertz method.

through an electric field to cause them to deflect. There are two ways of deflecting the drops in piezoelectric-driven CIJ. In the binary deflection method the droplets are directed either to a single pixel location in the medium or to the recirculating gutter. In the multiple-deflection method the deflection is variable so the drops can address several pixels. These two concepts are illustrated in Figs 3.2 and 3.3. There is a variant of CIJ called the Hertz method after Dr Carl H. Hertz of Sweden who invented it (Hertz et al., 1986). In the Hertz method the amount of ink deposited per pixel is variable. This is achieved by generating very small drops (of the order of 3 pL) at speeds of about 40 m/s with excitation frequencies of over 1 MHz (see Fig. 3.4). The drops not intended to reach the medium are charged and deflected to a gutter. The printing drops are given a smaller charge to prevent them from merging in flight. Iris Graphics has successfully commercialized this technology on digital color proofers. The company is now part of Kodak. Kodak has recently disclosed a CIJ system in which thermal pulses are used to uniformly break up the jet of ink (Hawkins, 2003; Anagnostopoulos et al.,

32

Digital printing of textiles

3.5 Continuous ink jet ± thermal excitation.

2001). In this version of the technology, each nozzle has an annular electrical heater that is pulsed at a certain frequency. The heat generated raises the temperature of the ink jet in the vicinity of the nozzle and locally lowers the viscosity of the ink. Because the heating pulse is periodic in time and the jet velocity is constant, the resulting jet breaks up into equally sized drops in a reproducible way. This type of drop ejector is illustrated in Fig. 3.5. The thermal CIJ technology lends itself to several deflection mechanisms. One could certainly charge the drops and use the standard electric field-driven method to achieve the deflection. Another option disclosed by Kodak is air deflection in combination with modulation of the drop size by the heating pulse so that when no drops are needed their size is reduced and an air current deflects them to a gutter. A third approach is based on dividing the annular heater that controls the drop breakup into two independently controlled heaters placed on diametrically opposite sides of the nozzle. By applying different energy to each heater, the direction of the jet can be steered at will. Because of the complexities associated with conventional CIJ (charge and deflection, ink recirculation, pressurization) such print heads tend to be costly. On the other hand, because the nozzles are actively refilled by the positive pressure operation, the operating frequencies of these devices are typically at least an order of magnitude higher than in DOD systems. For these reasons, CIJ systems are generally used in industrial applications.

3.2.2 Drop-on-demand piezoelectric ink jet In piezoelectric ink jet, the mechanism used to generate the droplets is a piezoelectric element, typically made of lead zirconate titanate (PZT). Depending on the architecture of the head, the piezoelectric transducer could be

Ink jet printing technology (CIJ/DOD)

33

attached to a membrane that forms an ink chamber wall or could actually constitute the chamber itself. In either case, when a voltage is applied to the electrodes of the piezoelectric element the volume of the chamber is typically reduced, which results in a droplet of ink being squirted out of the nozzle. PIJ print heads are sometimes subdivided in different classes according to the geometry of the drop ejector and/or how the piezoelectric element operates. The classes, shown in Fig. 3.1, are `shear mode', `bend mode', `push mode', `squeeze mode', `nozzle excitation', and `porous layer feed', A common configuration used, for example, by Spectra (Fishbeck et al., 1989) is the shear mode. In shear mode ink jet, the electric field is perpendicular to the poling direction of the piezoelectric material (see Fig. 3.6). The application of this field produces a shear motion in the piezoelectric material that makes the membrane move like an oil can. Xaar's drop ejectors (Temple et

3.6 Drop-on-demand piezoelectric ink jet ± shear mode.

34

Digital printing of textiles

3.7 Drop-on-demand piezoelectric ink jet ± bend mode.

al., 1995) also operate in the shear mode (i.e., the electric field is perpendicular to the poling direction), but in their version the firing chambers are grooves diced into the piezoelectric material and the electrodes are placed inside the chambers. For that reason, this configuration is also referred to as `shared wall'. Unlike Spectra's version of the technology, the walls approaching each other cause the volume reduction in the chambers during firing. In bend mode piezoelectric ink jet, the electric field and poling directions are parallel. The piezoelectric material is placed on the membrane and the membrane moves like an oil can. This configuration is illustrated in Fig. 3.7. Print heads made by companies such as PicoJet and Xerox (Tektronix) as well as some of Epson's print heads operate in this mode. In the push mode piezoelectric ink jet used by Trident, the electric field and polarization vectors are also parallel but the membrane is placed in the expanding direction of the piezoelectric material (see Fig. 3.8). In the squeeze mode the drop ejector is a hollow tube of piezoelectric material. Upon the application of an electric field, the inside volume of the tube (firing chamber) decreases its radius and ejects the ink in the direction of its axis (Fig. 3.9). A novel way of configuring a piezoelectric drop ejector has been disclosed by The Technology Partnership (Arnott et al., 2002) and, independently, by HP (Haluzak et al., 2004). In this configuration the piezoelectric elements are mounted on the nozzle plate (see Fig. 3.10). The simplicity of the fluid path achieved in this concept should result in significant cost advantages as well as robustness against the presence of air bubbles in the ink path. To our knowledge, this concept has not yet been commercialized. In the piezoelectric print head designed by Aprion, the actuator chamber is made out of a porous metal layer (e.g., sintered stainless steel) and the ink is fed to the chamber through this porous material. The concept is illustrated in Fig. 3.11.

Ink jet printing technology (CIJ/DOD)

3.8 Drop-on-demand piezoelectric ink jet ± push mode.

3.9 Drop-on-demand piezoelectric ink jet ± squeeze mode.

35

36

Digital printing of textiles

3.10 Drop-on-demand piezoelectric ink jet ± nozzle excitation.

3.11 Drop-on-demand piezoelectric ink jet ± porous layer.

In piezoelectric ink jet, waveforms with various levels of complexity can be used to control the whole ejection process (Lubinsky et al., 2002). Pre-pulses can be timed to get the nozzle meniscus to bulge out or in, thereby increasing or reducing the ink present in the front channel. This results in larger or smaller droplets, respectively. More complex waveforms are also used to effectively increase or decrease the drop volume by pumping small droplets that merge into a single (larger) drop shortly after leaving the nozzle. In principle, all of these techniques are capable of adjusting the drop volume over an order of magnitude, though this is not very common in commercial products.

3.2.3 Drop-on-demand thermal ink jet In TIJ an electric heater is typically built inside the nozzle, usually by microelectronic device fabrication techniques. A current pulse is allowed to flow through the heater to quickly raise the temperature of the ink in its vicinity to

Ink jet printing technology (CIJ/DOD)

37

over 300ëC. This causes a vapor bubble to violently nucleate and expand, ejecting an ink droplet through the nozzle orifice. Water tends to cause more explosive bubble growth than other solvents. For this reason, TIJ favors waterbased inks. The TIJ process resembles an explosion. Once the bubble nucleates and starts expanding, there is no point in continuing to provide power to the heater because the bubble is a poor thermal conductor. Thus, the pulse is usually tailored to stop shortly after bubble nucleation. As the bubble expands it cools and its pressure (which starts at over 70 atmospheres in water based inks) drops quickly. The bubble reaches its maximum size and then, just as violently, it collapses, retracting the meniscus to a region inside the channel. After the bubble collapses, capillary action drives the refill process, which continues until the channel is full again, ready to fire. Because of its explosive nature, there is little control over the process beyond the pulse length and power applied. Techniques of providing a short pre-pulse (or train of pre-pulses) to pre-warm the ink in the vicinity of the heater are sometimes used. With these techniques, one can control or modify in a limited way the total ejected ink volume. There are several configurations of TIJ drop ejectors, the most common being the `roof-shooter' and `side-shooter' types. In the `roof-shooter' type shown in Fig. 3.12, the plane where the heater resides is parallel to the nozzle plane. In the `side-shooter' type, the nozzle plane is perpendicular to the heater plane. This configuration is illustrated in Fig. 3.13. There are also `back-shooter' drop generator designs (Lee et al., 2004) where the heater is located on the back side of the nozzle plate, as shown in Fig. 3.14. Canon introduced in 1997 a sideshooter version with multiple heaters that enables drop modulation. A top view of this design is shown in Fig. 3.15. Sony has developed a roof-shooter type drop ejector (Eguchi et al., 2004) that has two independently-driven side-by-side heaters (see Fig. 3.16). This feature can be used to control the directionality of the ejected drop. Energy-efficient configurations with suspended heaters have also been proposed (Kubby, 1998; Hideyuki et al., 2004). In these configurations

3.12 Drop-on-demand thermal ink jet ± roof shooter.

38

Digital printing of textiles

3.13 Drop-on-demand thermal ink jet ± side shooter.

(shown in Fig. 3.17), because the heater is embedded in the ink, a larger portion of the total heat generated during the fire is transferred to the ink, resulting in higher energy efficiency than in the configurations where the heater is in a substrate. Finally, Canon has disclosed in a series of patents (see, for example, Kudo et al., 1998) a drop ejector design with a moveable member that, pushed during the vapor bubble expansion, prevents the ink from flowing into the ink

3.14 Drop-on-demand thermal ink jet ± back shooter.

Ink jet printing technology (CIJ/DOD)

39

3.15 Drop-on-demand thermal ink jet ± multi heater.

reservoir through the rear channel region. This feature would be expected to enhance the energy efficiency of the drop ejector (see Fig. 3.18). The fabrication methods used to make TIJ print heads are typically those that are used by the semiconductor industry. This enables the possibility of building a substantial amount of the drive and control electronics into the print head. As a

3.16 Drop-on-demand thermal ink jet ± double heater.

40

Digital printing of textiles

3.17 Drop-on-demand thermal ink jet ± suspended heater.

3.18 Drop-on-demand thermal ink jet ± moveable member.

result of this integration, a typical commercial TIJ print head can have hundreds of nozzles but only tens of leads. This, coupled with the batch processing economies of IC fabrication techniques, results in low cost, multi-nozzle print head arrays.

3.2.4 Other drop-on-demand ink jet technologies Though the most widely used DOD technologies are PIJ and TIJ, there are other DOD technologies at various stages of development that can potentially succeed

Ink jet printing technology (CIJ/DOD)

41

3.19 Drop-on-demand electrostatic ink jet.

in addressing the customer needs of some markets. In this section we give a sampling of this wide range of ideas by describing five of them. A more detailed description can be found in the book by Stephen F. Pond (Pond, 2000). Similar to a piezoelectric transducer, an electric field can be used directly to move the membrane of an ink chamber and thus produce drop ejection. This technology is enabled by MEMS (Micro Electro Mechanical System) fabrication techniques. This is the principle of operation of an electrostatic ink jet drop ejector shown in Fig. 3.19. Epson has the only commercial product (a point-ofsale printer) based on this technology, though other companies have some level of R&D focused on it. Xerox has developed an ink jet technology in which an acoustic excitation is focused on the free surface of the ink in order to eject a drop (Quate et al., 1991) (see Fig. 3.20). One advantage of this technology is that, in principle, no nozzle structure is needed. On the other hand, the ink level has to be tightly controlled

3.20 Drop-on-demand acoustic ink jet.

42

Digital printing of textiles

3.21 Drop-on-demand thermo-mechanical ink jet.

and patent literature shows (Hadimioglu et al., 1993) that nozzles can be used to control the ink level quite effectively. The technology has been demonstrated, but to our knowledge no product has been commercialized to date. The thermo-mechanical ink jet technology disclosed is another example of drop on demand ink jet (Silverbrook, 2001; Trauernicht et al., 2002). The principle of operation is based on the sudden motion of a composite structure caused by differing coefficients of thermal expansion induced by the heating of an electric resistor. Many embodiments have been disclosed for this concept. In one embodiment the motion of a paddle immersed in the ink behind the nozzle initiates the drop ejection process (see Fig. 3.21). In another, the nozzle structure itself is made to move inward thereby generating a drop. To our knowledge this method is not yet commercial. A mechanism sometimes referred to as `Electro-hydrodynamic Extraction' has been also used to generate ink drops (see, for example, Newcombe et al., 1999). The concept is illustrated in Fig. 3.22. Equilibrium is achieved in the nonprinting state between a negative pressure provided at the ink supply and a standby electric field generated by an extraction electrode located in front of the

3.22 Drop-on-demand electro-hydrodynamic extraction ink jet.

Ink jet printing technology (CIJ/DOD)

43

3.23 Drop-on-demand surface tension driven ink jet.

nozzle. When a drop is needed, a higher potential is applied to the extraction electrode, causing the drop ejection. A collection electrode behind the medium is also required to guide the drop to the medium. Casio commercialized this technology in the early seventies but, to our knowledge, no product is being sold at the present time. Silverbrook has disclosed in a series of patents assigned to Kodak (Silverbrook, 1999) the concept we refer to in Fig. 3.1 as `Surface Tension Driven Ink Jet'. The concept, also called `Liquid Ink Fault Tolerant' (LIFT) ink jet, consists of establishing equilibrium in the nozzle between a positive driving force and surface tension. This driving force can be a positive head pressure or a high voltage differential, both of which would cause the ejection of drops if the ink surface tension were lowered. When an electrical heating element is positioned at the nozzle is activated, the ink temperature increases, lowering the surface tension of the ink and inducing the ejection of a drop. We are not aware of any commercial product that utilizes this technology. The concept is illustrated in Fig. 3.23.

3.2.5 Strengths and weaknesses of the various technologies The technologies described in the previous sections have advantages and disadvantages depending on the application for which they are intended. In this subsection we illustrate this point by comparing the two main DOD technologies: PIJ versus TIJ. The displacement that can be achieved with a piezoelectric material sets a limit to the packing density of nozzles in PIJ. Current techniques and operating voltages typically produce displacements on the order of 0.1 m. Thus to generate a volume change of 30 pL (i.e., 3000 m3), a 30,000 m2 area per firing chamber is necessary. This should be compared to a heater area of about 1300 m2 of a comparable TIJ drop ejector. In reality, the situation is worse because the change in chamber volume required to eject a drop of a given size is

44

Digital printing of textiles

on the order of twice the drop volume. For this reason the native resolution (i.e., the number of nozzles per inch in the direction of the nozzle array) of commercial TIJ heads is significantly higher than for PIJ heads. Of course this problem can be dealt with by laying out extensive two-dimensional arrays of nozzles but at a cost of substantially more real estate. Another advantage of TIJ (mentioned previously) is that semiconductor fabrication techniques are used to manufacture these types of heads. It is therefore possible to integrate the electronics necessary to drive the heaters into the print head. This has been difficult to achieve with PIJ print heads and, to our knowledge, no such devices have yet been commercialized. For the two reasons stated above, TIJ print heads tend to be more compact and less costly than their PIJ counterparts. A problem intrinsic to ink jet technology is the detrimental effect on jetting of the presence of trapped air bubbles in the ink system. A bubble is a compliant element in the system and can absorb a substantial portion of the driving pressure pulse, rendering it totally or partially ineffective. There are many possible sources of air bubbles in ink jet devices. Air dissolved in the ink can nucleate at rough surfaces and sharp edges. Particulates suspended in the ink can also lead to air bubble nucleation. Another source of trapped bubbles is the presence of corners in the ink delivery system that can be difficult to fill in the priming process. PIJ waveforms typically tend to create areas of low pressure in the ink in portions of the firing cycle which tend to exsolve air through a process called rectified diffusion. Rectified diffusion occurs because the rate of diffusion of a gas toward the liquid during the compression portion of the cycle is smaller than the rate at which the gas leaves the liquid in the low pressure portion, causing the bubble to grow. Finally, the heating of the ink during the firing pulse in TIJ devices also causes air ex-solution. Air management is another area where the state of the art TIJ is superior to PIJ. The main reason for this is that TIJ devices have the drop generator energy source very close to the nozzle, which tends to flush air bubbles away from the critical regions more effectively. The ink path from the firing chamber to the nozzle tends to be more complex in commercial PIJ devices and in most cases degassed ink is used. An advantage of piezoelectric ink jet relative to thermal ink jet is ink latitude. Because the vapor pressure of water at the nucleation temperature is abnormally high, water is a very good `propellant'. Though examples of drop ejection of non-aqueous fluids from thermal ink jet devices have been disclosed, all commercially available TIJ print heads fire aqueous inks. Piezoelectric heads, on the other hand, can easily fire any fluid, within a given range of operating viscosity and surface tension. For this reason, most industrial non-conventional applications of ink jet use piezoelectric technology. UV inks, phase-change inks and solvent-based inks, for example, are jetted with PIJ devices. As discussed in Section 3.2.2, another PIJ advantage relative to TIJ is the

Ink jet printing technology (CIJ/DOD)

45

ability to control the volume of the drop through the shape of the waveform. Prepulsing techniques can be used in thermal ink jet to affect the volume of the drop (Becerra et al., 2004) but the effect produced is fairly modest compared to PIJ, where up to an order of magnitude of drop volume variation is possible. The management of the waste heat is an important issue in thermal ink jet. In TIJ devices only a small fraction of the heat generated by the heater is ejected with the drop in one cycle. Therefore, unless measures are taken to control this problem, the temperature of the head increases with use and duty cycle. As the temperature increases, the ink viscosity decreases and the thermal energy stored in the superheated layer of ink at the time of bubble nucleation increases (Freire, 1997). The end result of these effects is that the drop volume drifts upward, causing print quality issues. Heat sinks and/or fluid paths that enable selfcooling are typically used to manage this heat in combination with pre-warming algorithms. In contrast, for piezoelectric devices, most of the energy dissipation occurs in the driver electronics that are typically thermally disconnected from the actual print head. Thus, the operation of piezoelectric devices is naturally more isothermal. Drop ejector lifetime is another aspect where PIJ is generally considered to be more robust than TIJ. Two failure modes unique to TIJ contribute to this difference. The accumulation of ink-related deposits on the surface of the heater, called `kogation', is one of them. These deposits tend to thermally insulate the heater, causing non-uniform nucleation. Over time, drop ejection failure occurs. Bubble collapse is another cause of drop ejector failure in TIJ. This process is very violent and can erode the heater surface through a phenomenon called cavitation damage. Ink formulation and coating the heater surface with highly durable materials are common practices that bring the drop ejector lifetime up to acceptable levels for TIJ applications.

3.3

Aspects to consider and metrics to use in the print head selection process

The decision as to which technology and version to use for a given application has to take into account a variety of factors. These factors can be loosely grouped in four categories: image quality, cost, printer productivity (or throughput), and ink latitude. In this section we discuss these topics and, in some cases, introduce metrics that can help in the technology selection process.

3.3.1 Image quality The key image quality parameter to be considered in making a print head decision is the drop volume. In general, the lower the drop volume, the finer are the details that can be imaged. This is because, given the ink and medium, the drop volume determines the size of the printed dot. As expected, the drop

46

Digital printing of textiles

volumes ejected by commercial print heads have come down significantly over time. The first drop-on-demand thermal ink jet print heads for desktop applications produced drop volumes in excess of 100 pL. Nowadays photo printers use TIJ or PIJ print heads capable of delivering drop volumes as low as 1.5 to 2 pL. When addressing print quality, drop volume should not be confused with resolution. Resolution refers to the scale of the grid over which the dots are placed and it is customary to measure it in `dots per inch' or dpi. Often the addressable points are located in a rectangular grid. In such cases, two resolutions are quoted, one for each of the orthogonal directions. For example, a printer that uses a reciprocating carriage could print at a resolution of 2400  1200 dpi. This means that the dots are placed at a 2400 dpi spacing (i.e., 10.6 m apart) in the direction of the motion of the carriage and at a 1200 dpi spacing (i.e., 21 m) in the perpendicular direction. It follows that increasing the resolution improves print quality but only up to a point. If the resolution is increased to the point that the dot diameter is much larger than the resolution, the print quality improvement is insignificant (and other problems related to drying time and speed could be generated). Moreover, the resolution is essentially a feature of the system. Virtually any resolution can be achieved with a given print head using multi-pass printing and/or adjusting the print head effective resolution by rotation. The drop volume, on the other hand, is a parameter intrinsic to the ink± head combination. Gray scale (i.e., the ability to generate drops of variable sizes from the same print head) is another print head characteristic that is usually considered under image quality. We believe, however, that drop volume (in this case the smallest achievable one) is still the appropriate metric to describe image quality even in print heads with gray scale capabilities. This is because the ability to eject larger drop volumes does not impact image quality but, rather, productivity in solid tones and regions of high area coverage.

3.3.2 Cost The cost of the print heads obviously impacts the cost of the machine. Given the ink and media, the productivity of the printer is determined by the operating frequency and the total number of nozzles in the printer. For this reason, a metric frequently used to normalize cost is the cost per nozzle. The cost of ownership or running cost is obviously impacted by the price of the ink and media, but the print head lifetime is also a factor in this cost because it determines how often the print head needs to be replaced. It is common for print head manufacturers to test their print heads to failure with a recommended ink and use Weibull statistics to determine a minimum life. Many factors can affect print head lifetime. They can range from contamination in the ink delivery system to loss of hydrophobicity of the print head nozzle plate. At the center of

Ink jet printing technology (CIJ/DOD)

47

the problem is the ink±print head interaction. Therefore, unless the manufacturer's recommended ink is used, the quoted minimum life cannot be taken for granted and reliability tests will be needed. Another print head related cost that needs to be considered in the total running cost is ink royalty. Some head manufacturers would add a royalty cost that is usually computed as a percent of ink sales per print head.

3.3.3 Productivity The fastest way of printing at full area coverage is to have all nozzles fire at the maximum allowable frequency. The required amount of ink per unit area for full area coverage printing is a function of the ink and medium. It follows that the productivity of a print head is given by the amount of ink it can deliver per unit time. Thus, PH ˆ n  f  V where PH is the productivity, n is the number of nozzles in the head, f is the operating frequency and V is the drop volume. Note that if the desired printing resolution is not equal to the print head native resolution, more passes will be needed (or the print head would have to be placed so that the array direction is not perpendicular to the printing direction) but the productivity definition stated above still limits the maximum throughput. The productivity metric PH introduced above is clearly a print head centric metric. The more primitive nozzle productivity can sometimes be used (PN ˆ f  V ). From a performance standpoint this is probably the main discriminator between CIJ and DOD ink jet. This is because the maximum operating frequency in state of the art DOD ink jet is in the tens of kilohertz whereas CIJ typically operates at hundreds of kilohertz. Some of the metrics introduced above can be combined to address other questions. For example, one can define the print head productivity cost as the ratio between the head cost and its productivity. The print head productivity cost therefore measures the dollar cost of 1 liter per hour of productivity. Similarly, one can compute the productivity cost per nozzle, i.e., the dollar cost of 1 liter per hour of productivity per nozzle. For example, the cost of CIJ print heads is high. On the other hand, they tend to be very productive because of the high frequencies at which they operate. The productivity cost per nozzle of some CIJ systems is actually quite competitive, making them suitable for high-speed applications.

3.3.4 Ink latitude Another key factor in the technology decision is ink latitude. As discussed in the previous section, commercial TIJ print heads are typically effective for low

48

Digital printing of textiles

viscosity (i.e., less than about 4 cps) water-based inks. Therefore, PIJ print heads are generally used for industrial applications that require operating outside this region. We do not have a good metric to address ink latitude other than the manufacturer recommended range for the viscosity and surface tension of the fluid to be ejected. Within PIJ heads there are certain limitations regarding the ink vehicle. Some print heads are manufactured with materials (typically adhesives) that are affected by the presence of water. Those heads cannot be used with aqueous inks. The conventional continuous ink jet process requires that the droplets be charged after ejection. Therefore the conductivity of the inks used in conventional CIJ heads needs to be high.

3.4

Companies currently active in print head technology

Ink jet technology has been around for many years and many companies have entered and exited this field. In this section we provide the reader a sense of the size of the ink jet print head field by listing all the companies that we believe are actively working on print head technology. The list was constructed from attendance at trade shows, researching the patent literature, and by Internet searches. We do not claim to have a complete list since the field is highly populated but we believe the list captures the major players. The field of TIJ print heads has been historically controlled by four companies that had enough intellectual property to practice the art. These companies are Canon, Hewlett-Packard, Lexmark, and Xerox. Xerox exited the TIJ business in 2001 leaving only three major players. Patent activity shows that other companies are actively pursuing this field, motivated by the expiration of the some of the first TIJ patents. In the desktop market the main PIJ player is Epson, followed by Brother. In industrial DOD ink jet the three major players are Epson, Spectra and Xaar with its licensees. The main players in CIJ are VideoJet, Domino and Imaje. Due to the robustness required by the textile industrial application and the relatively higher viscosity of some textile inks, the textile market is currently dominated by PIJ technology. Most textile inks are also water-based so only those PIJ print heads that are water-compatible are being used. CIJ could also serve this sector, but the constraints coming from the roll-fed media combined with the high carriage speed required to take advantage of the high operating frequency of this technology make the implementation more difficult. Stationary multiple jet arrays could solve this problem but the cost (and throughput) would put such a machine in a completely different class. Osiris has announced a CIJbased printer but, to our knowledge, the product is not yet commercial. Accordingly, the print heads with significant presence in the textile market are currently those offered by Aprion, Epson, and Seiko.

Ink jet printing technology (CIJ/DOD)

49

The list of major print head manufacturers is shown in Table 3.1. The table includes the type of technology and whether the company is known by the author to have commercialized any print heads. We have also included a column to capture the companies that currently serve the digital textile printer market. Note that the table is print head centric. This means that printer integrators that outsource the print head technology are not included.

3.5

Future trends

There are several developments happening at the moment that are likely to shape the future of this technology. A major factor is that many of the original patents are currently or will be expiring in the near future. This is an incentive for new entrants into the various areas of the technology, including print heads. Thus the market may see new players in the future which, in turn, may generate new concepts as well as drive prices down. Companies in the Far East will likely take advantage of this opportunity and we can see this trend already in the patent literature with active players such as Samsung, ITRI, and BenQ, among others. Another factor that is likely to influence the field in the future is the development of page-wide array systems. From the early days of this technology several companies have worked on the development of a page-wide print head that could print a full page without the need of a reciprocating carriage. One key challenge of page-wide array systems is that they would operate in single-pass mode. Virtually all multi-nozzle printing systems currently have multi-pass printing modes to ensure the highest print quality by minimizing issues of directionality, missing jets, and other nozzle-to-nozzle non-uniformities in the head. Operating in single-pass mode requires a much higher print head quality level than that needed in current products. This is one of the technical reasons why page-wide printers are not widely available. Sony has recently announced a page-wide system using its proprietary double heater TIJ technology. The printer is being sold in Japan. Brother has also announced a page-wide array product and demonstrated it at the 2005 World Exposition at Aichi, Japan (`EXPO 2005') on 25 March 2005. From their disclosure, the print bar is made out of trapezoidal two-dimensional piezoelectric nozzle arrays. In the CIJ arena, we think that the thermal excitation technology developed by Kodak is quite promising. According to Kodak's disclosures, the productivity cost per nozzle is better than for any other technology. The ability of using the nozzle heaters to correct jet directionality in a large array of nozzles could potentially be very valuable as well.

3.6

Sources of further information and advice

Ink jet is an ever-evolving area and advances are being made constantly. It is therefore common for literature to become dated quickly. However, the funda-

Table 3.1 List of companies currently active in print head technology No. Organization

Technology

Sub-type

Commercialized

Textile application

1 Atlantic Zeiser 2 Danaher (Linx) 3 Danaher (VideoJet)

CIJ CIJ CIJ

4 Domino

CIJ

5 Imaje 6 Kodak

CIJ CIJ

Binary Multiple deflection Binary Multiple deflection Binary Multiple deflection Multiple deflection Binary Hertz Thermal excitation Multiple deflection Multiple deflection Hertz Shear mode Bend mode Push & bend modes

None known Yes Yes Yes Yes Yes Yes Yes Yes None known Yes Yes Yes Yes Announced Yes

± Bend mode Shear mode Shear mode Bend mode Roof shooter Side shooter Squeeze mode Squeeze mode Bend mode Bend mode Push mode

Yes Yes Yes None known Announced None known None known Yes Yes Yes Yes Yes

None known None known None known Special applications None known None known Osiris (announced) None known None known None known None known Digital Printing Systems Stork Amethyst None known None known Epson, Mimaki, Robustelli, Mutoh, Hollanders, USSPI None known None known Nassenger Series None known None known None known None known None known None known None known None known None known

7 8 9 10

Matthews Scitex (Jemtex) Stork Brother

CIJ CIJ CIJ PIJ

11 Epson

PIJ

12 IJT 13 Konica-Minolta 14 Kyocera

Electrostatic PIJ PIJ PIJ TIJ

15 16 17 18 19

Microdrop Microfab Panasonic Picojet Ricoh (Hitachi)

PIJ PIJ PIJ PIJ PIJ

20 Scitex (Aprion) 21 Seiko Instruments 22 Spectra

PIJ PIJ PIJ

23 24 25 26 27

PIJ PIJ PIJ PIJ PIJ Electro-hydrodynamic PIJ TIJ PIJ TIJ

Toshiba-TEC Trident Xaarjet Sharp The Technology Partnership 28 Samsung 29 Xerox

30 BenQ 31 Canon 32 33 34 35 36 37 38

HP ITRI Lexmark Microjet Olivetti Sony Fuji Xerox

39 iTi 40 Fuji Photo 41 Silverbrook Research

Acoustic Electrostatic TIJ TIJ TIJ TIJ TIJ TIJ TIJ TIJ TIJ PIJ Electrostatic Electro-hydrodynamic Surface tension Thermo-mechanical

Porous layer feed Shear mode Shear mode

Yes PIJ Yes

Bend mode Shear mode Push mode Shear mode Bend mode Nozzle excitation ± Bend mode Suspended heater Bend mode Side shooter Suspended heater ± ± Back shooter Side shooter Roof shooter Roof shooter Back shooter Roof shooter Roof shooter Roof shooter Double heater Side-shooter Bend mode ± ± ± ±

Announced Yes Yes Yes None known None known None known None known None known Yes Yes (abandoned) None known None known None known None known Yes Yes Yes None known Yes Yes Yes Yes TIJ None known None known None known None known None known

Reggiani DuPont Artistri 2020 DuPont Artistri 3210, Leggett & Platt, Kornit None known None known None known None known None known None known None known None known None known None known None known None known None known None known None known Canon None known HP, ColorSpan None known Encad None known None known None known None known None known None known None known None known None known

52

Digital printing of textiles

mentals of the technology and a large portion of the drop ejector designs shown in Fig. 3.1 have not changed significantly. For those topics, the book by Stephen Pond (Pond, 2000) is an excellent reference for further reading. Also, the paper by Hue Le (Le, 1998) contains a good description of the more traditional drop ejector designs. All the references cited in the text are listed below.

3.7

References

Anagnostopoulos C et al. (2001), US Pat. No. 6,217,163 and references therein. Arnott M et al. (2002), US Pat. No. 6,394,363. Becerra J et al. (2004), US Pat. No. 6,698,862 and references therein. Eguchi T et al. (2004), US Pat. No. 6,817,704 and references therein. Fishbeck K et al. (1989), US Pat. No. 4,825,227. Freire E (1997), `Effects of geometry on the drop volume sensitivity to temperature in TIJ printheads', IS&T's NIP13, p. 694. Hadimioglu B et al. (1993), US Pat. No. 5,229,793. Haluzak C et al. (2004), US Pat. No. 6,685,302. Hawkins G (2003), `Next generation continuous ink jet technology', 11th Annual European Ink Jet Printing Conference, Lisbon, Portugal. Hertz C et al. (1986), US Pat. No. 4,620,196. Hideyuki S et al. (2004), US Pat. No. 6,834,940 and references therein. Kubby J (1998), US Pat. No. 5,706,041. Kudo K et al. (1998), US Pat. No. 5,821,962. Le H (1998), `Progress and trends in inkjet printing technology', J. Imaging Sci. Technol., 42(1), 49±61. Lee C et al. (2004), US Pat. No. 6,749,762 and references therein. Lubinsky A et al. (2002), US Pat. No. 6,428,135 and references therein. Newcombe G et al. (1999), US Pat. No. 5,992,756. Pond S (2000), `Inkjet technology and product development strategies', Torrey Pines Research, Carlsbad, CA. Quate C et al. (1991), US Pat. No. 5,041,849. Rayleigh Lord (1878), `On stability of jets', Proc. London Math. Soc., 10(4), 4±13. Silverbrook K (1999), US Pat. No. 5,856,836. Silverbrook K (2001), US Pat. No. 6,243,113. Temple S et al. (1995), US Pat. No. 5,463,414. Trauernicht D et al. (2002), US Pat. No. 6,460,972.

4 Drop formation and impaction W W C A R R , H P A R K , H O K , R F U R B A N K and H D O N G , Georgia Institute of Technology, USA and J F M O R R I S , City College of New York, USA

4.1

Introduction

The dynamics of formation and impaction of drops are physical processes of clear relevance to design and control of inkjet printing technologies. While the processes are coupled, in the sense that the drop formation process influences the size, velocity and frequency of the impacting drops, the two processes have typically been studied separately. For that reason, the processes are described as distinct in this chapter. Topics which are important for understanding the processes in applications of inkjet printing to textile materials, in particular the role of suspended particulates and nonsmooth surfaces, are discussed in this chapter. We begin, in Section 4.2, with a discussion of particulate effects observed in drop formation. The studies described focus on slow drop formation from suspensions of noncolloidal particles in order to allow explicit consideration of the mechanical influence of particles, about which little is known. Engineering of processes has proceeded without firm scientific basis, including jetting of ceramic materials (Blazdell et al., 1995; Windle and Derby, 1999) as well as pigments and polymeric binders onto textiles (Tincher et al., 1998). Because the conditions in such applications are more rapid and at smaller scale than those of our studies, we have focused upon the role of particles in the necking and pinchoff processes, events which are thought to be generic as they force the flow scale to that of the particles, regardless of the rate, absolute size, or the relative sizes of particle and orifice. Note that the abundant prior study (as reviewed by Eggers, 1997) and more recent and ongoing research (e.g., Ambravaneswaran et al., 2002; Chen et al., 2002) on drop formation processes involves almost no work devoted to solid±liquid mixtures, with a few notable exceptions (Alaoui, 1991; Ogg and Schetz, 1985; Furbank and Morris, 2004); there are also a few studies involving viscoelastic liquids (Goldin et al., 1969; Christanti and Walker, 2001), but the rheology of suspensions is very different from that of these liquids. We follow the discussion of drop formation with a consideration of drop impaction. The size of a printed dot in inkjet printing, which greatly affects print

54

Digital printing of textiles

quality, is determined by spreading of an ink drop when it impacts the substrate (Asai et al., 1993). The study of impact dynamics is thus important in determining the ultimate spreading and will be covered in Section 4.3. Most prior studies have been conducted using homogeneous liquid drops impacting smooth surfaces. A general description of spreading without splashing for homogeneous liquid drops impacting on smooth surfaces is covered in Section 4.3.1. In textile printing, an understanding of the interaction of an individual drop with various textile surfaces is needed. However, obtaining sufficiently high resolution images of an inkjet drop impacting on a textile surface is difficult due to the small drop size (less than 100 microns) and high impact speed (around 5±20 m/s). For this reason, the interaction of an individual drop with various textile surfaces has not been studied. However, Park (2003) used a scaled-up experiment to simulate the impaction of an ink drop on a fabric. Drop impaction on a textile-like structure is presented in Section 4.3.2. As noted, a growing number of nontraditional applications of inkjet technology contain solid particles. These particles have various purposes, depending upon the application. They serve as colorant or binder in the textile printing applications, but may also be ceramic or metallic particles in other applications. Although progress has been made in the design, formulation and utilization of such inks, impaction of particle-laden drops on surfaces has received little attention (Carr et al., 2004). In fact, only one paper was found in the refereed literature on impaction of particle-laden drops on surfaces (Ok et al., 2004). An on-going investigation of the effects of particles on the impaction process is discussed in Section 4.3.3.

4.2

Drop formation from particle-laden liquids

It is to be expected that the introduction of particles to a liquid from which drops are to be formed adds to the complexity of the problem. The parameter space needed to describe the problem expands, and to the dimensionless groups needed to describe the problem for a pure liquid (typically Reynolds, Re ˆ vd=, and Weber numbers, We ˆ v2 d= , or the capillary number, Ca = We/Re in place of We), we must add at least the solid volume fraction , and the ratio dp =d of the particle size dp (diameter if spherical) to the orifice diameter d. The axial velocity v, liquid (or suspension) density , and gas±liquid surface tension are used in the above dimensionless numbers. This is by no means a complete description in the actual application, where particles may be small enough that thermal forces inducing Brownian motion (and hence non-infinite Peclet number) as well as colloidal forces need to be considered. Drop formation from an orifice, regardless of flow rate and length scale of the orifice, involves the formation of a neck which connects the forming drop to the fluid remaining at the orifice. This neck thins and stretches to a thread until the action of surface tension causes a pinch-off, or bifurcation, of the thread to form

Drop formation and impaction

55

the drop. The stretching is affected by the liquid viscosity, and it is well known that addition of particles to a fluid causes an increase in the effective viscosity, eff …†, of the mixture relative to the suspending liquid. This is by its very nature a continuum description of particle influence, and may only be expected to have validity above some minimum lengthscale (relative to the particle size, with the width of the thread measured in particle diameters using an example relevant to the drop formation). Here, the role of particles identified in slow drop formation (and transition to slow jetting), but believed to be generic to other conditions, is considered. Given space limitations, the manner in which the particles stabilize or destabilize the necking process leading to drop formation is the primary focus. For slow drop formation, it is useful to consider a two-stage necking model. Early in the process (first stage), eff describes the added resistance to necking and provides a robust description of particle effects. In the final pinching (second stage), rapid thinning through what must ultimately be a pure liquid region occurs at a relatively localized axial location. The first stage is a continuum description of the particulate effect, with only the need for a solid fraction, . The second stage is intrinsically a finite-size effect of the discrete particles, as it involves the fluctuations in , and hence large fluctuations in the ability of the mixture to resist thinning; these may be viewed as fluctuations in the local viscosity. The two-stage process, and the difference from a pure liquid drop formation event, is illustrated by the sequences in Fig. 4.1: these show that the pure liquid continues to thin and stretch over its entire length right up to bifurcation. The suspension thins in a similar fashion early, and then abruptly pinches at a localized position, leaving relatively thick and slowly retracting cone-like `spindle' structures up- and downstream of the bifurcation point (Furbank and Morris, 2004, 2006). Quantitative demonstration of this influence for  ˆ 0, 0.05, and 0.40 is shown in Fig. 4.2, where the evolution with time of the minimum radius for the thread (normalized by the suspended particle radius, a ˆ dp =2) for a drop formed at Q ˆ 0:25 cm3/min from a 0.16 cm orifice (same conditions as Fig. 4.1) is illustrated as a function of t0 ÿ t, where t0 is the time of the pinch event; the detected radius determined by automated image analysis reaches zero before t0 owing to resolution limitations, and the pinch time is corrected based on visual analysis of the image sequence. The behavior is seen to be similar between the various cases for the larger values of t0 ÿ t plotted, but the pure liquid radius curve shows an inflection to a slower thinning rate near the pinch, while the suspensions from  ˆ 0:05 and 0.4 thin more rapidly as pinch is approached; other solid fractions behave similarly as shown by Furbank and Morris (2006). This higher thinning rate is the result of a change from uniform thinning over the entire thread to localized thinning through the lower viscosity of the pure liquid. The first stage thinning is found to satisfy a scaling as Bd  … or eff †ÿ1=3 for pure liquids of several viscosities as well as suspensions with a range of values of  (and hence eff ). Here B is the thinning rate from the

56

Digital printing of textiles

4.1 Sequence of images approaching pinch-off of a drop in the slow formation of drops from a pure liquid,  ˆ 0 (left of heavy dark vertical line) and  ˆ 0:20 suspension, for d ˆ 0:16 cm and dp ˆ 106±125 m in the suspension. The time between images in both sequences is
t ˆ 0:004 s. The flow is at Q ˆ 0:25 cm3/min (Re ˆ 0.01) in both cases.

4.2 Minimum detected thread radius scaled by the particle radius, a, for pure liquid ( ˆ 0) and suspensions of  ˆ 0:05 and 0.40; d ˆ 0:16 cm and dp ˆ 106±125 m in the suspension. The flow is at Q ˆ 0:25 cm3/min in all cases.

Drop formation and impaction

57

Table 4.1 Fitting parameters to the model equation given in the text for the radius as a function of time from pinch, with the last column giving an indication of the quality of the fit through the value of the regression parameter d (cm)

dp (m)



A

B (1/s)

C

R2

0.16 0.16 0.16 0.16 0.32 0.32 0.32 0.32 0.32 0.32 0.32 0.32 0.32

106±125 106±125 106±125 106±125 212±250 212±250 212±250 212±250 212±250 106±125 106±125 106±125 106±125

0.05 0.10 0.20 0.30 0.05 0.10 0.20 0.30 0.40 0.10 0.20 0.30 0.40

15.7 14.5 14.0 12.4 14.8 13.7 12.5 10.6 9.3 26.5 24.8 23.0 21.2

36.0 34.6 32.3 25.1 19.0 18.1 17.5 14.7 10.8 17.9 17.9 13.1 11.5

ÿ2.3 ÿ1.5 ÿ0.8 0.6 ÿ2.9 ÿ1.5 ÿ0.5 1.2 2.1 ÿ2.7 ÿ0.7 0.2 2.0

0.999 0.999 0.999 0.999 0.994 0.994 0.995 0.997 0.998 0.993 0.995 0.998 0.999

Source: Furbank and Morris (2006).

fit to the experimental R…t† curve to the form suggested by Clanet and Lasheras (1999), R=a ˆ A…1 ÿ exp‰ÿB…t0 ÿ t†Š† ‡ C, where a ˆ dp =2 provides a convenient nondimensionalization. The fitting parameters determined are shown for the suspensions in Table 4.1. The second stage is found also to correspond roughly to the onset of increased fluctuations, both in the position of the minimum radius and in the length of the material attached to the orifice. Related behaviors found, but not described in detail here, include · Satellite formation: The introduction of particles is found to cause a pronounced reduction in the number of satellite drops formed in slow drop formation (Furbank and Morris, 2004). This is illustrated by the sequences in the figure, where the pure liquid forms a small satellite, but none appears at  ˆ 0:2. Note, however, that the few satellites formed from suspension are typically much larger, as these arise from pinch at two points through a much thicker thread. · Transition rate and length: Particles cause a transition from dripping to jetting at lower flowrates, and the coherent jet is greatly reduced with even 2% solids (Furbank and Morris, 2004). This is evidence of the fluctuations caused by the particles destabilizing the thread or column of fluid.

4.3

Drop impaction

When a drop impacts a rigid solid surface, the outcome depends on several factors including drop speed, drop volume, liquid physical properties (viscosity,

58

Digital printing of textiles

surface tension, and density), solid surface energy, drop/surface interaction, and surface characteristics. Certain of these parameters are commonly combined to form the dimensionless Reynolds number (Re ˆ vd=), Weber number (We ˆ v2 d= ), and equilibrium contact angle (). In these relationships,  is liquid density, v is drop impact speed, d is drop diameter at impact,  is liquid viscosity, and is liquid±vapor surface tension. At high Re, the drop may bounce or splash, forming secondary or satellite drops. The splashing threshold has been correlated with We, Re, and roughness (Ra ) (Stow and Hadfield, 1981). More recently, the experimental results of Range and Feuillebois (1998) indicate that the dimensionless numbers (Ohnesorge number, Oh, and Re) containing viscosity are not important and can be neglected in the description of splashing. For their data, Wec (critical We for splashing) is found to correlate with the ratio of drop radius (Ro ) to the surface roughness (Ra ) for a given liquid±surface combination. They also point out that Ra is not the only parameter characterizing the effect of the splashing limit. The surface profile is important, but is not entirely described by Ra . A complete understanding of splashing is still not available, especially about the influence of the solid surface parameters. Clarke et al. (2002) show that for typical inkjet printing, splashing will not occur on a smooth surface. For that reason, splashing is not discussed further in this chapter. Spreading of liquid drops on porous substrates has received much less attention even though it is common and important, for example in inkjet printing on paper and textiles. Experimental study for inkjet systems is challenging since drops are very small and the substrates vary widely in their properties. Dynamic spreading occurs very rapidly, and penetration on porous substrates can result in further spreading. The initial spreading phase after impact occurs very rapidly relative to penetration. Hence the dominant physical processes change, as kinetic and surface energies dominate during spreading, while capillary forces dominate during penetration. A full analysis of the penetration process is a formidable task, but some progress has been made, including a model describing the spreading and imbibition of liquid drops on a porous surface developed by Clarke et al. (2002).

4.3.1 Homogeneous-liquid drop impaction on smooth surfaces Since print quality is related to spreading of an ink drop when it impacts the substrate (Asai et al., 1993), impact dynamics is of central importance to inkjet printing. The parameter usually followed during impaction is the spreading ratio, D ˆ D=d, where D is the contact diameter and d is the drop diameter before impact. The variation of D with time during impaction is illustrated schematically in Fig. 4.3. Before impact, the energy of the impacting drop consists of kinetic energy, surface energy, and potential energy. After impact,

Drop formation and impaction

59

4.3 Schematic of drop impaction process on a smooth surface.

the drop spreads over the surface. Assuming uniform spreading, the wetted contact area remains axisymmetric (circular), and spreading is characterized by the diameter, D, of the circle. Spreading continues until it reaches a maximum, Dm . At Dm , the surface energy of the drop is at a maximum while the kinetic energy is zero. Excess surface energy causes retraction to occur. The amount of retraction depends on several factors including Re, We and equilibrium contact angle (). For very low values of Re and We and , little retraction may occur. For this case, viscous and liquid±surface interactions dominate, and Dm is close to the equilibrium value of D* which can be estimated using the relationship D ˆ ‰4sin3 =…2 ÿ 3cos ‡ cos3 †Š1=3 derived by Ford and Furmidge (1967). As equilibrium contact angle is increased, the tendency to retract increases. Also, as Re is increased, the amount of retraction increases. The liquid may retract to the equilibrium position and stop, or retract through the equilibrium position and rise in the region of the initial impact. Sometimes the liquid will separate from the surface, rise a short distance and return to the surface. This phenomenon,  90ë. After maximum referred to as rebounding, is only observed for values of  > retraction, the drop changes its direction of motion and begins to spread again. The liquid may spread to the equilibrium position and stop or expand through the equilibrium position until it reaches the second maximum spreading diameter which is smaller than initial maximum spreading diameter. This type of damped oscillatory motion will continue until excess surface energy is dissipated, and the drop reaches its equilibrium D . Since Worthington (1876) reported an investigation of drops of liquids falling vertically on a horizontal plate, there have been over 100 published

60

Digital printing of textiles

investigations on the subject. While some have been entirely experimental (Bergeron et al., 2000; SÏikalo et al., 2002), most studies have included theoretical and/or numerical modeling approaches for predicting the spreading phenomenon. The theoretical approach (Engel, 1955; Ford and Furmidge, 1967; Chandra and Avedisian, 1991; Asai et al., 1993; Fukai et al., 1998) involves the use of an energy balance on the system, which consists of the drop and the impacted surface, to develop an equation for predicting the maximum spreading ratio, Dm (ratio of the maximum spreading diameter to initial drop diameter) as a function of drop properties and contact angle. Numerical modeling (Harlow and Shannon, 1967; Fukai et al., 1993, 1995, 1998; Pasandideh-Fard et al., 1996; Bussmann et al., 1999) has been used to simulate the dynamics of transient flow and to predict the drop impacting process. These studies provide firm understanding of the effects of impacting velocity and liquid properties, i.e., viscosity and surface tension, on the impacting process. However, understanding of the influence of solid±liquid interaction during spreading and recoiling is far from complete, especially its relative importance during different stages. Several equations for predicting Dm based on correlations and/or energy conservation are available in the literature. One of the first correlation equations was presented by Engel (1955). Since then, there have been several efforts (Ford and Furmidge, 1967; Chandra and Avedisian, 1991; Asai et al., 1993; PasandidehFard et al., 1996; Mao et al., 1997; Fukai et al., 1998) to improve the accuracy of the prediction. The earlier investigations are summarized by Mao et al. (1997) and Fukai et al. (1998), who presented two of the most recent predictive equations. Their models accurately predict Dm for most cases except at low Re and We, where they overestimate the experimental values, particularly at high and low contact angles, or may give negative or imaginary values at high contact angles. Mao et al. (1997) improved the model of Chandra and Avedisian (1991) and Pasandideh-Fard et al. (1996). In the earlier work, surface±vapor and surface± liquid interaction energies during spreading were included, which resulted in cos, where  is contact angle, appearing in the predictive equation. While Pasandideh-Fard et al. (1996) used the advancing contact angle in their predictions, Mao et al. (1997) used the static contact angle. Fukai et al. (1998) presented a model also based on that of Chandra and Avedisian (1991). They improved the predictions by modifying the model to contain three empirical coefficients, which were determined by fitting to their numerical results. Park et al. (2003) developed a model which gives improved predictions for low drop impact velocities by assuming a spherical cap model of the impacting liquid (rather than a circular cylinder). A summary of several models for predicting Dm is given in Table 4.2. The impaction studies discussed above have used a single millimeter-sized drop impinging on a smooth surface. Very little is available to show that these results scale down to micron-sized drops used in inkjet printing, but this is an issue currently under investigation (Carr et al., 2004).

Drop formation and impaction

61

Table 4.2 Summary of previous models for predicting Dm Reference Chandra and Avedisian (1991)

Models     3 We 4 1 ÿ We ‡ 4 ˆ0 Dm ‡ …1 ÿ cose †D2 m 2 Re 3

Asai et al. (1993)

Dm ˆ 1 ‡ 0:48We0:5 exp‰ÿ1:48We0:22 Reÿ0:21 Š

Scheller and Bousfield (1995)

Dm ˆ 0:61…Re2 Oh†0:166

Fukai et al. (1998)

  1 We We 4 2 D ‡ 2:29…1 ÿ cos †Dm ÿ ‡4 ˆ0 2 Re0:772 m 3

Pasandideh-Fard et al. (1996)

Mao et al. (1997)

Park et al. (2003)

s We ‡ 12 p D ˆ 3…1 ÿ cosa † ‡ 4…We= Re† " !#   1 We0:83 We 2 3 …1 ÿ cose † ‡ 0:2 ‡ 1 D ‡ ˆ 0 D ÿ 4 12 3 Re0:33 

For high viscosity liquids …Re < 81…Dm †4 †     We 4 1 1 ÿ cos  1 0:53 ÿ Dm ‡ cos D2 mÿ Re 2 4 sin2  1ÿ

We
ESP ˆ0 ‡ 2 12 d LV

For low viscosity liquids …Re > 81…Dm †4 †    We 1 1 1 ÿ cos 0:33 p ÿ cos ‡ D2 mÿ 2 sin2  Re 4 We
ESP ˆ0 ‡ 1ÿ 12 d 2 LV where

p We  Oh (Ohnesorge number), Oh ˆ p ˆ Re d

ˆ dynamic contact angle a ˆ advancing contact angle between liquid and solid     cos 1 1 ÿ cos g…† ‡ 1 ÿ g…†
ESP ˆ d 2 LV 4 2 sin2  " #2=3 4sin3  g…† ˆ D2 ˆ 2 ÿ 3cos ‡ cos3  D2 m

#2=3 4sin3  ˆ 2 ÿ 3cos ‡ cos3  "

62

Digital printing of textiles

4.3.2 Drop impaction on a textile-like rough surface In textile printing, an understanding of the interaction of an individual drop with various textile surfaces is needed. However, obtaining sufficiently high resolution images of an inkjet drop impacting on a textile surface is difficult due to the small drop size (less than 100 microns) and high impact speed (around 5±20 m/s). For that reason, Park (2003) used scaled-up experiments to simulate the impaction of an ink drop on a fabric. A micrograph taken by SEM of a woven rayon fabric is shown in Fig. 4.4. Notice that the yarn is made up of many fibers running in the warp direction. The cross-section of the fiber is a serrated circular shape, and the fiber has lengthwise striations. Since the width of the fibers is about 25 microns, approximately 10 fibers will be on the surface of the 250-micron-wide yarn. Circles with diameters of 20 and 80 microns are drawn on one of the warp yarns to indicate the size of typical inkjet drops. The ratio of the diameter of typical inkjet drops to the width of the rayon fibers ranges from 0.8 to 3.2. A surface was made to simulate a yarn on the fabric surface. Monofilament yarns (polyester coated with ethylene tetrafluoride) with diameter of about 1.3 mm were placed next to each other on a smooth surface and glued together to produce the surface illustrated in Fig. 4.5. The diameter of the drops used to

4.4 A SEM photograph of a rayon fabric (circles with diameters of 20 and 80 microns are the size of typical inkjet drops).

Drop formation and impaction

63

4.5 Schematic of rough surface.

impact on the surface was 2.3 mm. Thus the ratio of diameter of the impacting drop to the diameter of the monofilament is 1.8, which falls in the range for typical inkjet drops on the rayon yarn mentioned above. Figure 4.6 shows the spreading of a water drop on a smooth surface and on the rough surface simulating a continuous filament yarn. Tests were conducted with the impacting drop hitting the rough surface at three different locations: the center of the filament-like structure (position 1), the middle of the valley between two of the filament-like structures (position 2), and between these positions (position 3). A series of images of drop impingement for the three impact positions and the smooth surface were recorded using almost identical time steps. For the smooth surface, the liquid flows radially outward from the impact point. In contrast, for the rough surface, much of the liquid flows in the filament axial direction rather than strictly radially, because of `roughness' elements blocking the direction perpendicular to the filament axis. The spreading and retracting shapes and maximum spreading ratios thus depend on the impact position. The maximum radial spreading ratio is largest for impact position 2 while the maximum spreading ratio in the filament axial direction is the largest for position 1. The equilibrium diameters in the radial direction for positions 2 and 3 are almost equal and are larger than for position 1 due to the noted structural barrier. We note that very recent work describing the influence of a small obstacle on a surface upon splashing of impacting drops (Josserand et al., 2005) may provide some guidance to understanding of the role of roughness.

4.3.3 Particle-laden drop impaction on smooth surfaces Pure fluid drop formation and drop impaction on solid surfaces have been studied for over 100 years. In contrast, impaction of particle-laden drops on surfaces has received little attention despite its importance in a variety of applications including inkjet printing. In fact, only one paper was found in the

64

Digital printing of textiles

4.6 Impact of a 2.3 mm water drop on a smooth Teflon surface and on a rough surface produced by aligning and gluing polyester monofilaments coated with ethylene tetrafluoride on a silicon wafer. Amplitude and texture of roughness: 1.25 and 13 mm, respectively. Impact speed ˆ 0.87 m/s, Re ˆ 2000, and We ˆ 24.

refereed literature on drop impaction of particle-laden drops on surfaces (Ok et al., 2004). In the study of the impact dynamics of particle-laden liquid, the parameters dp =d and  must be considered in addition to Re, We, and . Here dp =d is the ratio of particle diameter to drop diameter, and  is the volume fraction of particles in the liquid. Research by several of the authors is being conducted to provide insight into the effect of particles on drop impaction on solid surfaces (Carr et al., 2004). The goal of the study is to develop understanding of how and why solid particles at a range of concentration affect the drop impaction process. Some of the results of the experimental study on the effects of particles on drop impaction are now discussed. The study has revealed that particles can affect maximum spreading ratio and retracting/rebounding, but the effects depend on the conditions, characterized by dimensionless numbers. Some observations made are:

Drop formation and impaction

65

· At low Re, particles have little effect on impaction process for values of  up to 0.30. · At high Re, particles affect maximum spreading ratio and retraction if  is sufficiently high. For particle volume fractions typically found in inkjet inks, particles have little effect on the impact process. · Increasing dp =d from 0.007 to 0.014 for a 2900-m drop had little effect on the impact process. · Particle volume fraction greatly affects rebound behavior for a hydrophilic drop on a hydrophobic surface.

4.4

Future trends

Although drop formation and impaction have been intensively studied, they remain areas of active research interest. Even the best-understood case of drop formation from Newtonian liquids is under active study, in part because of the new experimental tools allowing extremely high rates of imaging and numerical techniques and CPU power allowing detailed calculations of the behavior near the bifurcation, or pinch-off point. Future research is expected to explore the process in detail for particle-laden liquids. This area is of significant and growing interest as it influences nontraditional inkjet technology, such as textile inkjet printing application, in which solids are pigmented particles and/or binder, but is also relevant to metallic inks used in digital application of electronic materials and to ceramics-laden inks (Blazdell et al., 1995; Windle and Derby, 1999; Tay and Edirisinghe, 2001). A key reason for the slower progress in mixtures is the lack of a continuum model, although much progress has been made in this area (Morris and Boulay, 1999); even with a reliable continuum model, such an approach is limited by the fact that eventually the intrinsic graininess of the particle-laden liquid has an influence, as in necking and bifurcation a finite-time singularity yields a flow scale going to zero, and in impacting the film resulting from a drop may be below the particle size. Studies are needed to further investigate the formation process and the impaction process. For the latter, an emphasis on smooth surfaces seems warranted for drop sizes typically found in inkjet applications, in order to determine whether experimental results for drops of millimeter size have validity down to drops on the order of 10 microns in diameter. Study is needed for pure-liquid drops, particle-laden-liquid drops and predictive models. · Pure-liquid drops: `The size of a printed dot in inkjet printing, which greatly affects print quality, is determined by spreading of an ink drop when it impacts the substrate' (Asai et al., 1993). Numerous studies on the impacting of pure liquid drops on solid surfaces have been conducted; however, drop size has typically been about two orders of magnitude larger than drops used in inkjet printing. The larger drops ranging in size from 1 to 5 mm have been

66

Digital printing of textiles

used because experimental study of the impacting and spreading process is much easier with drops of this size than with smaller drops. Demonstration that these results for the larger drop size apply for drop sizes typically found in inkjet printing is needed. The formation process is quite well understood for pure liquids, and while added valuable knowledge will come from further work, the engineering need is less in this area. · Particle-laden-liquid drops: Studies of the effect of particles in the liquid on the drop formation and impacting process are needed because solids, serving as colorant or binder, are required in `inks' needed for a number of nontraditional applications of inkjet technology such as textile printing, as well as in ceramic dispersions applied by the inkjet method. Studies of the basic influence of particulates in the drop formation and spreading, as well as the dependence on particle size, are needed, with particular emphasis on examination of drop formation at the small scales and high rates typical of inkjet applications. · Predictive models: Predictive models which are physically based, and incorporate best present understanding, should be developed for micron-size pure-liquid drops, millimeter-size particle-laden-liquid drops, and micronsize particle-laden-liquid drops. Such models cannot be expected to be complete given present knowledge, but are nonetheless immediately valuable for development, and also provide a critical framework for further study and utilization of experimental results. Development of continuum models of mixture flow behavior, coupled to a statistical description of the influence of particles when the continuum description breaks down (at small scales), appears to be a fruitful direction for both scientific and engineering advances in the inkjet application of solids-laden liquids.

4.5

References

Alaoui O (1991), Contribution aÁ l'eÂtude des instabiliteÂs de jets injection de liquides et de suspensions dans un autre fluide non-miscible. PhD thesis, l'Universite de Provence (Aix-Marseille I). Ambravaneswaran B, Wilkes E D, and Basaran O A (2002), `Drop formation from a capillary tube: comparison of one-dimensional and two-dimensional analyses and occurrence of satellite drops', Phys. Fluids, 14(8), 2606±2621. Asai A, Shioya M, Hirasawa S, and Okazaki T (1993), `Impact of an ink drop on paper', J. Imaging Sci. Tech., 37(2), 205±207. Bergeron V, Bonn D, Martin J Y, and Vovelle L (2000), `Controlling droplet deposition with polymer additives', Nature, 405 (15 June), 772±775. Blazdell P F, Evans J R, Edirisinghe M J, Shaw P, and Binstead M J (1995), `The computer aided manufacture of ceramics using multilayer jet printing', J. Mater. Sci. Lett., 14(22), 1562±1565. Bussmann M, Mostaghimi J, and Chandra S (1999), `On a three-dimensional volume tracking model of droplet impact', Phys. Fluids, 11(6), 1406±1417.

Drop formation and impaction

67

Carr W W, Park H, and Morris J F (2004), `Textile inkjet: drop formation and surface interaction', National Textile Center Annual Report, National Textile Center, Wilmington, DE, Project C02-GT07. Chandra S and Avedisian C T (1991), `On the collision of a droplet with a solid surface', Proc. R. Soc. Lond., 432, 13±41. Chen A U, Notz P K, and Basaran O A (2002), `Computational and experimental analysis of pinch-off and scaling', Phys. Rev. Lett., 88(17), article 174501 (4 pages). Christanti Y and Walker L M (2001), `Surface tension driven jet break up of strainhardening polymer solutions', J. Non-Newtonian Fluid Mech., 100(1±3), 9±26. Clanet C and Lasheras J C (1999), `Transition from dripping to jetting', J. Fluid Mech., 383, 307±326. Clarke A, Blake T D, Carruthers K, and Woodward A (2002), `Spreading and imbibition of liquid droplets on porous surfaces', Langmuir, 18(8), 2980±2984. Eggers J (1997), `Nonlinear dynamics and breakup of free-surface flows', Rev. Mod. Phys., 69(3), 865±929. Engel O G (1955), `Waterdrop collisions with solid surfaces', J. Res. Natn. Bur. Stand., 54(5), 281±298. Ford R E and Furmidge C G L (1967), `Impact and spreading of spray drops on foliar surfaces', in Wetting. Soc. Chem. Industry Monograph, London, Society of Chemical Industry, 417±432. Fukai J, Zhao Z, Poulikakos D, Megaridis C M, and Zhao Z (1993), `Modeling of the deformation of a liquid droplet impinging upon a flat surface', Phys. Fluids, 5(11), 2588±2599. Fukai J, Shiiba Y, Yamamoto T, Miyataka O, Poulikakos D, Megaridis C M, and Zhao Z (1995), `Wetting effects on the spreading of a liquid droplet colliding with a flat surface: experiment and modeling', Phys. Fluids, 7(2), 236±247. Fukai J, Tanaka M, and Miyatake O (1998), `Maximum spreading of liquid droplets colliding with flat surfaces', J. Chem. Eng. Japan, 31(3), 456±462. Furbank R J and Morris J F (2004), `An experimental study of particle effects on drop formation', Phys. Fluids, 16(5), 1777±1790. Furbank R J and Morris J F (2006), `Pendant drop thread dynamics of particle-laden liquids', to appear in Int. J. Multiphase Flow. Goldin M, Yerushalmi J, Pfeffer R, and Shinnar R (1969), `Breakup of a laminar capillary jet of a viscoelastic fluid', J. Fluid Mech., 38, 689±711. Harlow F H and Shannon J P (1967), `The splash of a liquid drop', J. Appl. Phys., 38(10), 3855±3866. Josserand C, Lemoyne L, Troeger R, and Zaleski S (2005), `Droplet impact on a dry surface: triggering the splash with a small obstacle', J. Fluid Mech., 524, 47±56. Mao T, Kuhn D C S, and Tran H (1997), `Spread and rebound of liquid droplets upon impact on flat surfaces', AIChE J., 43(9), 2169±2179. Morris J F and Boulay F (1999), `Curvilinear flows of noncolloidal suspensions: the role of normal stresses', J. Rheol., 43(5), 1213±1237. Ogg J C, and Schetz J A (1985), `Breakup and droplet formation of slurry jets', AIAA J., 23, 432±439. Ok H, Park H, Carr W W, Morris J F, and Zhu J (2004), `Particle-laden drop impacting on solid surfaces', J. Dispersion Sci. Tech., 25(4), 449±456. Park H (2003), Drop impingement and interaction with a solid surface, Georgia Institute of Technology, Atlanta, GA.

68

Digital printing of textiles

Park H, Carr W W, Zhu J, and Morris J F (2003), `Single drop impaction on a solid surface', AIChE J., 49(10), 2461±2471. Pasandideh-Fard M, Qiao Y M, Chandra S, and Mostaghimi J (1996), `Capillary effects during droplet impact on a solid surface', Phys. Fluids, 8(3), 650±659. Range K and Feuillebois F (1998), `Influence of surface roughness on liquid drop impact', J. Colloid Interface Sci., 203, 16±30. Scheller B L and Bousfied D W (1995), `Newtonian drop impact with solid surface', AIChE J., 41(6), 1357±1367. SÏikalo SÏ, Marengo M, Tropea C, and GaniÂc E N (2002), `Analysis of impact of droplet on horizontal surfaces', Experimental Thermal and Fluid Science, 25, 503±510. Stow C D and Hadfield M G (1981), `An experimental investigation of fluid flow resulting from the impact of a water drop with an unyielding dry surface', Proc. R. Soc. Lond. Series A, Mathematical and Physical Sciences, 373(1755), 419±441. Tay B Y and Edirisinghe M J (2001), `Investigation of some phenomena occurring during continuous ink-jet printing of ceramics', J. Mater. Res., 16(2), 373±384. Tincher W C, Hu Q, and Li X (1998), `Ink jet systems for printing fabric', Textile Chemist and Colorist & American Dyestuff Reporter, 30(5), 24±27. Windle J and Derby B (1999), `Ink jet printing of PZT aqueous ceramic suspensions', J. Mater. Sci. Lett., 18(2), 87±90. Worthington A M (1876), `On the forms assumed by drops of liquids falling vertically on a horizontal plate', Proc. R. Soc. Lond., 25, 261±272.

5 Industrial production printers ± DuPont ArtistriTM 2020 textile printing system M R A Y M O N D , DuPont Ink Jet, USA

5.1

Introduction

The current worldwide production of printed textile fabric is over 34 billion square meters per year and is dominated by rotary screen-printing. Digital printing for textiles has a compelling value proposition, which could be leveraged into a variety of related businesses. The worldwide opportunity in digital textile printing solutions could range from $4 billion to $6 billion within the next five years. Each 1% adoption from traditional textile printing to digital creates a potential for 3 million liters of pigment ink, 1.3 million liters of reactive dye ink, 800,000 liters of disperse dye ink and 500,000 liters of acid dye ink. Technology is evolving and partnerships are being created to exploit inkjet textile printing opportunities. Textile ink, inkjet printhead, color management software, fabric handling equipment and fabric pre- and post-processing technologies have been developed to work together as an optimized system. The development of these new technologies is replacing traditional screen-printing techniques and is creating new opportunities and markets that coexist with existing technology. To meet these opportunities, DuPont offers the ArtistriTM 2020 digital printing system to the market. The printing system, developed through a partnership between DuPont, Ichinose Toshin Kogyo and Seiko Printek, meets the production level requirements for short-run textile printing. The printer developed by Ichinose Toshin Kogyo of Japan utilizes Seiko Printek inkjet printheads. DuPont supplies ink and color management technology and markets the system to end-users.

5.2

Industry needs

The run lengths of textile print jobs have decreased dramatically. Designers are providing more options and retailers are demanding more product choices and fewer inventories. A quick restocking of a popular design can increase profitability. As run lengths decrease, the cost of traditional screen-printing rises. Digital printing can be cost-effective against screen printing for shorter run

70

Digital printing of textiles

5.1 The difference in cost between screen and digital printing setup processes.

lengths. This is because the cost of engraving screens and setup must be amortized over the length of the print run. See Fig. 5.1 to see how screen and digital printing setup processes differ. The first and foremost need from the industry is quality. Quality of the image and quality in the fastness characteristics (i.e. rub, wash and light fastness) are required. DuPont ArtistriTM inks have been designed and tested to meet or exceed the industry standards for fastness. The DuPont ArtistriTM technology provides high quality as demonstrated by the high-value scarves, ties, swimwear and other apparel being sold to the public at retail. Textile inkjet printing does not yet come close to the printing speeds possible from a rotary screen process. The industry needs higher speed digital printers. Commercially available digital textile printers operate in the range of 2 to 150 square meters per hour, whereas a rotary screen process can easily reach speeds over 1000 square meters per hour. Inkjet printing technologies are constantly improving to meet future demands. Nevertheless, they are still far away from achieving rotary screen speeds at reasonable reliability and cost.

5.3

Markets and applications

Each of the many textile-printing markets has their own requirements for image quality, color and fastness characteristics. Image quality is more important for fine silks, and color quality is extremely important for swimwear and team-wear. All printed fabrics need to be rub fast, wash fast and light fast to varying degrees depending on the market. See Fig. 5.2.

DuPont ArtistriTM 2020 textile printing system

71

5.2 Textile-printing market requirements from digital printing technology.

The silk accessories market is amenable to digital printing due to the inherent short run lengths and relatively high value to the customer. The quality of the image is extremely important for applications such as ties and scarves. Fine line printing is very important for design patterns. Extra care is taken by the owner with these fabrics so fastness characteristics are secondary to image quality. The market for digitally printed silk ties and accessories could be over $300 million in the next five years. The swimwear market is a good target for digital printing due to the short run lengths and many design cycles. Fastness characteristics are more important than image quality. Color quality is high on the requirements and spot colors are used for fluorescent and metallic colors. The inks have to meet the fastness requirements for chlorine and salt water exposure and, of course, light fastness is very important as well. The market for digitally printed swimwear could be over $250 million in the next five years. Home furnishings can be bed coverings, window treatments, upholstery, etc. The home furnishings market is more cost sensitive than silk or swimwear. Wash fastness is a high priority because bedding must stand up to numerous washings. The printing run lengths tend to be longer than those found in swimwear and silk and a lot of printing in this segment is on wide fabrics up to 3.2 meters. The market for digitally printed home furnishings could be over $400 million in the next five years. Apparel, teamwear and T-shirts comprise a very large market segment. Apparel varies widely in image quality and fastness characteristics. Cotton

72

Digital printing of textiles

fabric is the most widely used. Wool, nylons, and polyester are also used. The market for digitally printed apparel could be over $500 million in the next five years. Soft signage applications fall into these categories: point-of-purchase signage, trade-show signs and wraps, banners, flags and backdrops. Almost all fabric types are used for soft signage, such as silk, cotton, linen, nylon, rayon and polyester. The market for inkjet printed soft signage is expected to reach $713 million in 2006 (IT Strategies). The average retail selling price ranges from $7 to $15 per square foot ($75 to $161 per square meter).

5.4

ArtistriTM 2020 printer

The ArtistriTM 2020 printing system is offered commercially to meet the needs of short-run textile printing. See Fig. 5.3. The 1.8-meter wide printer has eight color channels available for printing. Greater color range and quality can be achieved than in six-color printers. Usually, the colors offered are cyan (C), yellow (Y), magenta (M), black (K), light cyan (lc), light magenta (lm), C1 and C2. CYMK are the base colors for process color printing while light cyan and light magenta allow smooth tonals to be printed and offer a higher perceived resolution. C1 and C2 can be gamut expanding or used as spot colors depending upon the customer's needs. Printing resolutions are 360, 540 or 720 dots per inch. Print speeds are from 15 to 60 square meters per hour depending upon resolution and other parameter settings. See Table 5.1. The 2020 printer utilizes Seiko Printek piezo drop-on-demand printheads. The printheads and the printer were developed concurrently with the ArtistriTM Ink and the ArtistriTM Color Management System. All four technologies were optimized together to achieve the best print quality and machine performance. The printer was developed and is presently manufactured by Ichinose Toshin Kogyo in Japan.

5.3 The ArtistriTM 2020 Printing System.

DuPont ArtistriTM 2020 textile printing system

73

Table 5.1 Print speeds in relation to resolution Resolution

Print speed (square metres per hour) High speed (draft) Standard interlacing Highest quality

360 540 720

5.5

66 44 33

45 30 22

23 15 11

Competitive environment

There are three performance ranges of textile inkjet printers: low-speed (2±15 square meters per hour), mid (15±100 square meters per hour) and high (>100 square meters per hour). The prices of the printers are relative to their performance. See Fig. 5.4. Most low range printers use Epson printhead technology and Mimaki is the leader in this range with the TX-1, TX-2 and TX-3. Other, Epson printheadbased textile printers are offered by Mutoh and Roland. There are Konica printers in this range that have been sold primarily by Stork. Konica's printers use their own printhead technology, which is based on a Xaar license. These low-end textile printers are typically wide format printers modified for printing on fabric. The primary applications are sampling due to the low speed. Prices in this arena are under $100,000.

5.4 The price of printers in relation to their performance.

74

Digital printing of textiles

The DuPont ArtistriTM 2020 printer is in the mid-range category in both speed and cost. The Robustelli Monna LisaTM printer, based on Epson printheads, is also in this range. Konica-Minolta has recently launched a new printer, the Nassenger V, using their own printhead technology. Printers in the midrange category are used for sampling and short run production. Prices for printers in this range vary from $185,000 for the ArtistriTM 2020 to approximately $250,000 for the Robustelli Monna Lisa. At the high end is the Reggiani DReAM printer. Performance is up to 150 square meters per hour. The DReAM utilizes Aprion (Scitex Vision) printhead technology. It is capable of printing up to six process colors. The primary application is short run production. The DReAM printer price is over $500,000.

5.6

ArtistriTM 2020 textile printing technology

The ArtistriTM technology employed in the 2020 printing system is a triumvirate of ink development, color management and the printer/printhead system. All three areas of technology affect the output quality and performance of the printer and have been developed concurrently, to ensure that the DuPont ArtistriTM 2020 provides a total, optimized, turnkey solution for DuPont's customers. Customers do not have to shop for a printer, determine where to get ink or obtain third party software for color management. Peak productivity is achieved months earlier than would otherwise be possible.

5.6.1 Ink, pretreatment and post-treatment There are four ink types to cover the full range of fabrics and applications: acid dye, reactive dye, disperse dye and pigment. DuPont has developed these inks to meet production level requirements for printed fabrics. Pretreatment and posttreatment processes are required for dye-based inks. Pigment inks do not require fabric pretreatment. See Table 5.2. All inks have been optimized to perform with the inkjet printhead technology. Acid dye inks are generally used for printing on nylon, wool and silk fabrics. Pretreatment of the fabric prior to printing is required to provide fixation between the dye and the fibers. A paste solution is applied uniformly to the material that provides an acid donor for fixation of the dye. Steaming at 100± 102ëC for 30±60 minutes is required after printing and drying, which redissolves the dyes and swells the fibers, thus fixing the dye to the fabric. Then the fabric must be washed to remove any unfixed dye. For example, acid dye inks are used for silk ties, scarves and LycraTM swimsuits. Disperse dye inks are formulated primarily for polyester material. Pretreatment of the fabric is required by applying a solution that will enhance the fixing process during steaming. Fabrics should be post-processed with high temperature steam under pressure at ~160ëC for 20±30 minutes. Dry heat at

DuPont ArtistriTM 2020 textile printing system

75

Table 5.2 Ink type, fiber type and required pre- and post-treatments Ink type

Fiber types

Pretreatment

Post-treatment

Acid dye Disperse dye

Silk, nylon, wool Polyester

Acid donor Thickener

Reactive dye Pigment

Cotton, rayon Cotton, polyester, blends

Alkali Not required

Steam and wash High temperature steam and wash Steam and wash Dry heat

~160ëC can also be used to fix the disperse dyes, but deeper colors are achieved by steaming. Washing is required to remove any unfixed dye. There are many applications for disperse dye on polyester such as soft signage, flags and banners. Reactive dye inks are used on cotton, wool and rayon fabrics. Reactive dye colors are brighter and deeper than other inks. Pretreatment of the fabric similar to acid dye ink is required. Steaming at 100±102ëC for 10±20 minutes is required to fix the dye to the fabric. Washing removes any unfixed dye. Reactive dye inks are mostly used for sampling due to the larger color gamut they offer. Pigment inks are applied to cotton, cotton±polyester blends and almost all other fabrics. The pigments are in water dispersion with a binder that attaches the particles to the fabric. Pigment printing is the most economical process because pretreatment and washing are not required. Thermal fixation via a calender or fabric oven is needed to obtain the best fastness characteristics. Pigment inks have excellent fastness to light and very good rub and wash fastness. The handle (softness of feel) of a fabric printed with pigment ink can sometimes be hard. About half of the world's fabric printing is done with pigment ink. All inks for the 2020 are supplied in one- or two-liter cartridges. A plastic cartridge surrounds and protects a bag inside where the ink resides. Prior to filling the bag, the ink goes through a degassing process. Dissolved air in the ink is removed to improve the jetting performance of the printhead. Dissolved air in the ink can escape during the printhead's jetting process (this is called rectified diffusion) and cause it to misfire or misdirect the drop. The problem is similar with all piezo inkjet technologies and, if ink is not degassed, the printhead will have to be operated at lower firing frequencies, negatively affecting productivity of the printer. The bag is designed to prevent any air ingestion into the ink. There are four laminated layers constructing the bag: polyethylene, nylon, aluminum and polyester. The aluminum layer is the primary barrier to keep air from reaching the ink. The other layers are for structural support and impact resistance. The cartridge also provides a clean process. The operator is never exposed to ink and, because it is self-closing, there are no spills. An electronic chip

76

Digital printing of textiles

contains data about the ink, such as type, color, level and batch. The chip is read by the printer and the information can be accessed by the operator. If the operator puts the wrong cartridge in a slot, the operator will be notified with an error. When a cartridge is empty the operator is notified and it can be replaced without stopping the printing process, improving productivity. The cartridge also facilitates ease of ink changeover. A flushing cartridge is used to flush ink from the lines and printheads. Then a different ink chemistry or color can be used.

5.6.2 Printhead The printheads utilized in the 2020 printer are developed and manufactured by Seiko Printek (SPT), Japan. See Fig. 5.5. They employ piezo shared-wall technology and are manufactured under a license from Xaar, Cambridge, UK. SPT developed the printhead to meet DuPont's requirements for aqueous based textile inks. In shared-wall technology, the piezo walls of the chamber are squeezed to eject a drop through a nozzle. A chamber shares its walls with its neighboring chamber. When a nozzle is firing, the adjacent channels cannot be fired. The printhead for the 2020 operates using the same mechanism, but every other channel is not used. Therefore, it is technically not a shared-wall configuration. Each nozzle can be fired without affecting its neighbor. The electrodes that electrically activate the piezo in the printhead are located in the chamber in contact with the ink and every other chamber has the opposite polarity. Making every other channel a `dummy' channel is required because the textile aqueous based inks are conductive. The ink would short the electrodes if operated in a shared-wall configuration. Of course, there are half the nozzles available for printing, but they operate at higher operating frequencies with less cross-talk between channels. Each printhead in the 2020 printer has 255 active nozzles and each nozzle delivers a 35 picoliter (35  10ÿ12 liter) drop at up to 20,000 drops per second. There are two identical printhead carriages on the 2020. Each carriage has eight printheads, one for each color. Carriage speeds are 20, 40 or 60 inches per second with 40 being the nominal operating point. Each of the printhead carriages adjusts up to 10 millimeters from the belt to accommodate thicker fabrics. An alignment procedure must be performed during installation or whenever a printhead is replaced. Printhead to printhead alignment must be done for each carriage. Then, a carriage to carriage alignment procedure is performed. Printhead replacement and alignment should take less than one hour. The printhead is designed for industrial use and has a very long life. Printhead failures most often occur from nozzle contamination or other external harm. The Seiko printheads are very robust and warranted to over 4 billion drops per nozzle. However, printhead life is expected to greatly exceed that.

DuPont ArtistriTM 2020 textile printing system

77

5.5 The printheads utilized in the 2020 printer.

5.6.3 Color management Color management is the technology related to the correct interpretation and rendering of color information. Matching colors from design to the digital printer is critical to the customer. It is also important that colors are matched to the screen printing process. Screen printers will visually look at the printed textile sample and mix different base colors of ink to obtain the final color. This is done in a `color kitchen' similar to mixing paint colors. A typical rotary screen printer can have

78

Digital printing of textiles

from 1 to 12 spot colors and from 1 to 12 screens. Each color is mixed individually and is called a colorway. To change a colorway, any one or more of the spot colors change, but will print the same pattern. The color gamut is very large due to the relatively high number of base colors available. In a digital textile printer the colors are mixed or dithered directly onto the fabric, and almost all systems use the four base colors CYMK. Some systems, such as the 2020, use up to eight colors. The color information is in a digital file and must be converted to the correct color by mixing or dithering the colors available in the printer. The most common color data format for printers is L*a*b*. L* is the lightness ranging from 0 (dark) to 100 (light). The a* value defines the colors of a red±green axis and b* defines the yellow±blue axis. Using a spectrophotometer to measure the colors in L*a*b* space, a printer's color gamut can be determined and a lookup table can be created. The lookup table tells the printer what colors to mix or dither to create the required color. This is measured against a color standard such as a CIELAB reference. The user can print a `color book' for a visual representation of the color gamut. One of the most difficult challenges is to match the process color of the printer to the spot color of the screen system. Digital printers are not economical for printing long run lengths. After sampling and short runs are complete, a large production run may need to be done on a rotary screen printer. If the color gamut of the digital printer is outside the available spot colors or vice versa, a color match may be difficult to achieve. The DuPont ArtistriTM Color Management System has been optimized to deal with these issues.

5.6.4 Fabric handling The ArtistriTM 2020 printer receives fabric from a roll on the input and rewinds the material back onto a roll when complete (roll to roll). The fabric must enter the printer without being stretched or otherwise dimensionally unstable. The printer must advance the fabric beneath the printheads after each pass of the printer's carriage. Accuracy is critical to achieving good print quality by reducing the probability for dot placement errors. The motion control system for movement of the belt and fabric must be accurate to the sub-pixel level. At 540 dots per inch, position accuracy needs to be better than 12 microns. The fabric position is controlled by a belt that has a sticky surface. This is technically called an adhesive print blanket but can also be called a `sticky belt'. It keeps the fabric dimensionally stable during printing. As the fabric is placed on the belt it cannot be stretched or have wrinkles. A mechanism unwinds the fabric roll and keeps as little tension as possible on the fabric. This is especially important when printing stretchy fabrics such as LycraTM (which is used for swimsuits and other sportswear) otherwise the image could be distorted after coming off the printer. The feed mechanism gently lays the fabric on the sticky

DuPont ArtistriTM 2020 textile printing system

79

belt after it goes over a de-wrinkling roller. The fabric is then pressed to the surface of the belt by a pressure roller prior to entering the print area. After being printed, the fabric is dried by passing over a heated platen. The heater dries the fabric so it can be rewound onto a roll without any ink transfer from the image. It is not intended to cure the ink. The fabric is then rewound onto a roll, that is tension controlled to reduce stretching. The adhesive print blanket is cleaned by a brush with water to remove residual ink and fibers and dried with a squeegee prior to receiving more fabric from the input roll.

5.7

Process color printing versus spot color printing

Textile screen printing is primarily a spot color process and most inkjet printers, such as the 2020, utilize process color. These two approaches differ in that the colorants used to color the textile are premixed in the case of spot color, and mixed on the fabric in the case of process color. Process color printing is generally composed of black, cyan, magenta and yellow inks that are mixed in varying proportions by jetting droplets onto the fabric to create a variety of colors. The color gamut achievable by mixing only four colors of the same chemistry is far less than the colors obtainable in spot color. Spot color printing uses a set of `mother' colors numbering between four and 12. In an attempt to produce more colors with process color printing, either dilute four-color process inks or up to eight different colors are used. While this provides improvements, it still does not reach the combination of color correctness and functionality of spot color printing in color-critical applications.

5.8

Cost of printing

One of the big advantages of digital textile printing is the ability to print immediately from a digital file without any setup. The other is the lower cost of printing for short runs. The cost per square meter for digital printing is relatively flat and doesn't change much with volume. With screen-printing, the setup required and the cost of the screens must be amortized over the length of the print run. When a printing job is above 1000 square meters, it is most likely economical to run the job on a screen printer. Below that, digital printing can be cost effective. See Fig. 5.1. The cost of printing of the 2020 consists of the cost of the machine plus the cost of ink, and possibly the cost of a service contract if the customer desires one. A customer may want to include cost of fabric, labor costs and overhead in the cost of printing, but with all else being equal we are discussing only the costs associated with the printer. Ink printed onto the fabric will usually have between 40% and 150% coverage depending on the pattern. At 100% coverage there is approximately

80

Digital printing of textiles Table 5.3 Total cost of printing with the 2020 Ink cost (per liter) Coverage Printed ink per square meter (mL) Waste ink per square meter (mL) Total ink per square meter (mL) Ink cost per square meter

$150 40% 7.0 2.6 9.6 $1.44

$150 100% 17.5 2.6 20.1 $3.02

$150 150% 26.3 2.6 28.9 $4.33

Machine cost $185,000 $185,000 $185,000 Number of months 60 60 60 Machine cost per month $3,083 $3,083 $3,083 Machine speed (square metres per hour) 25 25 25 Hours per shift 7 7 7 Shifts per day 2 2 2 Days per month 22 22 22 Square meters per month 7700 7700 7700 Machine cost per square meter $0.40 $0.40 $0.40 Total cost per square meter

$1.84

$3.42

$4.73

17.5 ml of ink per square meter. Therefore ink coverage can be expected to be from 7 ml to 26.3 ml per square meter, more or less. Waste ink must also be calculated into the equation. Ink used for priming and cleaning the printheads is considered waste ink. If the printer has been idle for a length of time, a priming operation may be necessary prior to starting a new print job. Periodically during printing, the machine will pause and move the printheads to a maintenance station. Some ink will be purged through the printhead to clean the nozzles and then the nozzle face is wiped of any excess ink before returning to the printing operation. The period for maintenance operations is set by the operator. For this analysis the waste ink is considered to be 15% for 100% coverage. Therefore, the amount of waste ink is 2.6 ml per square meter. Total ink usage will be from 9.6 ml to 28.9 ml per square meter. If ink costs $150 per liter, this equates to $1.44 to $4.33 per square meter. If the machine costs $185,000 and is amortized over five years, the cost will be about $0.40 per square meter, operating two work shifts per day. The total cost of printing for the 2020 will be approximately $1.84 to $4.73 per square meter. Again, this does not include labor and overhead expenses. See Table 5.3.

5.9

Opportunities and new markets

Digital inkjet printing of textiles opens doors to new opportunities and creates new markets. Creative designs can be digitally printed that cannot be screenprinted. The largest screen printers have no more than 12 screens, which equates to a limitation of 12 spot colors. With process color there can be an almost

DuPont ArtistriTM 2020 textile printing system

81

unlimited number of colors in a design, allowing much more than 12 colors in a specific design. Design cycle times are reduced and sample production can be done immediately. The ability to do economical short runs allows reductions in the size of inventories. Restocking of a `hot' apparel item is made easy by digital printing and the store doesn't have to discount its prices. Today's markets are changing faster and customers are becoming more demanding than ever. Digital textile printing allows the production of goods and services to match individual customers' needs. Personalization allows people to be unique. Customized auto upholstery, a room's upholstery, wall coverings and window treatments can all be decorated with an exclusive design. Gaming table covers in a casino can be printed with designs of a specific convention or trade show. Pool table covers can have a targeted advertising message. The ideas and opportunities are growing rapidly.

5.10 ArtistriTM Technology Center The ArtistriTM Technology Center (ATC) is located in Wilmington, Delaware, USA, at the DuPont Experimental Station. The ATC supports customer demonstrations, sampling, training and technical development. There are over 100 personnel supporting the center including ink chemists, color experts, applications engineers and technicians. The ATC also has the resources of DuPont's Experimental Station readily available, including research, engineering services, metalworking and materials support. The ArtistriTM Technology Center supports the sales effort by providing demonstrations and printing samples for customers. The normal course of a sales cycle starts with providing printed fabric samples to a customer. A customer's application can be pretreated, printed, post-treated and returned very quickly from a number of the ATC's printing systems. Customers are also invited to visit the technology center for personal demonstrations. The ArtistriTM Technology Center also supports customers after the sale. Further education and training after installation is provided in a classroom setting and a hands-on learning environment. The customer can also work with a team of experts to solve new application problems or create color profiles. Training for maintenance of the printing systems is provided if the customer wants to service their own equipment. Product improvements and developments are ongoing at the ATC. The type of developments include new ink chemistry, color science and calibration, pretreatment, post-treatment, raster image processing and workflow. All product improvements are tested thoroughly at the ATC before customers get newly developed ink or upgrades to their printing system.

82

Digital printing of textiles

5.11 Applications support, technical service and training When a customer purchases an ArtistriTM 2020 textile printer, they are fully supported by DuPont's technical service and applications organization. Support is provided worldwide and is centrally organized into three regions: the Americas; Europe, Africa and Middle East; and Asia and Australia. The goals of the service group are to get customers productive as soon as possible and to provide support to maintain productivity. When a customer buys an ArtistriTM printing system, a technical service team will install the printer and train the users on the proper operation and maintenance of the machine. They are taught how to calibrate the printer and handle various types of fabric. The operators learn the proper handling and maintenance of the piezo inkjet printheads and the ink supply system. An applications specialist then trains the users on workflow, raster image processing and color management. They learn how to move jobs from the designer or other source, to the printer for processing. If wanted by the customer, the specialist teaches how to create color profiles and match them to the correct job. After the file is RIPped (Raster Image Process) it can then be queued for printing.

5.12 Future trends Customers are always asking for faster, better and cheaper. All three areas will progress as the nascent digital textile market grows from the present adoption phase. Print quality and color gamut will improve, machines will get faster and printing costs will come down. The speed of an inkjet printer is highly dependent on the number and operating frequency of the printhead nozzles. Productivity can be measured by the rate at which ink can be applied, so more nozzles and fast jetting can produce higher speed printing. The present limitations to speed are primarily the number of printheads you can economically design into a printer. At present, the cost per nozzle is prohibitive to getting to speeds over 200 square meters per hour at under $1,000,000 machine cost. As manufacturing techniques improve the cost per nozzle will come down. Enhancements in image quality will be achieved by smaller drop sizes. Gray scale capability (rather than present single drop size technology) will improve tonals and smooth gradations. Expanded color gamut will be achieved by better dye and pigment ink technology. The cost of printing is primarily driven by the price of ink. The increasing penetration of printers into the market is being supported by higher ink production. As ink volumes increase, the costs decrease, allowing reductions in pricing. As the cost of printing is reduced, the market will grow faster.

DuPont ArtistriTM 2020 textile printing system

83

Digital textile printing will not replace screen-printing in the near future. The two technologies will coexist. Software and ink technology will improve to provide a seamless workflow from digital inkjet to screen-printing. Ink type changeover in an inkjet printer is time consuming and wasteful. It would be wonderful if there was one ink type that printed well on all fabrics. Presently four ink types are required to cover the range of fabrics but only one can be in a printer at a time. Research and development resources are working to discover the `universal' ink.

5.13 Sources of further information and advice DuPont Inkjet: http://www.inkjet.dupont.com/ Inteletex: http://www.inteletex.com ± World Textile Publications Ltd. Reggiani DReAM Textile Printer: http://www.cibasc.com/index/ind-index/indtxt_fib/ind-tex-textile_processing/ind-tex-inkjet_tex/ind-txt-ink-dream_ industrial_printing.htm Techexchange.com: http://www.techexchange.com/textile-printing.html

5.14 Bibliography F Cost, Pocket Guide to Digital Printing, Delmar, 1997. L W C Miles, Textile Printing, Society of Dyers and Colourists, 1994. S F Pond, Inkjet Technology and Product Development Strategies, Torrey Pines Research, 2000.

6 Industrial production printers ± DReAM L C A C C I A and M N E S P E C A , Reggiani Macchine S.p.A., Italy

6.1

The DReAM project in the present textile printing scenery

Inkjet printing of textiles has been developing in the last few years but has not brought about a rapid development of supporting technologies, because of limited practical success with inkjet printers, their experimental nature and their high costs. For some years most newly designed inkjet machines have been just a mere adaptation of the best graphic plotters available on the market to meet the latest textile printing needs. These machines are nevertheless only printing machines originally designed for paper printing. The maximum growth is actually being reached in new markets of textiles for home furnishing, fashion, advertising and automotive, where new commercial opportunities are offered by innovative applications, trying to satisfy customers who demand more and more personalized products. The industry of traditional textile printing, which is probably not yet ready for these big changes, is now facing a dilemma in entering this new market. On the other hand, the companies that have specialized in digital printing, often only recently established, come from completely different fields. Nonetheless, moving to this new technology could actually become crucial for the future survival of the textile industry in developed countries. The growing interest in textile printing solutions with the development of inkjet printer technology is based mainly on several factors (notwithstanding some limits such as speed, adaptability to different weaves, etc.) that are definitely more innovative than those for other conventional printing systems. These positive factors are rapid transformation from paper drawings into printed patterns, easy selection of colours, complete elimination of screen-engraving processes, a print cost almost independent of the volume of production, extreme reduction in the number of operations required, low environmental pollution, and small installation footprint. It has to be noted that the current developments in the field of inkjet inks have enlarged the application scope of inkjet systems to almost all fibres, pure or blended.

Industrial production printers ± DReAM

85

Nevertheless, most inkjet printing machines still face severe limitations, such as slow printing speeds (generally strictly related to the fabric width); problematic alignment between the textile surfaces and the printing nozzles; difficult application on elastic or knitted fabrics, or on relief or flat weaves; insufficient colour homogeneity on large surfaces; sparse or poorly controllable penetration of the dye; and inaccurate reproducibility of printing samples produced on rotary or flat printing machines. Notwithstanding the abovementioned drawbacks, interesting results (considered sufficient and economically interesting for certain niche applications) have been achieved. Some textile machinery manufacturers have committed themselves to quickly overcome the current shortcomings. From a global point of view, the technology available up to now has not provided the characteristics needed for a suitable printing process able to satisfy industrial application requirements. Until now, digital printing of fabrics has in fact been confined to extremely short runs of 10±20 square metres, producing strike-offs and samples of new designs, and occasionally small production runs of up to 50 square metres. For anything approaching production volumes, banks of these printers must work in parallel. This approach is dated and impractical; the other and more innovative one is simply to use a faster machine. Even so, there are many hurdles to be overcome to reach this target, because in order to manufacture machines suitable for the textile printing process it is necessary to select the right printing system among a wide range of inkjet technologies, from the head technology to the inks available. Alternatively, the software applications used to drive this technology from a digital file standpoint should be developed with specialized and innovative technology specifically designed for textiles. This complex task can be solved only through strict cooperation and synergistic interaction between specialists from three different sectors: textile machinery, information technology and colouring. For this reason, a collaboration has been established between three leading international technology companies: Reggiani, providing extensive knowledge of textile handling and print machine manufacturing; Scitex Vision, which has wide experience of digital printing on different substrates, providing the print engine using its proprietary Aprion technology, unique piezoelectric drop-ondemand print heads for the DReAM printer; and Ciba Specialty Chemicals, with its wide expertise and long-standing experience in chemicals and the pioneering role it has played in inkjet printing, which has developed suitable inks especially designed for the Aprion print heads, thus ensuring accurate performance. A few words need to be said about the three companies which have cooperated for the development of the whole system: · Reggiani Macchine is a leading manufacturer of traditional printing machines for textiles (rotary and flat screen-printing machines for clothing and household fabrics) as well as conveyance and control systems for fabric

86

Digital printing of textiles

feeding. It was originally founded as a supplier of specialist textile enhancement services, and later expanded into developing its own web-fed machines for this purpose. Reggiani, which has been synonymous with innovation and quality worldwide for more than 50 years, is regarded as the supplier of reference and considered by customers for the high standard and reliability of its printing machines and for its focus on today's market needs and trends. The company now has a worldwide base of over 1000 customers. Located in Grassobbio, Bergamo, 40 km from Milan, with a staff of 190 people, Reggiani has been certified to ISO Standard 9001 since 1995 and to UNI EN ISO 9001 since 2000. · Scitex Vision is a leading developer, manufacturer and service provider of cutting-edge digital printing presses and consumables for industrial applications including ultra-wide-format graphic arts, packaging and textiles. Backed by global marketing and support networks, Scitex Vision is committed to continuously providing high-quality, flexible and cost-effective solutions to printing houses all over the world. The company owns a core technology based on Aprion's patented drop-on-demand piezo-inkjet print heads and water-based inks. Scitex Vision employs more than 460 people worldwide with headquarters located in Netanya, Israel, and subsidiaries in Atlanta, Hong Kong and Brussels. · Ciba Specialty Chemicals (SWX: CIBN, NYSE: CSB) is a leading company dedicated to producing high-value effects for its customers' products. It strives to be the partner of choice for its customers, offering them innovative products and one-stop expert service. Ciba creates effects that improve the quality of life ± adding performance, protection, colour and strength to textiles, plastics, paper, automobiles, buildings, home and personal care products and much more. Ciba Specialty Chemicals is active in more than 120 countries around the world and is committed to be a leader in its chosen markets. In 2004, the company generated sales of 7 billion Swiss francs and invested 288 million SFr in R&D. Years of hard work and commitment have led to the creation of a new machine, the DReAM: this is not an adaptation of a wide-format graphics printer but a new generation of inkjet printing machine specifically designed for textile printing processes. It is supplied in a standard roll-to-roll configuration; the transport mechanism can, however, be tailored to meet special customer requirements. Many of the innovations featured by DReAM are really unique and ground breaking, such as the following. · The machine can print on virtually any flat surface (from textile to leather) thanks to its adjustable-height printing heads. · Its output rate exceeds 150 square metres per hour (DReAM 160) and 190 square metres per hour (DReAM 220). · The fabric width can reach 1600 mm and 2200 mm.

Industrial production printers ± DReAM

87

· It can easily print on elastic and knitted fabrics, since the fabrics are not retained during the printing process but are permanently bonded to a nondeformable blanket. · It washes the blanket continuously with optimized washing intensity. · It allows direct and instantaneous change of design pattern. · It uses high-level inks produced by Ciba Specialty Chemicals and adapted to the needs of Scitex inkjet heads: ± reactive inks for cellulose fabrics ± acid inks for silk and polyamide/Lycra blends ± disperse inks for polyester inkjet printing by transfer or direct, and for fashion and high light fastness applications ± pigment inks for all fabrics. · It dries the fabric in-line after printing with adjustable temperature, and polymerizes the bonding agent. · It reduces ink consumption remarkably in comparison to similar applications. · It allows continuous ink feeding during the printing process. · It applies piezoelectric Scitex Vision Aprion heads featuring the `drop-ondemand' technology. · It utilizes seven printing heads for each of the six colours available (42 printing heads on one bridge). · It prints with 600 dpi resolution (real). · It grants maximum reliability, stability and reproducibility. · It is equipped with a software program allowing perfect matching between digital and conventional prints. · It integrates with all graphics software. All these features combined deliver characteristics superior to other inkjet printing systems: high productivity; very short response time to market requirements; excellent cost-effectiveness for the production of small lots and samples; reduction of environmental pollution to a minimum; and reduction of the staff to a single operator.

6.2

Goals of the project and description of the DReAM machine (technical and technological parts: Reggiani, Ciba Specialty Chemicals and Scitex Vision)

6.2.1 Goals of the project There has been a dramatic change in global trends in textile manufacturing over the past decade. Average production runs in developed countries have been reduced to less than 1000 square metres per design. An industrial inkjet printing machine such as the Reggiani DReAM is ideal because it ensures the highquality and cost-effective production demanded by the market. The new digital

88

Digital printing of textiles

technology is in fact appropriate for applications requiring high-quality printing for short to medium production runs. It is suitable for printing top fashion apparel, including high-end fashion niche markets, as well as home furnishing applications, flags and banners, swimsuits and technical textiles as well as new applications. The advantages are numerous. First of all, the cost saving aspect should not be underestimated. Since there is no cost for preparing cylinders/screens, this digital printing system is cost-effective for runs up to 1000 metres. Storage space for the cylinders/screens is no longer required. Production time can be cut down, also. Preparing the cylinders or screens and colours for traditional printing is a time-consuming task and takes at least 3±5 weeks. This is compared with just a couple of minutes with the DReAM system, making it much easier to produce samples, short runs or even medium runs in a short time. The Reggiani DReAM is the first and so far unique digital printing system perfected to work on an industrial scale with high resolution (600 dpi). A new collection can be printed in just a couple of days. With current inkjet plotters, roughly 20 machines and several weeks are required to achieve the same production output. The Reggiani DReAM inkjet printer is effective from creation to production, and facilitates the production of highly differentiated, added-value printed fabrics. Moreover, according to DReAM partner Ciba Specialty Chemicals, inkjet printing is `clean, creative and competitive'. It is clean because all the colour goes onto the fabric and not into the waste water; the inks are liquid and thus non-dusting; water and energy consumption is low, and there is no cleaning of equipment between runs. It is creative ± as well as fast and flexible ± because not only can any design be printed, but also colours and designs can be changed with a click on the computer, making short runs efficient and cost-effective. It is competitive because it is highly cost-effective, with fully reproducible results ± `a short cut to customized and personalized production'.

6.2.2 The DReAM machine The DReAM machine is quite compact, occupying approximately 3 m  6 m (see Fig. 6.1). A slightly larger footprint is necessary for models with additional drying or polymerizing units. The standard configuration is a roll-to-roll system. However, Reggiani also provides alternative configurations to suit specific customer needs. The standard operating power is about 10 kW (excluding the power needed for the eventual auxiliary drying unit). Machine installation is quite simple and does not require special connections. The machine must be connected to the mains, water and drainage system (its operation needs only electricity, compressed air, de-ionized water and normal water, but in any case before installation Reggiani provides DReAM users with a detailed guide for site preparation). The recommended printing conditions are

Industrial production printers ± DReAM

89

6.1 The DReAM machine.

room temperature of 20±25ëC (68±75.2ëF) and 50±60% humidity. The DReAM therefore requires a ventilated, temperature-controlled environment. All the operations are programmed according to the desired printing sequence by means of an on-board computer connected with another remote PC, which transmits the digital file with all the instructions about the print, its colours and variants. Once the setup of all the variables has been carried out following the operator's instructions, the digital file is stored in the PC for possible repetition. The DReAM runs a Windows NT operating system. Concerning RIP and printing process, the DReAM is provided with an entrylevel RIP utilizing standard Photoshop software and a Pantone plug-in (Hexachrome), which has been designed specially for use with six-colour process printing. This RIP software is provided as a standard integral part of each delivery. It is very simple to use and takes advantage of Photoshop being an industry standard. The entry-level RIP package, including the Photoshop RIP with the Pantone Hexachrome plug-in, is included in the package. Other options for RIP software may also be purchased with the system if more complex functions are required. In any case, RIP software is a must with the system. To be more precise, the DReAM currently supports the following RIPs: Photoshop with Pantone plug-in (entry-level), Ergosoft, Wasatch, Hightex, NedGraphics and Aleph. Ergosoft, Wasatch, Hightex, Eidocolor and Aleph provide more complex features and capabilities which may be required for certain applications. In the event that the customer wishes to use alternative software, a special driver must be developed to interface between the chosen RIP software and the printer. It is here necessary to explain what exactly is the input for the printing system. There are in fact two types of files that can be entered into the DReAM workflow (into the RIP): · Combined colour information (either RGB or LAB) in TIFF or other graphical formats; or

90

Digital printing of textiles

· Separated colour information (suitable for traditional printing) in TIFF or other graphical formats. In this case each individual separation may be provided as a grey-scale TIFF. The operator may develop unlimited colourways via the RIP software or in the CAD software used to prepare the files, depending on the workflow that he or she implements. Before starting to print, the system has to be calibrated for each fabric±ink combination, and after calibration a colour book can be created using the `recipes' chosen by the user. A spectrophotometer allows (after calibration) sample colours to be measured, for colour matching. The measurement gives a recipe that can be used for the colourways. A certain number of colour calibration profiles (tables) are provided by Reggiani together with the RIP. In addition, the customer may develop and define additional colour profiles. (Instructions for creating additional colour profiles may be provided as part of the Reggiani training package.) Figure 6.2 shows the sequence of machine operation. A single application specialist can manage alone all the above-mentioned operations. He or she should have a working knowledge of pre-press software, colour management and printing processes and know how to use the RIP software for preparing files for printing. The pre-press person performs basic calibration procedures like producing linearization curves. An expert in this field can also perform colour calibration and create colour profiles. Training on the specific software provided as part of the DReAM is also provided by Reggiani. In addition to the application specialist who is recommended to prepare the files for printing, as is done today with computerized workflows (CAD/CAM operators), a single operator is sufficient to control the machine and oversee printing. This operator should have a technical background, know the machine components, and know how to operate the machine to print different jobs. Furthermore, the operator performs routine maintenance activities. The DReAM machine prints starting from standard rolls or from a single big roll. The inkjet printer (whose design is quite similar to conventional continuous printing machines) delivers the printed fabric in the form of a roll or a folded piece ready for subsequent fixing processes. Any fabrics in fact may be printed on the DReAM system. The appropriate pre-treatments have been developed by Ciba in order to maximize print quality for each fabric type together with the appropriate ink type. Moreover, the fabric should be treated before and after printing, in order to allow fixation of the colours as well as to improve the colour depth and visual quality. Reggiani Macchine S.p.A. and Ciba Specialty Chemicals Inc. will provide specific recommendations for the preparation, pre-treatment and finishing of the fabric. It is now interesting to examine some technical details of the DReAM inkjet printing system.

6.2 Schematic of the DReAM machine.

92

Digital printing of textiles

Entry to the machine On entering the machine, the fabric roll to be printed slides on two cylinders that unwind the fabric and automatically adjust the speed with respect to the preset tension level. A load cell system detects the tension, while a reciprocating compensating device transforms the constant motion of the unwound fabric into forward stepping motions. These movements are controlled by the blanket on which the fabric is bound by means of a permanent thermoplastic adhesive. The bonding power is the minimum required to retain the fabric (the high bonding performance required for flat screen printing is not necessary in this case). The printing phase The printing unit, which includes six colours, travels crosswise through the machine at a controllable speed. The quantity of ink sprayed by each head is adjusted by means of software and ranges between +200% and ÿ70%. In any case the ink consumption, based on CMYKOB reactive inks for medium coverage with medium-dark colours, is approximately 10±11 g/m2. The distance between the spraying nozzles and the blanket surface can be adjusted from 0 to 40 mm, thus granting excellent handling for any type of fabric, velvets, non-woven, leather, finished garments, etc. Each head is fed from both sides for more uniform spraying. The feeding system for each of the six colours is continuously recirculated into a system equipped with microfilters and gas exhausts, ensuring maximum efficiency of the spraying nozzles. On one side of the machine the ink feeding system is equipped with six tanks, each containing 10 kg of ink (one for each colour). Moreover, there is a buffer quantity of ink in the machine at all times (supporting approximately 30 minutes of printing) so that the empty tanks can be replaced while printing without hindering machine operation. The blanket The flat and non-deformable blanket is tensioned continuously and uniformly. The special stepper motors connected with encoders grant an accuracy of a few microns, as a result ensuring optimum fabric feeding and transport. After passing the return rollers, the external surface undergoes a powerful washing carried out by means of a large rotating brush sprayed with water and followed by a squeegee that removes the excess water. The intensity must be adjusted according to the degree of dirt on the surface and to the type of fabric. The maximum water consumption is approximately 300 litres per hour. The drying phase The printed fabric is separated from the blanket and conveyed into a drying unit operating according to the hot air impulse principle, which can be adapted to

Industrial production printers ± DReAM

93

different manufacturing requirements. The printed and dried fabric is wound on rolls (or folded according to the customer's requirements) which are then sent out to the subsequent development, fixing and washing stages depending on the type of dye applied. The inks: Ciba Specialty Chemicals The inks are developed and manufactured at Ciba Specialty Chemicals Inc. in Basel, Switzerland, and are sold and distributed via Ciba's international distribution structure. Ciba Specialty Chemicals has been a market leader in digital printing ink development and manufacturing for many years and is thus a reputed partner in the field. The chemistry and R&D teams of Ciba Specialty Chemicals and Scitex Vision have worked extensively together to develop Ciba's unique inks, which are designed exclusively for the Aprion inkjet heads, thus ensuring optimum results. Reggiani offers a package to the customer, a seamless integration of the DReAM printer with Aprion heads and Ciba inks. In addition, the one-year inclusive warranty covers only the use of Ciba Specialty Chemicals inks. Finally, using Ciba's inks ensures the overall quality of the printed results as well as reliability of the DReAM system. Today, the following reactive, acid, disperse and pigment inks are available: · CIBACRONÕ RAC reactive inks, specifically designed for cellulose fabrics (cottons, viscose, etc.) · CibaÕ LANASETÕ RAC acid inks, for silk, polyamide and wool fibres · CibaÕ TERASILÕ RAC disperse inks, suitable for polyester applications and transfer printing · CibaÕ TERASILÕ RAC TOP disperse inks, suitable for apparel and automotive polyester direct printing applications · CibaÕ IRGAPHORÕ RAC pigmented inks, suitable for all fabrics. On the DReAM machine six process colours are used: yellow, orange, magenta, blue, turquoise (cyan) and black (CMYKOB), allowing the widest colour gamut. Alternatively, the customer may choose to use four process colours together with light colours in order to maximize colour smoothness (CMYK and lights turquoise and magenta). If the difference between using spot colours and process colours is not clear, we can say that spot colours in traditional printing require that the specific, and exact, colours needed for each job and each colourway are mixed in advance in the colour kitchen. Subsequently, each colour is loaded onto its screen for printing. After printing, the screen needs to be cleaned and the new colour loaded. Digital printing uses process colours, whereby the print heads are always loaded with the same basic colours: cyan, magenta, yellow, black, orange and blue (or light magenta/cyan). Using sophisticated software, unlimited

94

Digital printing of textiles

combinations of colours can be created from these basic colours. In such a manner, many advantages and benefits are achieved: there is no need to prepare colors in the colour kitchen, there is no need to load and unload inks from the printer, and there is no need to clean the printer between jobs or colourways. Moreover, this is a much faster way to produce short to medium jobs or jobs with a large number of colourways. Most digital printers available on the market today, including the DReAM, use process colours, based on CMYK and either light colours or supplementary colours (such as orange and blue). The print heads: Scitex Vision The secret of the Aprion technology provided by Scitex Vision lies in the multilayer construction of its inkjet heads (Fig. 6.3). These are only 1.5 mm thick. The heads measure 5.9  0.8 inches and print at 600 dpi resolution. The top layer of the `sandwich' is a grid consisting of hundreds of piezoelectric drivers. Next is a layer of porous metal, which allows the ink to flow to the bottom layer, which contains the nozzles. This structure allows the ink to flow reliably at high firing rates over a very wide cross-section ± as opposed to other technologies involving micro-channels (unlike continuous-flow inkjet technology, there is no ink waste and unnecessary recycling). The heads are of the drop-on-demand type, which means that the piezoelectric driver above each nozzle creates a shock wave, causing a droplet to be emitted only when required. Although the ink flows slowly through the porous layer, each shock wave pulls it through rapidly, thus eliminating crosstalk between nozzles. The print heads installed on the DReAM run at

6.3 Scitex Vision digital drop-on-demand print heads.

Industrial production printers ± DReAM

95

Table 6.1 The principal technical characteristics of the Aprion print heads Head dimensions Resolution Number of nozzles per print head Ejection frequency Mode Structure Printing method

5.9 inches  0.8 inches True 600 dpi 512 Over 30 kHz Binary drops Rooftop piezo drop-on-demand; multi-layer structure that enables high rate, reduced crosstalk Scanning array

speeds exceeding 30,000 droplets per second; however, the Aprion inkjet heads already demonstrate a potential of up to five times that speed. For each of the six colours on the machine there are seven heads. The total number of heads in each DReAM printer is 42. Each head has 512 nozzles. The printing resolution is true 600 dpi (dots per inch). The principal technical characteristics of the Aprion print heads are listed in Table 6.1. The technology is protected by numerous patents and pending patents that have been awarded in the USA, Europe, Japan, Canada and Israel.

6.3

New opportunities offered by the new Reggiani digital printing machine: Digital Technological Center (DTC)

6.3.1 What is the DTC? The DTC, or Digital Technological Center, is a recent realization by Ciba Specialty Chemicals Inc. and Reggiani Macchine S.p.A., aimed at furthering industrial digital printing. Established on 1 March 2004 in Grassobbio (Bergamo) at Reggiani Macchine, the DTC is an independent organization and the result of a common project between the two partners. This modern structure is dedicated to the study and research of the digital workflow, production process and the pre/post treatment of digital printing. It is specifically geared to: · Development of new products with Ciba: inks, pre-treatment and posttreatment · Development supporting Reggiani: software and hardware, colour calibration, CAD driver releases, upgrades and workflow research · Coordinating and hosting events and conferences to foster industrial digital print penetration · Supporting textile universities in promoting the vision of industrial digital printing.

96

Digital printing of textiles

6.4 The Reggiani Digital Technological Center (DTC).

The DTC's scope is to provide a higher added-value service, offering clients constant improvement in the digital printing experience and consistent service. See Fig. 6.4.

6.3.2 The DTC's activities The facility will include a DReAM system, RIP and software facilities in a demonstration center, training facilities, and a lab and offices. The DTC's mission is research and development into the specific technology aimed at promoting industrial digital printing; contributing to new business development by offering clients complete assistance and aiding their industrial needs; and leading clients to attain the know-how necessary to set a benchmark for digital printing internationally. The technological center offers the following services: · · · · · · ·

Sampling Printer driver acceptance for third parties Short-run production Collection development Colour calibration Pre-treatment service Post-treatment service

Industrial production printers ± DReAM · · · ·

97

Online technological support Education Client start-up Set up of turnkey industrial textile processes.

The DTC's team of highly qualified personnel display a gamut of experience unmatched in the industry with experts in machine engineering, industrial textile know-how, inks, chemistry, software and colorimetric science. The DTC has been structured to efficiently offer clients support and know-how through all phases of industrial growth from start-up to uptime guarantees, pointing on technology as a tool and asset. The applied technology offers: · Live monitoring between the DTC and client machinery · On-site and remote support and service. DTC offers training to fulfil all client needs (machine operators will be trained by Reggiani staff and their partners); all DTC courses are offered at the Reggiani Macchine Grassobbio location and can also be held at a client site. Training includes books and materials. Training materials (student manuals) and the training schedule will be shipped to the client site one to two weeks before classes start. Reggiani's presentations are delivered using a high-resolution PC projection unit (1024  768 resolution) and an erasable whiteboard writing surface, both required for each class.

6.4

Bibliography

Fabio Viviani, Sinergia di competenze, Tecnologie Tessili, May 2003, Reed Business Information. Internal documentation.

7 Industrial production printers ± Mimaki's Tx series H K O B A Y A S H I , Mimaki Industries, Japan

7.1

Evolution of digital printing

Fabric printers using ink-jet technology were introduced to textile printing vendors at the end of the 20th century. The growth of the ink-jet printer market is attributed to advances in ink-jet technology and rapid acceptance of industrial wide-format color ink-jet printers (WFP) for sign graphics starting approximately 10 years ago. Establishing textile printing technology, WFP manufacturers aimed to expand print targets to fabric, which would provide a much larger market than sign graphics. After careful consideration, WFP manufacturers adopted one of three approaches according to their business circumstances: · Developing original textile printers based on WFPs for paper or film printing · Leaving such development to customizing companies that modify and market such printers · Taking no action for textiles. At the International Textile Machinery Exhibition in 1999 (ITMA 99), six printer manufacturers exhibited printers they had specially developed for textiles to fulfill the first of the above options. Also, several manufacturers showed printers they had extensively modified in terms of fabric feed mechanism, bulk ink system, etc., based on WFPs for graphics. Both ITMA 99 in June 1999 and another textile show, HimeTex 2000 in January 2000, contributed much to gain recognition and spread textile ink-jet printers rapidly into the market. Especially, HimeTex 2000 introduced 24 textile ink-jet printers at 11 stands in total, announcing that textile ink-jet printing was finally becoming available commercially. Of the 24 textile ink-jet printers at HimeTex 2000, 15 machines (62%) were manufactured by Mimaki Engineering Co., Ltd. Also, five printers based on WFPs modified for textiles were all modifications of Mimaki's Tx-1600S. This machine, the first version of the Tx series, was easy to modify for textiles because of its high quality piezo-head and low cost. To meet market needs for production of sample products in small lots, printer manufacturers and customizing companies have enhanced printer functionality,

Industrial production printers ± Mimaki's Tx series

99

including the print speed, fabric transportation, allowable weight of rolled fabric, supported ink types, and dryer equipment. These improvements have further accelerated the marketing of textile ink-jet printers and continue to do so today.

7.2

Marketing profile of Mimaki's Tx series

The Tx series lineup includes the Tx-1600S, the first model released in October 1998 (the year before ITMA 99), the Tx2-1600 (August 2001), and the Tx31600 (October 2004), with total sales of approximately 1500 units: see Figs 7.1±7.3. Figure 7.4 shows annual worldwide sales of the Tx series (three models) and those of digital fabric printers in the market (our estimation). Figure 7.5 shows a regional sales breakdown of the Tx series.

7.1 Tx-1600S.

7.2 Tx2-1600.

100

Digital printing of textiles

7.3 Tx3-1600.

7.4 Annual worldwide sales of the Tx series (three models) and those of digital fabric printers in the market (our estimation).

7.5 Regional sales breakdown of the Tx series.

Industrial production printers ± Mimaki's Tx series

7.3

101

Market needs for digital textile printing

There are various market needs for digital textile printing that differ from WFPs for graphics, compared to conventional screen textile printing: Productivity: 20±30 m2/h print speed for practical use High resolution: enabling printing of fine tie patterns Support of various fabric types: stretching/shrinking, thin, raising fabric, etc. Color reproduction: equal or higher color gamut to screen textile printing; high color reproduction when reprinting or between different models · Strike-through: color permeability to rear surface (especially for scarves) · High fastness: equal fastness to screen textile printing · Low running costs: slightly higher printing costs than manual textile printing.

· · · ·

The importance and achievement level of these needs differ depending on the intended purpose of the digital textile printing. For small lot production, productivity and running costs are especially important because of printing end products. For color correction printing, color reproduction and high resolution are important. Before now, many users who introduced the Tx series had used several printers for small lot production. Therefore, the highest market need is demand for productivity and running costs.

7.4

Technical issues and solutions

7.4.1 High resolution images The quality of digitally printed images is determined by printer resolution, variable dot size, and fabric feed accuracy of digital textile printers, and those factors are now discussed. Resolution Printers with finer resolution produce higher-quality output. Conventional screen printers typically use 100 to 300 mesh screens, resolution of which is comparable to 254±770 dpi (100 m down to 33 m) for the resolution of digital images. For textile printers, 720 dpi is a practical and necessary plotting resolution. To print actual images with resolution equivalent to that of paper inkjet printers, it is important to inject the proper amount of ink that gives the right dot length on the textile appropriate to the resolution from the nozzle. The Tx series adopts a 720 dpi plotting resolution. Realizing 80±100 m print dot length by 5±25 pl (1 pl ˆ 10ÿ12 liter) ink drop size, high-resolution printing is available. The nozzle hole pitch of the ink-jet head incorporated in the Tx series is 180 dpi (141 m). However, it is capable of performing real 720-dpi resolution (35 m) printing by scanning four times separately between nozzles. Also, for enhancing image quality, it is possible to print a scanning dot line

102

Digital printing of textiles

every dot using two types of nozzle. In this case, 720  720 dpi meshes will be coated by scanning eight times. In the same way, a printing mode using 16 times scanning is available. Printing using multiple scanning is time-consuming. So it is necessary to use the appropriate mode depending on the desired print quality. Variable dot size Popular methods for color tone control are digital dithering, pseudo halftone reproduction using an error diffusion method, and combination with light colors. In gradation of dark colors only, graininess is often apparent in the highlight area where basic dots are large and dark. Use of smaller dots with those methods is effective in reducing graininess and preventing a tone-jump phenomenon. Tx2 and Tx3 printers are capable of manipulating dots in three sizes as shown in Fig. 7.6. These dots in different sizes allow smooth tone gradation with less graininess. Fabric feed accuracy (banding prevention) When feed fluctuations for every head scan occur, a striped pattern in the scanning direction (banding) is seen on the resulting print and leads to poor image quality. Banding occurs chiefly because of faulty fabric feed and irregularities in the fabrics themselves. Its causes are as follows: · Existence or nonexistence of slippage at fabric clamping mechanism part (printer) · Tension uniformity of fabric (printer) · Stretch fabric or not (fabric, material) · Fabric wet expansion/shrinkage by ink (fabric, material) · Slippery surface of fabric (fabric, material)

7.6 Tone gradation by variable-sized dots: (a) with variable dot sizes; (b) with a single dot size.

Industrial production printers ± Mimaki's Tx series

103

· Concavity and convexity of fabric (embossment, crape) (fabric, material) · Uniformity of edge of roll fabric (in shape) (fabric, pre-treatment) · Meandered or curved fabric (fan-like, deformation to s-shape) (fabric, pretreatment). The feeder must be equipped with a mechanism that prevents a fed fabric from slipping, and fabric can withdraw easily from it to prevent fabric from imperfection. As the anti-slip mechanism, the Tx2-1600 (Fig. 7.7) and Tx3-1600 (Fig. 7.8) have knurled rollers and a feed system using an adhesive belt (table adhesive method), respectively. The knurled rollers of Tx2 have fine projections on the stainless-steel surface and have strong friction in the thrust direction when winding fabric. If the height of toothing is too great, the amount of toothing to be pierced into the fabric will increase and lead to imperfection in the fabric. Also, fabric cannot be easily withdrawn. Feeding is performed under fixed tension, pulling the entire fabric by the roller with special surface treatment at the point immediately after printing (Fig. 7.9). Applying clingy paste to the wide endless belt surface, an adhesive belt feeds fabric by sticking it to the belt. Because the entire fabric sticks to the belt, stretch material such as knit can be fed and textiles are prevented from wet expansion/shrinkage by ink. As the adhesive power of paste will deteriorate with use, it is necessary to put paste on the belt periodically. Also, varying thickness

7.7 Tx2 feeding roller.

104

Digital printing of textiles

7.8 Tx3 feeding belt.

7.9 Tx2 fabric tension mechanism.

Industrial production printers ± Mimaki's Tx series

105

7.10 Tx3 detecting the travel distance of the belt.

of the feeding belt and non-uniform paste lead to accidental errors in fabric feed accuracy. In order to correct feed fluctuations, movement of the belt is controlled by detecting its travel distance at the final stage by a rotary encoder with a feedback system to the belt driver motor (Fig. 7.10). To reduce variations of the tension applied to the fabric, the Tx2-1600 has a torque limiter and a conditioning roller mechanism followed by the drive section; the Tx3-1600 has parallel tensioning bars. Generally, when feed accuracy has 20 m or more accidental error, visible banding tends to occur. As mentioned above, fabric characteristics such as expansion/shrinkage, wet expansion/ shrinkage by ink, slippery surface, concavity and convexity also affect banding. Pre-treatment of fabric consists of impregnation by an agent mainly consisting of paste and then drying. This pre-treatment can prevent bleeding of dye inks and improve feed performance. (Section 7.4.3 explains the details of prevention of bleeding.) For fabrics woven with a hard twist and large expansion/shrinkage, pre-treatment paste coating should be increased to improve feed performance. Adjustment of the composition of the agent, width of the tenter, speed and strength of take-up by the fabric all affect the result. If the make-up of the agent is not suitable, fabric may be starched or stressed, causing skewed feeding, meandering, and non-uniform pitch. These cause banding. Figure 7.11 shows typical examples of results of pre-treatment. Skewing of the fabric causes banding, fabric slip upon feeding, and wrinkles, eventually leading

106

Digital printing of textiles

7.11 Skewing caused in the pre-treatment.

to the head nozzle dragging on the print fabric surface and to fabric jamming in the printer.

7.4.2 Color reproduction Figure 7.12 compares the color gamuts of reactive dyes, acid dyes, disperse dyes, and water-based pigment inks. The figure suggests that the color gamut of dye inks compares favorably with that of pigment ink.

7.4.3 Prevention of bleeding Conventional screen printing and ink-jet printing use different pastes for print inks. To prevent ink bleed, screen printing uses a volume of pastes that makes inks much more viscous than those for ink-jet printing for dye. We call the mixture of textile dye ink and pastes `printing pastes'. Hand screen printing, automatic flat-bed screen printing, and rotary screen printing use pastes with lower viscosity in that order. Generally, screen printing uses pastes with viscosity of some hundreds to tens of thousands of mPa
s. Ink-jet printing adopts pastes with much lower viscosity, from a few to over 10 mPa
s. It requires light pastes for spraying ink droplets of a few to several tens of picoliters in size at a high jet frequency of approximately 10 kHz from the printer head nozzles. Ink droplets with low viscosity pastes would produce ink bleed on a fabric. Therefore, coating the target fabric with a pre-treatment agent, the main element of which is a paste, is necessary. Ink bleed and strike-through on the printed fabric occur due to a capillary phenomenon. The ink penetration length through the capillary is calculated with the Lucas±Washburn equation, which expresses the relation between the penetration length L and the viscosity  as follows: L / ÿ1=2

…7:1†

Industrial production printers ± Mimaki's Tx series

107

7.12 Color gamuts of fabric printing inks.

This equation proves that the penetration length becomes shorter as the ink viscosity increases. In conventional screen printing, an appropriate ink viscosity is chosen according to the fabric type for printing, the amount of ink to be applied, and the print speed to achieve the optimum print quality and avoid ink bleed. In the ink-jet printing, however, the nature of the ink-jet head prohibits the use of high-viscosity ink. For this reason, fabrics need to be coated with pastes in the pre-treatment to minimize risk of ink bleed in ink-jet printing. Pastes used in pre-treatment contain various agents to improve overall print quality: coloring, color stability and print fastness against washing. These agents are usually added to inks for conventional screen printing. Only small amounts

108

Digital printing of textiles

of those agents are contained in the ink-jet inks to control dye properties, help the dyes to be dissolved in the liquid, and preserve inks (required for assuring ink-jet performance and extending ink life). Without application of those agents in the pre-treatment of target fabrics, ink-jet printing would be unable not only to prevent ink bleed but also to ensure sufficient coloring. Agents included in conventional pastes to keep their functionality are as follows: · · · · · · · · ·

Paste and anti-dyeing agents Penetrating agents Preservatives Dye dissolving agents Level dyeing agents Softeners Anti-color bleed agents Discharging agents Adhesive: for foils and gold pigments.

For ink-jet printing, ink consists of dyes, dye dissolving agents, moisturizing agents, small amounts of binders, preservatives, pH adjusters, and surfactants.

7.4.4 Strike-through Uniform strike-through is one of the important factors in determining textile printing quality. In conventional screen printing, inks of all colors and tones are prepared according to screens to be used in the printing. Therefore, amounts of the inks applied on the fabric are irrelevant to color tones, and the amount of inks consumed is almost constant regardless of color or tones, resulting in uniform strike-through. Ink-jet printing, however, controls color tones by the amount of YMCK inks injected; thus, the amount of inks consumed can be varied depending on tones. Strike-through in ink-jet printing is, therefore, reduced in light-colored print areas, which clearly contrast in strike-through by color tones when observing them from the opposite face of the printed areas. Strike-through spotting can be effectively prevented by using a variation of inks blended for required color tones. For reproduction of YMCK colors in ink-jet printing, use more inks of light colors and spray them on as large an area as possible to consume them evenly. Figure 7.13 shows the difference in ink consumption with and without gradation inks (the light color is 33% of the dark color) for a solid color shading model. Another advantage of using dark and light inks is that smoother gradation can be achieved with more flexible tone control. Figure 7.14 gives 2  2 matrices to distinguish differences between three gradation methods: (a) a dark ink only and no ink (two-level dither method), (b) dark and light inks and no ink (three-level dither method), and (c) dark, medium, and light inks and no ink (four-level dither method). Assuming the matrix size is expressed as N  N and the number

Industrial production printers ± Mimaki's Tx series

109

7.13 Gradation and consumed ink volume.

7.14 Smoother gradation using light and dark inks.

of ink types is represented as k, the number of reproducible tones L can be expressed as: L ˆ …k  N 2 † ‡ 1

…7:2†

7.4.5 Fastness Table 7.1 shows ink fastness on fabrics used in the test. Tables 7.2 to 7.4 show the results for the fastness of digitally printed images with acid dye, reactive dye, and disperse dye inks, respectively, by a digital textile ink-jet printer.

110

Digital printing of textiles

Table 7.1 Inks and fabric fibers tested Type of ink

Cellulose fibers (cotton, hemp, rayon, TencelÕ)

Reactive dye

Available

Acid dye Disperse dye Pigment

Not available Not available Available

Protein fibers (silk, wool)

Polyester fibers

Can be used under Not available a certain condition Available Not available Not available Available Can be used under Can be used under a certain condition a certain condition

Nylon fibers

Not available Available Available Can be used under a certain condition

7.4.6 Applicability of fabrics: feed mechanism Fabrics can be grouped as follows in terms of fabric feeding: · Type 1: Stretch (including knit) · Type 2: High wet expansion/shrinkage · Type 3: Low strike-through resistance. Type 1 fabrics are elastic and stretch literally by external forces; they need to be conveyed with minimum tension applied or with a certain degree of tension constantly applied in pre-process or during plotting. It must be noted, however, that conveyance without tension will easily meander (move) the fabric by distortion inherent to the fabric itself (variation of strand/thickness/weaving density) and from the pretreatment, which eventually causes skew and dislocates the fabric position. Thus, it is best to convey the fabric by applying the minimum tension required for correcting fabric distortion inherent in the fabric or caused by the pretreatment. Type 2 fabrics, similar to Type 1, exhibit numerous sags or wrinkles due to fabric stretching in the printing process. Type 2 wet shrink fabrics will shrink toward the fabric center; both right and left edges lose straightness, shaping like steps according to the scanned bandwidth. Those fabrics start shrinking just after being scanned (printing) and will be shrunk upon the next scanning. Thus, the edges are deformed as shown in Fig. 7.15. The shrinkage causes greater deviation toward the fabric edges; print images are thus distorted, although white and black stripes do not appear. To solve this problem, some means for controlling the fabric is necessary; flat belt printing used for screen printing is effective. Type 3 fabrics involve a problem where inks running through webs or penetrating to the back of woven fibers leave spots on the platen surface, which makes the fabric stained as conveyed. To avoid this stain problem requires either a contact-free system in which the fabric is not in contact with the platen, or a synchronized conveying system in which the fabric is conveyed with the platen.

Table 7.2 Color fastness ± reactive dye ink and a cotton fabric pre-treated with Mimaki's recipes Unit: grade Item

Test method

Category

Fabric pre-treated with the Mimaki recipes BL

C

GR

GY

K

LC

LM

M

O

R

Y

Light

JIS L 0842

E reactive ink cotton

4 or higher

4 or higher

3±4

4 or higher

4 or higher

3±4

3±4

4

4

4 or higher

4 or higher

Laundering

JIS L 0844

Fading/discoloration Cotton Staining Silk

5 5 5

5 5 5

5 4±5 5

5 4±5 5

4±5 4±5 4±5

5 5 5

5 5 5

5 4±5 5

4±5 4±5 5

4±5 4±5 5

5 5 5

Drycleaning

JIS L 0860 Perchloroethylene JIS L 0860 Petroleum-based

Fading/discoloration Staining Fading/discoloration Staining

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

4±5 5 ± ±

4±5 5 ± ±

5 5 ± ±

Sweat

JIS L 0848 Acid

Fading/discoloration Cotton Staining Silk Fading/discoloration Cotton Staining Silk

5 4±5 4±5 5 4±5 4±5

4±5 3 3±4 5 2±3 4

4±5 4±5 5 4±5 4±5 5

5 4 5 5 4 5

4±5 4 5 4±5 4 5

5 4 4±5 5 4 4±5

5 4 4±5 5 4 5

5 3±4 4 5 3±4 4±5

4±5 4 4±5 4±5 4 4±5

4±5 4±5 4±5 4±5 4 4±5

5 4±5 5 5 4±5 5

Alkali

Water

JIS L 0846

Fading/discoloration Cotton Staining Silk

5 4±5 4±5

5 3±4 4±5

5 4±5 5

5 4±5 5

4±5 4±5 5

5 4±5 5

5 4±5 5

5 4 5

5 4±5 5

4±5 4±5 5

5 4±5 5

Friction

JIS L 0849 Type II

Dry Wet

3 3±4

3 3±4

4±5 4±5

3 3±4

2±3 2±3

4±5 4±5

3±4 4

2±3 3

2±3 3

2±3 2±3

3±4 4

Hot pressing

JIS L 0850 A-2 Dry

Fading/discoloration Staining

5 5

5 5

5 5

5 5

5 5

5 5

5 5

5 5

5 5

5 5

5 5

DTK no.

2697

The test was commissioned to theTokyo Office of Japan Dyer's Inspection Institute Foundation/JapanTextile Finisher's Association. The mark `±' means that the test was not conducted. Printing condition on each medium: 720 dpi, four-pass, 100% of single color. Post-treatment conditions for the test media were all the same. The bold numbers mean a failure.

Table 7.3 Color fastness ± acid dye ink and a silk creªpes de chine fabric pre-treated with Mimaki's recipes Unit: grade Item

Test method

Ink color

Category K

Y

M

C

LM

LC

R

Light

JIS L 0842

7

5

3±4

3±4

3±4

3

4

Laundering

JIS L 0844

Fading/discoloration Staining Cotton Silk

4±5 4±5 4±5

4 4±5 4±5

3 4±5 4±5

3±4 4 5

3 4±5 4±5

3±4 4±5 5

3±4 2±3 4±5

Drycleaning

JIS L 0860 Perchloroethylene JIS L 0860 Petroleum-based

Fading/discoloration Staining Fading/discoloration Staining

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 4 ± ±

Sweat

JIS L 0848 Acid

Fading/discoloration Staining Cotton Silk Fading/discoloration Staining Cotton Silk

4±5 4±5 4±5 4±5 4±5 4±5

5 4 3 5 4 3

5 4 1±2 5 3±4 1±2

4±5 4±5 5 4±5 4 4±5

5 4±5 3 5 4±5 2±3

5 4±5 5 5 4±5 5

5 4 3±4 5 3±4 2±3

Alkali

Water

JIS L 0846

Fading/discoloration Staining Cotton Silk

4±5 4±5 4±5

5 4 2±3

5 3±4 2

4±5 4±5 5

5 4±5 3

5 4±5 5

5 4±5 2±3

Friction

JIS L 0849 Type II

Dry Wet

5 2±3

5 3±4

5 3±4

5 3±4

5 4±5

5 4

5 2±3

DTK no.

1043

The test was commissioned to Japan Dyer's Inspection Institute Foundation/JapanTextile Finisher's Association. The mark `±' means that the test was not conducted. Pre-treatment: pre-treatment 1for acid dye inks of Mimaki Engineering Co., Ltd (Mimaki's original recipe: see page 4 of the ink guidance). Post-treatment: same as pre-treatment. Fabric used for the test: silk creªpes de chine. Printing condition: 720 dpi, eight-pass, uni-direction, 100% of single color.

Table 7.4 Color fastness ± disperse dye ink and a polyester fabric pre-treated with Mimaki's recipes Unit: grade Item

Test method

Light

JIS L 0842

Laundering

JIS L 0844

Drycleaning

Sweat

Category

Ink Color C LM

K

Y

M

LC

GR

3

6

5

Less than 3

4

Less than 3

3

3

Fading/discoloration Staining Cotton Silk

4±5 4±5 5

4±5 4±5 4±5

4±5 4±5 5

4±5 4±5 5

4±5 4±5 5

4±5 5 5

4±5 4±5 5

4±5 4±5 5

JIS L 0860 Perchloroethylene JIS L 0860 Petroleum-based

Fading/discoloration Staining Fading/discoloration Staining

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

5 5 ± ±

JIS L 0848 Acid

Fading/discoloration Staining Cotton Silk Fading/discoloration Staining Cotton Silk

4±5 4±5 5 4±5 4±5 5

4±5 4±5 4±5 4±5 4±5 4±5

4±5 4±5 5 4±5 4±5 5

4±5 4±5 5 4±5 4±5 5

4±5 5 5 4±5 5 5

4±5 5 5 4±5 5 5

4±5 4±5 5 4±5 4±5 5

4±5 4±5 5 4±5 4±5 5

Alkali

BL

Water

JIS L 0846

Fading/discoloration Staining Cotton Silk

4±5 4±5 5

4±5 4±5 4±5

4±5 4±5 5

4±5 4±5 5

4±5 5 5

4±5 5 5

4±5 4±5 5

4±5 4±5 5

Friction

JIS L 0849 Type II

Dry Wet

5 5

5 5

5 5

5 5

5 5

5 5

5 5

5 5

Hot pressing

JIS L 0850 A-2 Dry JIS L 0850 A-2 Wet

Fading/discoloration Staining Fading/discoloration Staining

5 5 5 5

5 4±5 5 4±5

5 5 5 5

5 5 5 5

5 5 5 5

5 5 5 5

5 5 5 5

5 5 5 5

Fading/discoloration

5

5

5

5

5

5

5

5

Chlorinated Water JIS L 0884 Method B DTK no.

1213/3162

The test was commissioned to Japan Dyer's Inspection Institute Foundation/JapanTextile Finisher's Association. The mark `±' means that the test was not conducted. Pre-treatment: pre-treatment for disperse dye inks of Mimaki Engineering Co., Ltd. Post-treatment: same as pre-treatment. Fabric used for the test: polyester specially designed for evaluation by Mimaki's Development Department (microfiber). Printing condition: 720 dpi, four-pass, uni-direction, 100% of single color.

114

Digital printing of textiles

7.15 Step-shape edges due to wet shrinkage.

The former provides an open space between the printing area and the platen, and the latter conveys the belt or table itself, keeping the printed fabric on it. The model Tx2-1600 has adopted features that avoid such problems inherent in the fabrics (see Fig. 7.16). For Type 1 fabrics, the unit suppresses abrupt fluctuation of fabric tension with a feed torque limiter and a conditioning roller mechanism located just before a driving roller. A tensioning roller, located ahead of the printer head, runs at a slightly higher peripheral speed to minimize fabric sagging. And finally, a roll-up torque limiter rolls up the fabric with no abrupt fluctuation of the tension. For Type 2 fabrics, addressing applicable fabrics for the unit helps users to determine the choice of machines for their specific task. For Type 3 fabrics, the contact-free (ditch) system is adopted. A tensioning roller is provided ahead of this ditch system to prevent the fabric from sagging by gravity. The model Tx3-1600 solves the problems by the following features (see Fig. 7.17). For Type 1 fabrics, parallel tensioning bars are provided. The feed/take-up

Industrial production printers ± Mimaki's Tx series

115

7.16 Tx2-1600 textile path.

control mechanisms control the bar positions via feedback and suppress the fluctuation of tension. For Type 2 fabrics, the table adhesive method is adopted. For Type 3 fabrics, the flat belt method is adopted to prevent the fabric back from becoming stained by inks.

7.17 Tx3-1600 textile path.

116

Digital printing of textiles

7.4.7 Running cost Ink-jet printing does not require stencils; it achieves cost-effective printing for a much smaller print run than does conventional screen printing. Table 7.5 compares costs of ink-jet printing to conventional screen printing. Figure 7.18 shows the production cost (variable cost) per square meter depending on the production volume. As shown below, 1200 m2 or less is a rough standard below which ink-jet printing is more economic. Actually, extinguishing of facilities, cost of washing stencils, number of production staff, etc., also affect the curve chart of cost and favour ink-jet over screen printing. As the cost of ink accounts for a high proportion of ink-jet printing cost, to reduce this ink cost is the key to wider adoption of ink-jet printing.

7.4.8 Ink-jet stability and reliability Disincentive elements of stable and reliable ink-jet performance As ink-jet printing tends to be used more in production, stable and reliable performance is very important. For ink-jet printing, possible disincentive elements of stable and reliable performance are the injected inks themselves and fabric feed. This section explains how to ensure ink-jet stability and reliability. Stable and reliable performance is hampered mainly by the following causes and leads to defects such as white lines, non-uniformity of colors, spots, etc.: · Ink physicality and mismatch with head. As mentioned in Section 7.4.1, the Tx series printers inject a very small ink drop of 5±25 pl (1 pl ˆ 10ÿ12 liter) at high speed with a frequency of about 10 kHz from the nozzles. Therefore, the

Table 7.5 Comparison of costs of ink-jet printing to those of conventional screen printing

Cost of inks and color pastes Volume of inks and color pastes Cost of pre-treatment Stencil Printing cost Environmental impact Quality comparison: Concentration Definition Texture

Ink-jet printing

Screen printing

¨150±222 per kg 20 g/m2 Required < ¨100 per m2 Not required ¨3.7±5.2 per m2

¨1.48±2.22 per kg 100 g/m2 Not required

Minimum

Required ¨0.15±0.22 per m2 + stencil cost: ¨296±370 per stencil Serious

Poor Good Poor

Good Poor Good

Industrial production printers ± Mimaki's Tx series

117

7.18 Comparison of 1 m2 unit cost.

flow channel geometry of ink in the head is complicated by fine processing technology. If ink grains that have too wide a diameter are mixed in the ink, clogging will occur in the head. Also, if ink physicality values cannot follow the high-speed movement of the piezo head, ink drops will not be injected but ink mists are generated instead, or ink will not be injected at all. · Lint and dust adhered to the head nozzles. Tiny lint particles and dust adhered to the fabric are stirred up by the air current of the scanning head and easily adhere to the surface of nozzles wet with aqueous ink. Adhered substances near the nozzle hole prevent ink from injecting and lead to white lines. In addition, adhered substances may get into nozzles during cleaning or wiping operations. In this case, unrecoverable deflection will occur or ink will not be injected at all. Though fabric pre-treatment can reduce such problems, it is realistically impossible to eliminate them totally. · Dirt on nozzle surface generated by ink mists. If ink mists are generated from any cause while ink is injecting, the mists may be floating in the air and they land on the head and become a bigger ink drop. When the drop reaches a certain size, it may drop onto the fabric during scanning and make stains. Measures taken for the Tx3 The Tx3-1600 is equipped with the function to detect an injection failure automatically. The ANR (Automatic Nozzle Recovery) unit is designed to assist users to check for clogging. It outputs a unit-unique print check map periodically, as shown in Fig. 7.19, on a designated medium and reads by a sensor. The unit itself is shown in Fig. 7.20. Users can configure the unit

118

Digital printing of textiles

7.19 Print check pattern used for ANR unit.

according to their designed print parameters, including checking intervals, thresholds for error determination (nozzle(s) causing void), and recovery procedures (cleaning levels). The Tx3-1600 is also equipped with the function to vacuum mists up while they are floating in the air and trap them in one place.

7.4.9 Productivity Direct ways to improve the productivity of the ink-jet printer while maintaining print quality can be as follows: · Improving the print head speed (improving the jet frequency) · Increasing the ink-jet nozzles of the print head. Indirect ways can be as follows: · Capability for a large-sized long roll-type fabric (reducing the process steps)

Industrial production printers ± Mimaki's Tx series

119

7.20 ANR unit.

· Capability for a large-sized ink tank and a toggle cartridge (reducing the process steps) · Improving stability and reliability of unattended operation. Some of these are explored below. Increasing nozzles by introducing a stagger head Improving the print head speed means developing a new head; this requires time and the initial investments. Multiplying nozzles of a printer head may be technically difficult; however, aligning multiple heads can solve the difficulty. This modification involves undetermined technical and maintenance tasks, such as head±head position adjustment (inclination, gap), head variations, and maintenance work for the service shops. Toggle-switch type cartridge To reduce the process steps, adoption of a large-sized roll-type fabric or a largesized ink tank seems an easy option. In reality, a toggle-switch type ink cartridge is much superior to those options. It is a device for switching an ink cartridge when its ink level calls for replacement. One-color ink cartridge or one-color

120

Digital printing of textiles

multiple cartridges that simultaneously operate (in parallel) must be replaced when the ink runs short or the predetermined ink-end level is reached. If the ink runs short while the printer is running at night without operators, printing must be stopped until operators return, causing unnecessary downtime. Users often reluctantly install a new cartridge much earlier than the ink-end level indicates because of forecast downtime during operator absence. As a result, they discard cartridges in which much usable ink remains. Conventional ink cartridges can be replaced only as long as an operator attends the printer; however, the operator must reset the printer quite frequently. The toggle-switch type cartridge, in contrast, allows users to use the ink right to the end, freeing the operators from frequent cartridge replacement. Therefore, the Tx2-1600 has adopted the toggleswitch type. Stability and reliability of unattended operation To maintain productivity, it is very important to ensure stability and reliability of unattended operation, including night operation. As explained, fabric print performance can be assured by lint and dust control. Without thorough control, an all-night print job may end up in vain the next morning because of voids and defects on the print. Tx3-1600 models, equipped with ANRS, assure users reliable non-stop printing.

7.5

The future of digital printing

This section discusses tasks that need to be overcome for digital printing to grow in the future.

7.5.1 A simple and easy digital printing system A digital printing system is not limited to the ink-jet printer itself; it involves many other operations to provide a simple and user-friendly printing system that will gain wider acceptance in society: · · · ·

The fabrics to be printed Agents, techniques, tools, and vendors for pre-treatment A textile ink-jet printer of low cost but high performance Post-treatment (steaming, washing).

Currently, one company, Seiren Co. Ltd, has successfully established such a complete printing system on its own and has been the only player so far in the digital printing market. Some of the keys to popularizing digital printing are disclosure of pre/post-treatment recipes, encouraging vendors to welcome small lot printing, and marketing and publicizing printing vendors for easy accessibility for self-print users. Mimaki has been making efforts to collect

Industrial production printers ± Mimaki's Tx series

121

information on pre/post-treatment and to disclose technical information to a wide range of users in support of fabric treatment vendors and users.

7.5.2 Improving productivity Once digital printing speed exceeds 200 m2/h, it will be able to replace conventional screen printing in terms of printing speed. Currently, digital printing speed is 30 m2/h or less for practical use; thus, multiple printer operations cover the productivity. As well as improving the printing speed, maintaining high quality of printed images is essential. Achieving 30±50 m2/h in a high quality mode of 720 dpi is the market demand that is technically expected. At the same time, ink cost must be reduced, as mentioned in the following section.

7.5.3 Promotion of digital printing technology to a higher stage In Italy and France, digital printing technology has been widely promoted. Mimaki has shipped 1000 sets of textile ink-jet printers to both countries in total. Successful promotion of digital printers in those countries has been backed up with branding of digital printing. The ability of digital printing to create eye-catching designs and brands, as well as to mark smash hits by small runs of special or luxury items, should facilitate the popularizing of digital printing. In addition to the hardware package, such as offering the print technology as a system, fostering its software side, including collaboration with high-fashion designers and hosting a T-shirt design contest for future designers in fashion design schools, for example, will gain further recognition for digital printing.

7.5.4 Creating new markets for digital printing ± highresolution images and small-lot production of a variety of products Now, entirely new and revolutionary possibilities for digital printing are being explored: gradation printing of photographs and graphical paintings, printing on leathers, to name but two. Here, digital printing is not replacing conventional techniques. Creating novel products is an essential factor to make this technology distinctive.

7.5.5 Lowering the cost Lowering running costs, including cutting costs of printers, inks, and pre/posttreatment, is also one of the essential factors for digital printing to become attractive, competitive and acceptable in the printer market.

122

Digital printing of textiles

7.5.6 Developing more advanced technologies Increase of ink-jet heads, creation of special color inks and additional lighter/ deeper colors are some of the important tasks to increase color reproduction capability and uniform strike-through. The technical problems of current digital ink-jet printing technology will be solved one by one.

8 Integration of fabric formation and coloration processes B R GEORGE, D WOOD, M GOVINDARAJ, H U J I I E , M F R U S C E L L O , A T R E M E R E and S N A N D E K A R , Philadelphia University, USA

8.1

Introduction

Traditionally, fabric formation and coloration processes have been two separate distinct processes. First the fabric was produced via weaving, knitting, or through other methods such as those employed in nonwovens, and was wound onto a take-up roll. In the case of weaving, a polymeric coating must be employed on the warp yarns for additional strength and resistance to abrasion during the weaving process.1 Specifically known as warp sizes, starch, polyvinyl alcohol, and carboxymethyl cellulose are most commonly used when working with a cotton warp. A typical sizing mixture consists of a combination of starch, partially hydrolyzed polyvinyl alcohol, and a small amount of a lubricating agent, such as fat, to ensure a more secure weaving process.1 After fabric production, the fabric is desized to remove the sizing agent, cleaned to remove any undesirable particulate matter, and sometimes bleached to provide a uniformly white surface on which to print. The fabric is then printed, usually through the use of roller screens, with a screen for each color in the design. Although roller screen printing is a fast process, from 30 up to 100 meters per minute, setting-up is time consuming, as each screen must be individually created, and changing patterns requires creation of new screens, thereby rendering roller screen printing time intensive and relatively inflexible.2 Digital inkjet printing on fabric is similar to the inkjet printers utilized with many computers to print on paper. Currently it has a relatively slow production speed, approximately 12 meters per hour, especially when compared to roller screen printing.3 Although the speeds of digital inkjet printers are increasing, their use in the textile industry has mostly been limited to producing samples and small production runs of exclusive designs.1 However, digital inkjet printing on textiles shows promise as speeds increase, especially as it offers the possibility of quick pattern changes and as such can be used to produce mass customized textiles.1,4

124

Digital printing of textiles

As the speeds of digital inkjet printing approach those of weaving looms, it might be possible to combine these two processes so that woven fabric is printed as it leaves the weaving area prior to take-up, enabling the creation of mass customized textiles as well as fast pattern changes. This could revolutionize fabric formation and coloration, as less time and handling would be required for processing which could result in lower cost and improved fabric quality. However, for this idea to be successful a link needs to be established between weaving and printing. One possibility is to desize the fabric on the loom so that it can then be printed. This would require several additional pieces of equipment to desize and dry the fabric after weaving. Additionally, the fabric would have to be treated with a print fixative, required to affix the dye to the fabric, prior to printing. This would not be much of an improvement upon the current process of fabric production and printing as it would still require several processing steps. An alternative link exists: utilizing the print fixative as a sizing agent. With this scenario the fixative would be applied to both warp and weft yarns prior to weaving, and would protect the warp yarns from abrasion during the weaving process. After weaving, the fabric would move directly into a printing area where it could be printed without the need for additional dye fixation treatments to be utilized. Thus, the goal of reducing processing steps and time may be realized. This idea has been undergoing research at Philadelphia University for the past several years and exhibits promising results for the integration of fabric formation and coloration, in this case weaving, and digital inkjet printing. In the research conducted thus far, the fiber type being examined is cotton while the dye type being utilized to print on cotton fabric is reactive dye. Cotton is the most widely used of the natural cellulosic fibers and is well known to most consumers, making it a critical area of investigation for integration. Cellulose is a polysaccharide made up of cellobiose units, which combine to form a cellulose molecule, as depicted in Fig. 8.1. In this figure, the hydroxyl (OH) groups are clearly present. These hydroxyl groups are critical for several reasons, but most importantly directly pertaining to integration, cellulose fibers absorb watersoluble dyes and finishes, which aids in chemical processing. Reactive dyes are water-soluble anionic dyes, which react with the hydroxyl groups of cellulose to become covalently bonded to the cellulosic fiber.1 The chemical reaction between a reactive dye containing a chlorinated reactive group (RG) and a cellulosic fiber takes place in the presence of a base, such as sodium carbonate (Na2CO3).1 Urea ((NH2)2±C=O) is used in the fixation step to absorb moisture. The resultant covalent bond provides good washfastness properties, and is much stronger than the weak hydrogen bonds formed between direct dyes and cellulose, making reactive dyes the preferred choice when working with cotton.

8.1 Schematic diagram of a cellulose molecule.

126

Digital printing of textiles

8.2

Experimental

8.2.1 Size evaluation In order to determine the feasibility of combining weaving and digital inkjet printing, it is necessary to find a chemical solution that can function as both a size and a dye fixative for a particular fiber type. In the case of cotton, there are several possibilities for this solution. The first possibility is size, currently utilized by the textile industry to protect the yarns from abrasion during the weaving process. Another possibility is the solution of sodium carbonate, urea, thickener, and silica, currently utilized as a cotton fabric pretreatment prior to digital inkjet printing using reactive dyes. In this solution the thickener and silica are added to increase the viscosity of the solution, required to prevent the dye solution from wicking to other areas of the fabric where that dye is not desired. There are also the possibilities of a basic pretreatment solution consisting of sodium carbonate and urea, and a solution with a slightly higher viscosity of the preceding solution but lower than the first pretreatment solution: sodium carbonate, urea, and thickener. The first experiments conducted were to determine the effectiveness of size as a fabric pretreatment solution for digital printing. To do this, 20/2 Ne cotton yarns, supplied by Huntingdon Mills, were woven into a plain weave fabric with a Sumagh 400 end sample loom, depicted in Fig. 8.2. The fabric was divided into three pieces: a control with no treatment, one treated with traditional size, and one treated with sodium carbonate, urea, thickener, and silica. To

8.2 Weaving cotton fabric on a Sumagh 400 end sample loom.

Integration of fabric formation and coloration processes

127

approximate traditional size, corn starch and water in a ratio of 15% starch and 85% water was used. The pretreatment solution consisted of 85% water, 10% urea, 2.5% sodium carbonate, 1.5% Noveon Carbopol 2491 WC thickener, and 1% Degussa Aerosil 200 silica. The solutions were padded onto the fabrics with a Werner Mathis lab padder at a wet pickup of 80% and dried with a Tsuji Senki Kogyo through air oven at 130ëC for two minutes. The three fabrics were then digitally printed with four pure color stripes of Ciba reactive dyes: CIBACRON Turquoise MI700, CIBACRON Red MI500, CIBACRON Yellow MI100 and CIBACRON Black MI900, using a Mutoh inkjet printer with an Epson print head, as displayed in Fig. 8.3. Afterwards, the fabrics were subjected to the usual post-inkjet printing processes of steaming, rinsing, and soaping at the boil, which are used to affix the dye to the cotton fibers and remove any unfixed dye from the fabric. Steaming was performed with an Arioli steamer at 103ëC for eight minutes, which was followed by a fiveminute cold water rinse. The final process consisted of soaping the fabric at a boil for two minutes with a mixture of water and Synthrapol detergent, done with a beaker and a laboratory hotplate. After conditioning at standard temperature and humidity levels for 24 hours the fabrics were evaluated for colorfastness to laundering, wet and dry crocking, and light. Laundering colorfastness was determined via AATCC 61 with an Atlas Launderometer. Lightfastness testing, performed via AATCC 16 with a QSun 1000 XENON test chamber, was done for both 20 and 40 hour exposure times. Crockfastness was evaluated with an Atlas vertical rotary crockmeter according to AATCC 116.

8.3 Printing plain weave cotton fabric with color stripes on Mutoh inkjet printer.

128

Digital printing of textiles

8.2.2 Print fixative as a possible sizing agent Based on the results of the fastness evaluations, the three possible pretreatment solutions of water, sodium carbonate and urea; water, sodium carbonate, urea, and thickener; and water, sodium carbonate, urea, thickener, and silica were applied to yarns to evaluate their potential as sizing solutions. The untreated control is hereafter referred to as Solution A. The formula for water, sodium carbonate, and urea, hereafter referred to as Solution B, is 85% water, 10% urea, and 5% sodium carbonate. The formula for water, sodium carbonate, urea, and thickener, hereafter referred to as Solution C, is 86% water, 10% urea, 2.5% sodium carbonate, and 1.5% thickener. Solution D consists of 85% water, 10% urea, 2.5% sodium carbonate, 1.5% thickener, and 1% silica. All components of these solutions were those mentioned previously. Solutions B±D were padded onto 20/2 Ne yarns, supplied by Huntingdon Mills, which were also utilized for the control, Solution A. The yarns were padded with the solutions and dried via the same method that was used to pad and dry the initial fabrics. Afterwards, the yarns were allowed to condition under standard temperature and humidity conditions for 24 hours prior to evaluation. The yarns were evaluated for tenacity before and after abrasion in order to determine if solutions B±D would be effective as sizing agents. Twenty yarn specimens of solutions A±D were evaluated for single end tenacity according to ASTM D2256 with a Testometric SDL constant rate of elongation tensile tester. These yarns were the pre-abrasion specimens. The 20 post-abrasion specimens from solutions A±D were subjected to an abrading force of 454 grams for 75 cycles of a CSI flex tester, usually utilized to measure flat abrasion, similar to the type and amount of abrasion that would be incurred on a loom during the weaving process. After abrasion these yarns were also tested to determine single end tenacity, as described previously.

8.2.3 Evaluation of different print fixatives Based on the results of the initial sized fabric evaluations as well as the results of the single end yarn tenacity tests, it was decided to evaluate the effects of Solutions B±D when utilized to affix dyes to fabric via digital inkjet printing. Solutions B±D were padded onto cotton plain weave fabric produced with 20/2 Ne yarns on the Sumagh loom. The padding and drying process utilized was that aforementioned. The control in this instance was photo quality glossy paper, which currently provides the highest quality print line. The three fabric samples and the paper sample were digitally printed with CIBACRON Black MI900 reactive dye using a Mutoh Full Color Inkjet printer. Printed on each sample were lines of 0.25, 0.50, 0.75, and 1.0 points width. Afterwards, the textile samples were steamed, rinsed, and soaped at the boil utilizing the method described previously. The four samples were then analyzed to determine the line quality. As the paper provides the best printed line quality, the fabrics were compared to the

Integration of fabric formation and coloration processes

129

results of the paper to determine which solution provides the best results. The line quality was measured with a Personal IAS (Image Analysis System) by Quality Engineering Associates, and three variables were evaluated according to ISO 13660: line raggedness, line width, and line density. Line raggedness, commonly referred to as rag, is the average value of the leading and trailing edges of the line. Generally, the leading edge is the left edge of the line, while the trailing edge is the right side of the edge for a vertical line, whereas for a horizontal line the leading edge is the top edge and the trailing edge is the bottom edge. Rag is used to determine geometric distortion of a line from its ideal position. Line width is a measure of the actual width of a line compared to its theoretical width. Line density is the evaluation of the darkness of the line, expressed in terms of optical density (OD) units, determined with a density standard and color filter, as specified in ISO 13660.

8.2.4 The final evaluation Based on the results of all evaluations completed thus far, it was decided to focus on solution D ± thickener, silica, urea, sodium carbonate, and water ± as a combination size and print fixative. To determine its appropriateness for this task, solution D was padded onto yarns which were then woven into fabric, which was then printed, treated, and finally evaluated. The yarns utilized for this segment of the project were 100% cotton 12 Ne rotor spun yarns, produced on a Rieter rotor frame. The yarn was wound into 70 skeins each with a length of 210 yards and then padded with solution D with a Werner Mathis lab padder to give a wet pickup between 60% and 80%. After padding, the yarns were dried with a Tsuji Senki Kogyo through air oven at 150ëC for 6 minutes and 15 seconds. Using a Sumagh 12 SL 7900 sample loom, the treated yarns were woven into plain weave fabric, utilizing the treated yarns in both the warp and weft directions. Using the four previously described reactive dye colors, color swatches were printed on the fabric using the Mutoh inkjet printer previously described. Lines of the four point sizes, as described previously, were also printed on the fabric at this time, with the Mutoh printer, but only with the black dye. The fabric was then steamed, rinsed, and soaped at the boil, as described previously. The fabric was subjected to line quality analysis after printing, after steaming, and after steaming, rinsing, and soaping to understand how each process affects line quality as well as to determine if solution D survived the weaving process to also function as a print fixative.

8.3

Results and discussion

8.3.1 Size evaluation The first experiments compared traditional size to the current print fixative used for cotton printed with reactive dyes, as well as to untreated fabric. This was

130

Digital printing of textiles

done in order to determine if the traditional size could also function as a print fixative as well as to establish a baseline of print fixation with the control, untreated fabric. Both the control and the starch treated fabrics lost the dye applied to them during the steaming and soaping at the boil post-print treatments, while the sodium carbonate and urea treated fabric retained its dye. This indicates that the desired bond between the polymer chains of the cotton fibers and the reactive dyes did not form in the control and starch treated fabrics, while it did form in the sodium carbonate and urea treated fabric. As the control and starched fabrics did not retain color, their colorfastness properties were not evaluated. However, the sodium carbonate and urea treated fabric was evaluated in terms of its colorfastness properties, with the results listed in Tables 8.1, 8.2, and 8.3. In these tables the grades are based on the AATCC colorfastness scale which ranges from 1 (poor) to 5 (excellent). For the colorfastness evaluation, the fabric was evaluated after 20 hours of exposure and again after 40 hours of exposure. For each exposure time, the four shades exhibited good colorfastness properties, as displayed in Table 8.1. However, with the increase in exposure time the colorfastness properties did degrade slightly. The four shades all displayed excellent resistance to color loss during laundering as depicted in Table 8.2. Table 8.3 displays the results of both Table 8.1 Results from AATCC test method #16: colorfastness to light Sample

Colorfastness to light 20 hours

A: Control B: Cornstarch C: Sodium carbonate/urea formula with thickener and silica

40 hours

N/A N/A Cyan Magenta Yellow Black

5 4±5 5 4±5

5 4 4±5 4

Table 8.2 Results from AATCC test method #61: colorfastness to laundering Sample A: Control B: Cornstarch C: Sodium carbonate/urea formula with thickener and silica

Colorfastness to laundering (color change)

Cyan Magenta Yellow Black

N/A N/A 5 5 5 5

Integration of fabric formation and coloration processes

131

Table 8.3 Results from AATCC test method #116: colorfastness to crocking Sample

Colorfastness to crocking Wet

A: Control B: Cornstarch C: Sodium carbonate/urea formula with thickener and silica

Dry N/A N/A

Cyan Magenta Yellow Black

4 3 4 4

5 5 5 5

wet and dry crocking, or rubbing, evaluations. These tests measure the amount of dye transferred from the test specimen to a white cloth via rubbing. The colors all exhibit excellent resistance to color transfer via crocking when dry, but transfer some color when wet, which may be indicative of some unfixed dye within the fabric. This could lead to problems with staining of lighter colored fabrics under certain conditions, such as drying of recently washed fabrics, if this problem is not remedied. In particular, the decrease in magenta was much greater than that of the other shades. Cotton dyed with shades of red has exhibited difficulty with wet crocking. This is often due to problems affixing the red dye to the polymer chains of the cotton, and the dye remaining in the fiber, unaffixed, until the fiber swells when wet, at which time the dye is released. This problem may be minimized, as could the other decreases in wet crocking values, with a more vigorous soaping process. Overall, the initial size evaluations indicate that traditional size cannot be used to assist in the bonding of reactive dyes to cotton and as such may be ineffective for combining weaving and printing into one process. Additionally, the fabric cannot be untreated prior to printing, but requires a solution such as sodium carbonate, urea, thickener, and silica in order to retain the dye and hence color. The treatment of sodium carbonate, urea, thickener, and silica promotes bonding between the cotton polymer chains and the dye molecules, which in turn provides good to excellent colorfastness properties.

8.3.2 Print fixative as a possible sizing agent Although sodium carbonate, urea, silica, and thickener is the usual formula used to affix reactive dye to cotton in digital inkjet printing, an investigation into variations of this formula was undertaken to determine if modifications could result in similar properties in adhering the dye to the fiber as well as in the area of sizing. To determine the effectiveness of the three different print fixatives as sizing agents, yarns coated with the different formulas were evaluated for the

132

Digital printing of textiles Table 8.4 Tensile properties of yarns treated with different print fixative solutions Substrate

Average tenacity (g/den)

Tenacity standard deviation (g/den)

Average modulus (g/den)

Modulus standard deviation (g/den)

A unabraded A abraded B unabraded B abraded C unabraded C abraded D unabraded D abraded

0.92

0.09

18.95

2.33

0.72

0.08

18.15

3.66

1.08

0.10

12.22

2.06

0.89

0.13

11.51

3.01

1.30

0.11

11.32

2.65

1.15

0.20

13.42

3.03

1.34

0.08

14.24

2.80

1.29

0.12

13.19

2.55

tensile properties of tenacity and modulus, in both abraded and unabraded forms. The abraded yarns were subjected to enough abrasion to simulate that encountered during weaving to determine if the fixatives would function as a typical size and protect the yarns from loss of properties during weaving. The results are depicted in Table 8.4 and Figs 8.4, 8.5, and 8.6, where substrate A refers to the control without any fixative, substrate B refers to yarns padded with sodium carbonate and urea, substrate C refers to yarns coated with sodium carbonate, urea, and thickener, and substrate D refers to yarns treated with sodium carbonate, urea, thickener, and silica. As depicted in Table 8.4 and Figs 8.4 and 8.6, the yarn tenacity decreased with abrasion, which is to be expected. However, the percentage change in tenacity loss has a wide range depending upon the different solutions utilized. The untreated control lost 14.9% of its original tenacity after abrasion, while formula B lost 17.3% of its original tenacity. Yarns treated with formula C lost 11.4% of their tenacity after abrasion compared with unabraded formula C yarns. Yarns treated with formula D had a much lower tenacity loss of 3.7% after abrasion compared with the unabraded yarns. The addition of some sort of solution to the yarns usually provides an increase in tenacity retention compared to the untreated yarns, as the finished solution generally abrades off the yarn prior to the yarn itself abrading. The addition of silica to the solution of sodium carbonate, urea, and thickener to make solution D provides a great increase in tenacity retention compared to the two other solutions evaluated, indicating that

Integration of fabric formation and coloration processes

133

8.4 Average tenacity values of yarns treated with different print fixative solutions.

the silica itself may play an important role in preventing damage to the yarn during abrasion. As with the tenacity evaluations, most of the yarns suffered a decrease in modulus, the initial resistance to deformation, after being abraded. With the exception of formula C, the range of loss in modulus values is not as great as that for tenacity. In terms of modulus, formula A, at 4.2% loss, had the least amount of loss after abrading, compared to the non-abraded yarns. Out of the solutions applied to the yarns, solution B, at 5.8% loss, had the closest modulus loss values to the untreated control, while formula D had a value of 7.7% loss. The abraded yarns of solution C exhibited an increase in modulus compared to the non-abraded yarns. This is unexpected, as abrasion should lead to a decrease in

8.5 Average modulus values of yarns treated with different print fixative solutions.

134

Digital printing of textiles

8.6 Tensile property changes between unabraded and abraded yarns.

modulus. The cause of this abnormality is unknown, but may stem from handling or errors in the abrasion process. It would be expected that the yarns coated with solutions B±D should have a greater modulus than the untreated control, A, as increased modulus generally accompanies increased tenacity. However, in this instance A has greater moduli values in both abraded and unabraded forms. As solutions B±D exhibit lowered unabraded moduli values compared to the control, it is possible that handling of these yarns during padding and drying or in preparation for evaluation degraded the coatings on them, thereby leading to a decrease in stiffness and thus lower moduli values. However, the average values of the non-abraded and abraded yarns within each sample are within one standard deviation of each other, and as such, the exact extent of changes in modulus due to abrasion require further evaluation. Although the change in modulus as a result of abrasion for each set of treatments cannot be effectively gauged, the losses in tenacity values of the different treatments are easier to discern. The yarns with a treatment generally displayed a lower tenacity loss than those without, and among the treated yarns those of substrate D, treated with sodium carbonate, urea, thickener, and silica, had the greatest retention of tenacity values after abrading. Therefore, it can be concluded that formula D may provide the greatest protection from abrasion to the yarns during the weaving process.

8.3.3 Evaluation of different print fixatives After evaluating solutions B±D as possible sizing agents, they were evaluated for their ability to provide quality when printed upon. Although inkjet printing

Integration of fabric formation and coloration processes

135

has been used in conjunction with paper for years, digital inkjet printing upon fabric is relatively new. Print quality of digitally inkjet-printed textiles is often related to the appearance of the print upon the fabric, and line quality analysis plays a major role in determining the quality of the print.5 Line quality often pertains to text quality as they share many of the same desirable properties: line density, sharpness, accurate width, and edge quality.5 ISO, the International Organization for Standardization, has created standard ISO 13660 to address some of these issues. Included in this standard are the properties of line raggedness, line width, and line density. The printed textile samples were compared to printed photo glossy paper, depicted as substrate A in Tables 8.5±8.7 and Figs 8.7±8.9, because this material provides the highest quality print line possible. Therefore, the closer the values of the printed textiles to that of the printed paper, the higher the quality of that print. The results of the various line quality evaluations are depicted in Tables 8.5, 8.6, and 8.7, as well as in Figs 8.7, 8.8, and 8.9. The first evaluation of line quality analysis measured average line raggedness, or straightness, of the four line sizes for solutions B through D in comparison with the glossy paper, A. In three of the four point sizes, solution D obtained average values closest to those measured for the glossy paper. This is extremely important as one of the most crucial roles of the thickener and silica in solution D is to increase viscosity and create as precise a print as possible. Solution C obtained the closest value to the control in the remaining point size not obtained by solution D, and was second in two of the remaining three point sizes. Solution B, which did not contain thickener nor silica, did not allow for the production of a straight line in comparison with glossy paper. These results are expected, as the decrease in line quality follows the decrease in print ingredients from that of solution D to that of solution B, which contains only sodium carbonate and urea. This analysis indicates that thickener and silica are essential in providing decreased line raggedness. However, the results seem to indicate that the lines printed on fabric are not as straight as those printed on paper. This could be related to the hairiness of the surface of the fabric in relation to the smoothness of the paper or to other physical or chemical differences between the substrates. Table 8.5 Line quality analysis results: average line raggedness (m) Point size Substrate A B C D

0.25 pt

0.50 pt

0.75 pt

1.00 pt

7.0 33.9 30.3 21.4

7.8 46.7 31.6 33.5

7.8 41.6 39.3 24.7

8.2 30.6 31.0 19.1

136

Digital printing of textiles

8.7 Line quality analysis results: average line raggedness (m).

The second line quality analysis component measures the average line density, or darkness, of each printed line. Line density is extremely important in digital inkjet printing because it measures the amount and quality of dye that is bonded to the surface of the substrate. A low value of line density is indicative of reduced chemical bonding or mechanical print issues, such as clogged print heads. As expected, the control glossy paper measured the highest values of line density, with solution D possessing averages second highest to the control in two of the four point sizes. The line density value range was much smaller than that of the average line raggedness, yet large enough to see definite strengths by specific solution for each point size. As solution D outperforms the other two solutions in two of the line widths, it appears that the addition of silica assists in adhering the dye to the surface of the fabric. Although the three solutions provide some results similar to that of the control, it is obvious that there is Table 8.6 Line quality analysis results: average line density (OD) for a given line width (pt) Point size Substrate A B C D

0.25 pt

0.50 pt

0.75 pt

1.00 pt

1.54 1.27 1.17 1.08

1.41 1.21 1.15 1.21

1.27 1.01 1.14 1.04

1.16 0.81 1.03 1.04

Integration of fabric formation and coloration processes

137

8.8 Line quality analysis results: average line density (OD) for a given line width (pt).

slightly less bonding between the fibers and the dye then there is between the paper and the dye. Average line width was the third and final print quality property evaluated in this study. Interestingly, this analysis demonstrated the smallest variation between the three solutions. Although the control glossy paper is much closer in width to the actual point size than the remaining samples, the range of averages for solutions B, C, and D is much smaller. Solutions B and D were each closest to the control in two out of the four point sizes. As a result of the close range in data for this portion of the line quality analysis, the results for line raggedness and line density are crucial in determining which solution is most successful. The results of measuring the average line width indicate that although solutions B and D have the potential to provide a line closer in width to the control than Table 8.7 Line quality analysis results: average line width (m) for a given line width (pt) Point size Substrate

0.25 pt

0.50 pt

0.75 pt

1.00 pt

A B C D

374.9 532.0 520.8 480.3

335.9 434.4 425.3 421.5

221.7 333.7 363.8 335.2

152.1 252.1 282.7 267.0

138

Digital printing of textiles

8.9 Line quality analysis results: average line width (m) for a given line width (pt).

that of solution C, all three solutions when printed on appear to allow for wicking of the dye, resulting in larger than desired lines. From the three line quality evaluations it is clear that none of the three solutions provides line qualities quite as good as that of the control, glossy paper. There may be several causes of this, such as fabric hairiness and absorbency compared to the paper, or perhaps differences in the chemical make-up of the treated paper compared to the solutions used on the fabrics. However, given the differences, solution D, composed of water, sodium carbonate, urea, thickener, and silica gives the results closest to the paper in several of the evaluations. It is also more consistent in providing results closer to the paper than the other two solutions. As solution D also provided better results functioning as a size than the other two formulas, it seems that it is a possible link between combining the weaving and printing processes.

8.3.4 The final evaluation Based on the results of all the yarn and fabric evaluations, it was decided to utilize solution D as a combination size and print fixative. In this use, solution D will protect the yarns from abrasion during the weaving process, but also enough of it should remain on the yarn to act as a print fixative during the printing process. To measure this, yarns sized with solution D, as described previously, were woven into a plain weave fabric on the Sumagh loom. These yarns were used in both the warp and weft directions to provide a fabric that contained

Integration of fabric formation and coloration processes

139

solution D on all yarns, in order to provide the best results. After weaving, the fabric was printed, steamed, rinsed, soaped at the boil, and then subjected to a line quality analysis to determine if the print quality changed as a result of a particular treatment. As with previous line evaluations, the results are compared to those obtained for the photo quality glossy paper, reported previously. The results of the different line quality evaluations are depicted in Tables 8.8, 8.9, and 8.10. Table 8.8 Line quality analysis results: average line raggedness (m) Point size Substrate Glossy paper Fabric after printing Fabric after steaming Fabric after soaping at the boil

0.25 pt

0.50 pt

0.75 pt

1.00 pt

7.0 30.9 40.2

7.8 81.3 41.4

7.8 36.0 41.4

8.2 28.6 44.8

36.4

37.7

36.8

36.2

Table 8.9 Line quality analysis results: average line density (OD) for a given line width (pt) Point size Substrate Glossy paper Fabric after printing Fabric after steaming Fabric after soaping at the boil

0.25 pt

0.50 pt

0.75 pt

1.00 pt

1.54 0.65 1.10

1.41 0.73 1.32

1.27 0.73 1.44

1.16 0.80 1.51

0.72

0.83

0.94

0.97

Table 8.10 Line quality analysis results: average line width (m) for a given line width (pt) Point size Substrate

0.25 pt

0.50 pt

0.75 pt

1.00 pt

Glossy paper Fabric after printing Fabric after steaming Fabric after soaping at the boil

374.9 137.7 462.96

335.9 205.6 548.4

221.7 267.7 642.5

152.1 342.9 756.7

196.5

241.2

256.0

239.0

140

Digital printing of textiles

The different evaluations all give similar results in that the line quality values for raggedness, width, and density after printing and after the final fixation process of soaping at the boil are similar. However, after steaming, the values increase. The increase in values between printing on the fabric and after steaming may be due to steaming causing the fibers to swell and hence distort the print, resulting in larger values. The steaming process may also cause unfixed dye to migrate to the surface of the fabric. After rinsing, loose dye that was previously on the surface of the fabric has been removed, leading to lower readings, closer to those of the control glossy paper. Table 8.11 displays the results of crockfastness evaluations for both the initial fabric treated with solution D and then printed as well as the fabric woven with yarns padded with solution D that was then printed. In this evaluation the treatment of the yarns prior to weaving yields increased crockfastness results compared to those of the padded fabric. The cause of this difference in results is currently unknown, but it is possible that since the yarns were treated, dye in the interstices of the fabric was fixed to the yarn, whereas in the padded fabric solution D may not have penetrated the interstices of the fabric, resulting in dye not affixing to the fabric in these areas, and thus being removed during the wet crocking evaluations. Tables 8.12, 8.13, and 8.14 compare the control glossy paper with the fabric treated with solution D and then printed and with the fabric woven with yarns that were treated with solution D and then printed. It must be noted that the two fabrics are different from each other in that they are made of different sized yarns which may influence the results. However, by comparing the different evaluations, it is evident that there are some differences between the two fabrics, in comparison to the control. In the areas of line raggedness and line density the fabric treatment yields closer results to the control than the yarn treatment. This is not unexpected, as the take-up of solution D during yarn padding varied Table 8.11 Comparison of fabric treated with solution D, and yarns treated with solution D woven into fabric: crockfastness Sample

Colorfastness to crocking Wet

Dry

Fabric treated with solution D and printed

Cyan Magenta Yellow Black

4 3 4 4

5 5 5 5

Fabric containing yarns treated with solution D and printed

Cyan Magenta Yellow Black

5 5 5 5

5 5 5 5

Integration of fabric formation and coloration processes

141

Table 8.12 Comparison of control, fabric treated with solution D, and yarns treated with solution D woven into fabric: average line raggedness Point size Substrate Glossy paper Fabric treated with solution D and printed Fabric containing yarns treated with solution D and printed

0.25 pt

0.50 pt

0.75 pt

1.00 pt

7.0

7.8

7.8

8.2

21.4

33.5

24.7

19.1

36.4

37.7

36.8

36.2

Table 8.13 Comparison of control, fabric treated with solution D, and yarns treated with solution D woven into fabric: average line density (OD) for a given line width (pt) Point size Substrate Glossy paper Fabric treated with solution D and printed Fabric containing yarns treated with solution D and printed

0.25 pt

0.50 pt

0.75 pt

1.00 pt

1.54

1.41

1.27

1.16

1.08

1.21

1.04

1.04

0.72

0.83

0.94

0.97

Table 8.14 Comparison of control, fabric treated with solution D, and yarns treated with solution D woven into fabric: average line width (m) for a given line width (pt) Point size Substrate

0.25 pt

0.50 pt

0.75 pt

1.00 pt

Glossy paper Fabric treated with solution D and printed Fabric containing yarns treated with solution D and printed

374.9

335.9

221.7

152.1

480.3

421.5

335.2

267.0

195.6

241.2

256.0

239.0

142

Digital printing of textiles

between 60% and 80%, as opposed to a constant 80% for the fabric padding. This, combined with abrasion of the yarns during the weaving process, would give the yarn treated fabric decreased levels of solution D, and thus result in greater line raggedness and lower line density, or lighter line shades. This trend is also evident in the line width evaluations. However, in this instance, the line width values at 0.75 and 1.0 points of the yarn treated fabric are closer to the control than those of the treated fabric. The cause of this is unknown, but it is likely that lower wet pickup and abrasion from weaving may result in greater width variability.

8.4

Conclusions

The idea of combining weaving and printing may be feasible if a link connecting the two processes can be discovered. In weaving, a size is required to protect the warp yarns from abrasion during the weaving process. For printing with reactive dyes on cotton, a print fixative solution must be applied to the fabric prior to printing, in order to adhere the dye to the polymer chains of the fiber and prevent loss of dye during the steaming and soaping processes. In order to combine the weaving and printing processes, the sizing agent and print fixative agent must be the same compound. This will allow fabric that has just been woven to be immediately printed without requiring any preparation processes. By combining these two processes, handling of the fabric is reduced, which may result in less fabric or print defects due to fabric handling. The combining of the two processes also allows for increased production, as the fabric does not have to be further prepared for printing, nor does it have to be moved manually from one process to another. Various solutions that could be used to combine the weaving and printing processes together have been evaluated. Additionally, printing on fabrics without a treatment was also attempted. Some solutions displayed good results in some areas but poor results in others. However, one solution seems to offer good results in both areas: a mixture of water, sodium carbonate, urea, thickener, and silica. This treatment has been applied to yarns that were then woven into fabric and printed upon. It has functioned as both a size and a print fixative and appears to be successful in both. The fact that this solution has displayed good results in laboratory evaluations as well as in both weaving and printing indicates that it has the potential to link weaving and printing for cotton. A solution of water, sodium carbonate, urea, thickener, and silica appears to be the key to integrating weaving and printing.

Integration of fabric formation and coloration processes

8.5

143

References

1. Rivlin, Joseph, The Dyeing of Textile Fibers, Theory and Practice, 1992, p. 137. 2. Zoomer, Wim, `Get production rolling with rotary screen printing'. Screen Web, http://www.screenweb.com/index.php/channel/4/id/689, 16 December 2002. 3. `New high speed digital textile printer'. Melliand International, March 2003, pp. 68±71. 4. Hudson, Peyton B., Anne C. Clapp, and Darlene Kness, Joseph's Introductory Textile Science, 6th edn. Harcourt Brace Jovanovich, Fort Worth, TX, 1993, pp. 51± 52. 5. Tse, Ming-Kai, and John C. Briggs, `Measuring print quality of digitally printed textiles'. IS&T NIP14 International Conference on Digital Printing Technologies, Toronto, Ontario, Canada, October 1998.

Part II Printer software

9 Digital image design, data encoding and formation of printed images T L D A W S O N , formerly of University of Manchester, UK

9.1

Introduction

In factories, offices, homes and outdoors the past two decades have seen a revolution in the use of digital devices which are now commonplace for audio and video capture, data transmission and playback. The textile printing industry has not been slow to adopt such systems, particularly computer aided design (CAD) to aid pattern production and computer aided manufacture (CAM) to facilitate reproducible machine settings and print paste and screen production.1 Some 90% of all printed textiles are produced on screen printing machines and despite the advent of ink jet printing machines for textiles the initial application of textile CAD±CAM systems was for conventional printing. However, it was Stork's introduction of the first ink jet printer for textiles, with its associated special reactive dye inks in 1991, which offered printers a much faster means of producing sample and pre-production prints for customer approval. Wide format ink jet printers were already well established in the reprographics industry and still continue to gain market share from more conventional colour printing methods. Jet printing may be defined as a process by which the desired pattern with its individual colours is built up by projecting tiny drops of ink of different colours, in predetermined micro-arrays onto the substrate surface, each of these arrays representing one picture element (pixel) of the design. Usually a set of inks is used consisting of at least three primary colours, namely cyan (turquoise), magenta, yellow and optionally black, the so-called CMYK inks known in the reprographics industry as `process colours'. The proportion of each of these primaries in any area of a print determines the perceived overall colour in that particular region. Similar primaries are employed in conventional gravure and offset lithographic printing, although in this case the colours are usually applied as patterns of dots of varying diameter (amplitude modulation) or as uniformly sized dots in various randomised density arrangements (density modulation). As most ink jet printers were originally designed for paper printing, the technical terms encountered such as dots or lines per inch (dpi or lpi) continue to be used,

148

Digital printing of textiles

whereas a textile printer would refer to the screen mesh or raster, and the colours are inks rather than dye liquors or print pastes. It is essential to appreciate that the technology of ink jet printing is fundamentally different from that of all other textile printing techniques, not only because of the non-contact mechanics of the print head but also in the means by which individual colours of the design are produced. Traditional textile printing uses one screen for each colour in a design for which individual print pastes are prepared to match the desired shade. In jet printing a great deal of computation is necessary to produce each of the millions of pixels in a design and this continues for as long as the machine is printing the fabric. In the past printing machines were adjusted entirely by mechanical means using the operator's experience and judgement, and although modern impact printing machines may nowadays be fitted with more refined electromechanical feedback devices, these are still relatively unsophisticated compared with the electronic controls required for jet printers.

9.2

Computer aided design, editing and data storage systems

Print patterns can be produced on many standard graphics-based programmes but, because of the special editing requirements in textile designs, it is normal practice to employ proprietory software which gives not only full design/editing capabilities but can be augmented with many other features to form an integrated colour management system (CMS). In particular the CMS will assist in achieving accurate and reliable reproduction of the colours of the original design by characterising the scanner, monitor display and the jet printing machine to be used (see Chapter 10). More complex design systems are also available which record not only the colour information for printed designs but also the 3D surface texture effects of woven and knitted fabrics. Table 9.1 gives examples of design software available and the computer systems by which they operate. Traditionally the most popular operating platform used for professional reprographics was the Macintosh computer and operating system and this was also true of several early systems offered for textile jet printing. However, in recent years there have been a number of amalgamations within CAD vendors and PC-based systems, with MS Windows XP operating systems now more common. Some design system software offers 3D modelling features such as viewing the effect of the design as a garment or even draped on a figurine with rendered shading effects. Although textile patterns may be produced entirely on the colour graphics monitor of a CAD system using a `Paintbox' or similar software package, with a pressure-sensitive stylus on a digitiser tablet, it is still more common for designers to originate their artwork manually on paper or card with paint- or airbrushes. To obtain a digital image from paper designs suitable for editing, it is

Digital image design, data encoding and printed images

149

Table 9.1 Some CAD systems suitable for textile print pattern editing Company (Website)

Design system

Computer/system base

Aleph (Italy) (http://www.alephteam.com)

Smartcolor

MS Windows

B-Tree (Italy) (http://www.btree.net)

TreePaint

Macintosh/MacOS

DGS Dua Graphics (Italy) Matchprint II (www.ceam-group.it/DuaGraphic)

PC/Linux

DP Innovations (USA) (http://www.DPInnovations.com)

StudioMaster

PC/MS Windows

Lectra (France) (http://www.lectra.com)

Primavision Print

PC/MS Windows

NedGraphics (Holland) (http://www.nedgraphics.com)

Vision Printing Studio

PC/MS Windows

SpeedStep (Germany) (http://www.speedstep.de)

ProPainter 2000

PC/MS Windows

Stork (Holland) (http://stork.com)

Image PC/BestImage

PC/MSWindows

necessary to scan this artwork using a flatbed scanner (up to A3 size) or a rotary drum scanner for the larger A0 sized originals. The larger rotary scanners utilise a xenon light source, red, green and blue (RGB) interference filters and photomultiplier sensors which give a superior signal to noise response compared with the CCD sensor system in flatbed scanners. Rotary scanners require more skill to operate than flatbed ones as the artwork needs to be carefully attached to the rotary scanning drum. Scanners can usually capture an image at up to 2000 dpi or more and opinions vary as to the minimum scan definition required for a textile pattern, but 300± 600 dpi is usual. Image capture is also possible using high resolution digital cameras but this is less common. The image acquired by the camera or scanner may be captured in digital RGB in 24, 32 or 48 bit colour, although for textile designs these are usually stored as a 24-bit Trucolor images (i.e. with eight bits each for the R, G and B values). The captured image can be displayed on a computer monitor or VDU (visual display unit) but a great deal of editing is then required before a final design whose data is suitable for inputting to the printing machine is obtained. A scanner records every brushstroke and nuance of shade and even the smallest blemish in the designer's artwork which therefore requires electronic cleaning up and editing. The design will usually need breaking down into a selected number of colours, and further editing so that it has a defined pattern

150

Digital printing of textiles

repeat. This editing is achieved by the use of proprietory CAD programs, the graphics handling being usually based on well-known editing suites such as Adobe Photoshop, Corel Photo-Paint or Jasc Paintshop Pro. For most textile designs the image then needs to be broken down into a specified number of colours (say a maximum of 8±12), the precise shade of each being then adjusted individually, while stray defects such as isolated small groups of adventitious pixels can also be deleted and outlines cleaned and smoothed. After the colour separation operation the pattern elements of each colour can be displayed as design `layers' which can be examined, rescaled and `fitted' either individually or in any combination with the other layers. By contrast special effects such as tonal photographic designs, shadow and layer effects and those without conventional pattern repeats can all be jet printed and may consist of a very large number of individual shades. Such tonal effects cannot easily be produced by screen printing for which designs are saved as the individual colour separations. If the design is suitable, however, the digital information can also be used to control the making of individual printing screens, either using a conventional photoexposure method with printed film positives or by direct application of special black waxes onto the screen mesh using scanning, piezo-type printheads before the photoexposure stage.1

9.2.1 Capture of digital design data Irrespective of the type of scanner used, the graphics industry has adopted a common communication protocol known as TWAIN (not an acronym), a standard devised in the early 1990s by an industry-wide working group called the TWAIN Coalition. The adoption of this standard provides a consistent integration of image data between all input devices and software applications. Most modern flatbed scanners are single-pass devices using cold cathode fluorescent tubes or LEDs to illuminate the design and either a system of dichroic mirrors or special filters to separate the R,G and B signals before they reach the three linear CCD photosensor arrays. Capture of images by digital cameras differs in that the photosensor array consists of a mosaic of red, green and blue cells usually arranged in a so-called Bayer pattern, there being two G cells for every R or B one.2 The combined data for the individual sensors (known as RAW format) can be stored and subsequently processed on an external computer, but many cameras perform so-called demosaicing calculations `on board', using algorithms of varying complexity, to interpolate the two missing values (e.g. the R and B values at a G cell location) for each pixel. Recently a camera chip has been developed by Foveon and Texas Instruments in which each individual cell gives an RGB response and this obviates the need for demosaicing calculations.

Digital image design, data encoding and printed images

151

9.2.2 Encoding, compression and storage of pattern data Over the years various methods of encoding digital images have been devised.3 Many systems record the RGB values of each pixel location in a specific order (e.g. all MS Windows displays are based on the Bitmap (BMP) display system for pixel data, the RGB values requiring three bytes of data per pixel, being read, line by line, from left to right of a displayed image starting at the bottom). For high definition and large pattern repeats this results in very large files which, apart from requiring a large memory, may result in slower times of access, display or transmission. Another basic method involves run length encoding (RLE) which reads the data, for example as two bytes, in the form [RGB values][number of consecutive pixels with these values]. This may at first sight appear to be little better than the BMP format and indeed it is not if each pixel differs from its neighbours, but is particularly useful for encoding traditional textile designs which tend to have relatively large areas of uniform colour. Some other methods of encoding allow the data to be compressed but it is important that when the full file is restored the recovered data conforms exactly to the original (i.e. that the compression system is lossless). File compression usually takes place in three stages: · Transformation: Lossless mathematical shifting of image signals (e.g. by applying a discrete cosine transformation) · Quantisation: Simplification of image information (in JPEG the image is processed as 8  8 blocks of pixels, which is the main source of loss) · Encoding: The data is usually then subjected to lossless RLE (run length encoding). Most data encoding systems carry much additional information (so-called metadata) in the form of file headers. These identify the encoding system type, file and image sizes, the full list of possible colours in the palette, the start and end file address locations for pattern data, the bit/byte order used (which varies between the Motorola and Intel processors used on Macintosh and IBM computers respectively) and, very importantly, characterisation data such as device profiles (see Chapters 10 and 11). For real-time access, data systems tend to be device-dependent, the result of which is a diverse range of data formats to meet different needs. As a result many different image encoding systems have evolved, some of which (such as BMP itself) are becoming obsolete whilst newer methods are still being devised. Some design and colour management systems employ proprietory file formats which are application-specific, but for maximum flexibility it is preferable to use those (such as TIFF) for which all data is openly readable and therefore usable on a variety of platforms. Some formats allow additional CMS information such as ICC (International Colour Consortium) profile data to be incorporated. Table 9.2 gives some of the commonly used encoding systems with additional comments on their pros and cons.

152

Digital printing of textiles

Table 9.2 Some principal file formats Acronym for Explanation file format BMP TIF(F)

GIF PNG JPEG

TGA EPS PDF RAW DNG

9.3

Bitmap. Format for digital images as basis for all Windows-based graphic displays. Large file sizes. Tag(ged) image file format. Developed in 1987 as a universal file format based on BMP with data `tags'. Commonly used in textile design systems. Presently in Revision 6 (1992). 24 or 48 bit colour in RGB or CMYK. Capable of lossless (5 times) compression using the LZW (Lempel-Ziv-Welch) algorithm. Graphic interchange format. Compact image format (compression 5± 10 times) but limited to 8-bit colour. Widely used on the Internet but being replaced by PNG. Portable networks graphics. ISO/IEC 15948; 2003. Supports 24±48 bit colour. Capable of greater lossless compression than standard TIFF. Joint picture experts group. Strictly speaking JPEG (ISO/IEC 10918; 1991) is not a format but a compact picture file encoding system.4 When JPEG compressed files are coded using TIFF this is known as JTIF. Standard JPEG compression greater than 2±3 times is lossy but a lossless (JPEG-L) system exists. The new, more versatile standard, JPEG2000, is now available.5 Best system for photographic images. Targa (a trade name). Best known as format for storing quickreference, `thumbnail' images. Supports up to 32-bit colour and metadata. RLE compression possible. Encapsulated postscript. Common format for text, graphics and desktop publishing systems and also for printer drivers. Now largely replaced by PDF. Portable document format (Adobe). Frequently used file compression system for both graphics and text on the Internet. Many device-specific RAW file formats have been devised for storing pixel data from camera and scanner sensors but no standards exist, hence proposal for DNG. Digital negative format (Adobe). A suggested standardised means of encoding raw data from camera sensors together with metadata recording the basic camera settings.

Pixel and image formation by ink jet printers

The formation of the tiny ink drops produced by jet printers of either the mechanical pulse (piezo) or thermal (bubblejet) type follows a remarkably similar sequence, and the droplets normally form a roughly circular spot when they land on the substrate surface. On leaving the jet orifice the ink initially tends to form a `tail'. Ideally this collapses into the head of the main droplet which becomes spherical before reaching the substrate. If the applied driving pulse is too strong the tail may be so long as to break up into satellite droplets, a particular tendency with bubblejets. The volume of the droplets varies con-

Digital image design, data encoding and printed images

153

siderably, but is typically 5±10 picolitres (pl, 10ÿ12 litres) depending on the printhead design and the nature of the electrical driving pulse(s). Each pixel of any design being jet printed is actually composed of multiple droplets deposited onto the substrate as an ordered group called a superpixel, usually on the basis of either a 4  4, a 6  6 or an 8  8 matrix, and each of the coloured ink drops can be directed within these matrices. Hexagonal matrix configurations are also used on some printers. Most basic jet printers can project only one drop of each colour into any one or more of the 16, 36 or 64 matrix locations. This is known as a binary system. More sophisticated printers, particularly the continuous ink jet type, can modulate the total amount of ink in any one location within the superpixel matrix by projecting several drops. Some modern thermal and piezo-type printers can also produce such halftone effects by projecting multiple ink drops into each superpixel location. An example of this is the Hewlett-Packard C-REt (Color Resolution Enhancement technology) . Binary and multi-level halftoning is shown diagramatically in Fig. 9.1. Table 9.3 shows the relationship between the matrix size and the number of individual shades that can theoretically be produced with a binary system printer using just cyan, magenta and yellow inks. Jet printers used in reprographic applications can print at up to 1200/1440 dpi but those used for textile printing are normally 300/360 or 600/720 dpi models. There is often confusion between the stated dpi of a printer and the actual definition of the printed design, measured in pixels per inch (ppi), the value of which depends on the size of the superpixel matrix. Thus the printed image definition of a 300 dpi printer using a 4  4 matrix is the same as that of a 600 dpi printing machine using an 8  8 matrix, namely 75 ppi. By comparison the finest screen mesh size (lpi; lines per inch) used for conventional printing is usually120 lpi (48 raster) and for an average design 80±100 lpi (32±40 raster), which can theoretically be equalled with a jet printer operating at 600 dpi and a 6  6 superpixel matrix.

9.1 Binary and multi-level halftoning.

154

Digital printing of textiles

Table 9.3 Grey levels and colours from CMY inks Superpixel matrix

Grey levels (including white)

Number of colours

44

17

4 913

66 (or 6  4 sided hexagon)

37

50 653

88

65

274 625

257

16.7 million

16  16 (or 4  4 with 16 halftone levels per location)

The linear speed at which textiles can be jet printed depends on: · The maximum frequency response of the printheads as they are scanned across the width of the textile fabric (typically 12±25 kHz; modern piezo systems tend to be faster than bubblejets) and whether the print scanning is single-pass or bidirectional · The width of the strip of fabric covered in each scan and hence the distance the fabric is moved forward under the heads after each pass of the printhead · The width of the fabric · The print definition and superpixel matrix size (the greater these are the slower the production rate). Thus high production rates and optimum print quality are mutually exclusive. Although improvements have been achieved, the production speed of the smaller jet printers is still only 0.1±0.3 m/min and for the larger, more costly, units 0.5±1.0 m/min, as compared with say 30±50 m/min for a rotary screen machine. Ultimately the printing speed can be also adversely affected by the rate at which the drive data can be computed and transmitted. It is usual therefore to control one or more jet printers with a dedicated high-specification PC or workstation unit, pattern design and editing being carried out elsewhere.

9.3.1 Drop placement within a superpixel (dithering) To avoid undesirable chevron, moire or mottled colour effects in digital prints, the placing of individual droplets within the superpixels is carefully controlled to give smooth gradations of shade. Figure 9.2 illustrates the simpler case of a monochrome printer where the black ink spots may be placed systematically in an ordered 4  4 matrix arrangement to achieve 17 shade gradations, or grey levels. However, the placing of individual ink drops is usually `dithered' i.e. randomised from pixel to pixel within the matrix. In the case of a colour image each of the R, G and B pixel components are dithered independently. Dithering algorithms vary considerably from one proprietory system to another, but are

Digital image design, data encoding and printed images

155

9.2 Illustrates the simpler case of a monochrome printer.

usually classed as `ordered' (such as the Bayer system) or as `randomised', or `irregular dispersed' dithering, which uses error diffusion computation methods (as in the StuÈcki and Floyd Steinberg methods).6 Originated in 1975, the Floyd Steinberg method is simple and in the public domain, being based on a system in which the input signal intensity of each pixel is compared to a fixed threshold (e.g. a grey level of 50%) and the output signal is generated according to this comparison, with the difference between input and threshold values being distributed to the nearest four neighbouring pixels in predetermined proportions (in the original version these were one-, three-, fiveand seven-sixteenths).7 Several improved versions of Floyd Steinberg have been

156

Digital printing of textiles

evolved, but error diffusion methods in general are slower to compute than ordered dithering, although they yield better quality results with smoother tonal gradations required for photorealistic prints. Without further modification, ordered and some error diffusion dithering methods do, however, tend to produce banded effects at certain grey level values. Figure 9.3 shows a magnified photographic image of a succession of printed superpixels, for four different grey levels in a simple 4  4 matrix, 300 dpi print, with large unprinted areas between each pixel for the sake of clarity. Note the variations in both the placing and the total number of the ink spots between adjacent superpixels resulting from dithering. The greater the number of drops

9.3 A magnified photographic image of a succession of printed superpixels.

Digital image design, data encoding and printed images

157

9.4 Selected dyes to extend the limit of the colour gamut of reactive dye printing inks.

of ink placed into each superpixel the deeper the depth of shade, but the relationship between these is not linear, and it is observed that when the drops saturate the total area of the superpixel so that no white substrate remains, there is little further increase in the strength of the print (see Fig. 9.4) It therefore follows that to attain optimum build-up of colour the drops should be small and should suffer minimum lateral spreading on the fabric surface, whilst at the same time being absorbed to a limited extent into the substrate to aid drying, all of which are helped considerably by pretreating the fabric (see Chapter 12). As the number of drops in each pixel increases some in effect fall on areas prewetted by other ink drops, which leads to the overall area covered by the superpixel increasing, giving rise to a so-called dot gain effect with a consequent increase in depth of shade. When CMY inks are used to produce neutral grey shades (i.e. when R = G = B) it might be supposed that the number of spots of each ink present in the superpixel matrix would be equal, but this is not so because of differences in the manner in which colour values of the cyan, magenta and yellow increase with increasing depth of shade which is allowed for by the printer driver software.

9.3.2 Colour gamuts that can be attained using CMY inks The range of shades in a design that can be perceived on a VDU display (where the colour mixing of light from the RGB phosphors is additive) is much wider than can be achieved by dyeing or printing a fabric and wider still when compared to the colours that can be jet printed. The extent of a colour gamut should not be confused with the number of individual shades that can, at least theoretically, be produced that lie within the confines of that gamut. Thus the use of standard CMY inks with a black ink or a weaker strength cyan or magenta

158

Digital printing of textiles

ink (e.g. when using so-called photo ink cartridges) does not increase the extent or boundary of the colour gamut, only the number of possible individual colours from within the gamut. With a suitably pretreated fabric and in pale shades, the ink drops deposited in a superpixel remain separate and distinct and the colours produced are governed by the proportional area of the matrix covered by each of the primaries, which is the principle of colour production in reprographics printing and is known as partitive colour mixing. As the depth of shade increases and more drops of ink fall into any one superpixel the spots from each of the primary shade inks tend to impinge on one another and to spread. The range or gamut of shades attainable by the same three primaries in partitive colour mixing is rather more limited than that of subtractive mixing, which, of course, occurs in all conventional textile printing where colours are premixed. The mixing process within a matrix of ink spots has been studied on paper substrates and a unifying predictive theory developed. A practical trial of this theory showed good correlation between predicted colour values and those actually produced on two commercial paper printers, the software for which employed two dithering methods.8 As the depth of shade increases, individual ink spots spread increasingly as they impinge on locations occupied by other spots, leading to the formation of increasingly large superpixels. This can, however, be allowed for in mathematical calculations of the colours attained.9 On textile substrates, even when pretreated, ink drops tend to spread more than they would on a coated paper, and on woven fabrics the spread is most pronounced in the warp and weft directions due to capillary action in the fibre bundles. Where a jet printer also has a black ink supply this may be used in two ways, namely purely to produce dense black shades or as a fourth component, along with the CMY inks, for the duller shades. In fact a combination of CMY inks alone usually produces a poor brownish-black shade. As explained in Chapter 10, it is necessary to introduce still further colours to extend the gamut limits, by providing inter alia bright orange, red, blue and green inks. Most wide printers for textiles can now operate with up to seven or eight different inks, thereby attaining a gamut which is considerably extended in three critical directions, namely blue/violet/purple, turquoise/green/yellowish-green and red/scarlet/ orange. This places an additional burden on the printer driver software (RIP or CMS) as it is necessary, depending on the target shade characteristics, for the appropriate ink combination to be selected.10 The theoretical extent of the colour gamut for any combination of colour primaries can be determined by a number of mathematical methods of varying complexity.11 Suitable algorithms can therefore be incorporated into graphics editing software and proprietory CMSs which automatically indicate when an `out of gamut' shade in a design is selected and the nearest shade match (depending on the desired rendering intent) can be illustrated on the computer

Digital image design, data encoding and printed images

159

VDU, usually as either a 2D or 3D display. The extent of any achievable gamut depends on the colour characteristics of the primaries chosen, which varies with the class of dye used in the inks. In general reactive and acid dyes tend to yield a wider gamut than disperse and pigment formulations.

9.4

Control of the printing machine

The control of even the simplest computer-driven printer is achieved by software known as a printer driver, supplied via either the computer's software (usually IBM or Macintosh) or printer manufacturer or vendor. A printer driver takes the alpha-numeric or graphics file data and together with other user-specified information (e.g. required definition, single or multi-pass printing options and substrate settings) converts it into output data, which are `spooled' and when desired transmitted for further processing to the printer's microprocessor/ memory. The final instructions from the onboard processor and its memory module then control the electromechanical devices and the printhead firing systems within the printer. Although the drive software supplied by the printer manufacturer can often produce acceptable results, it is more common to use a proprietory RIP (raster image processor) package which offers much faster preprocessing and transmission of image data ± indeed some can translate the data `on the fly' as printing proceeds. Such systems require a correspondingly fast computer/printer communication for which an IEEE-1394 `Firewire' serial link usually replaces the more conventional parallel or serial USB (universal serial bus) printer connection. RIPs also include many other control features which to some degree duplicate those offered by CMSs. Whatever the software used there are usually a number of options that can be selected before printing begins, chief among which is to choose settings appropriate to the substrate, for its nature has a strong influence on the results. Particularly when using inks based on soluble (acid and reactive) dyes, the fabric needs pretreatment with certain chemicals, such as acids and alkalis, which are necessary for efficient dye fixation, yielding finished prints having optimum fastness properties. Film-forming agents are also applied at the pretreatment stage so that the best colour yield on a particular fibre is achieved, particularly in heavy shades. Thus colour yield can be almost doubled when compared to the use of an untreated substrate.12 The most universally applicable jet printing coloration system involves the use of pigments, but the colour yields attained depend very much on the choice of the binder and on the processing sequence selected.13

9.5

Machine performance monitoring

In general machines with thermal inkjet printheads (e.g. Canon, Encad, ColorSpan) can be made with a closer packing density of the jet orifices than

160

Digital printing of textiles

is possible with piezo heads such as are used in Mimaki, Ichinose, Reggiani and Robustelli machines, but in all jet printers it is clearly essential that they should run with minimum trouble from jets becoming inoperative. All machines incorporate an automatic flushing cycle for the jets at the start of a run and also at intervals during print runs to avoid problems of unsatisfactory performance which may, for example, arise because loose fibre lints from the fabric have adhered to the jet orifice plates. More seriously, individual jets in a thermal printer can eventually suffer total failure when the tiny heaters (which reach peak temperatures around 400ëC) burn out or become blocked by solid burnt-on deposits from an ink. Some manufacturers give their printheads a guaranteed minimum life, e.g. Stork guarantee the jets in the Amber and Zircon printers for 3000 hours (say 6±12 months), although the replacement of the piezo printheads would require a technician. By contrast Encad (Lexmark) print cartridges can be quickly replaced by the machine operator and are guaranteed to deliver 500 ml of ink. Some bubblejet cartridges are fitted with an attached supervisory electronic chip which signals when this point has been reached. When changing from one ink type to another, piezo printheads can be flushed with a special cleaning fluid and the ink supply bottles changed quickly, whilst on bubblejet printers each printhead cartridge/reservoir system is unclipped and changed. Inoperative jets, which are usually seen to produce stripiness in the weft direction of the printed fabric, are most noticeable when printing dark shades at higher production/lower definition machine settings and when two or more adjacent jets are faulty.11 It may seem rather surprising that such faults can be seen with a printer having a minimum drop definition of 300 dpi, but it must be remembered that such a printer may be giving a true pattern definition of only about 75 ppi. It is therefore prudent at convenient intervals to print special test patterns which will highlight any such jetting failures. The test patterns, which are usually made available as part of the printer driver software, may either be examined visually or be used as part of an automated system check on some machines such as the ColorSpan Fabrijet XII and Encad 700 series. In general such faults are less likely to be observed with 600/720 dpi printers which also have the advantage of showing less colour mottling with some dithering methods.

9.6

Future trends

Any major improvements in digital printing of textiles are likely to come from machinery developments rather than from the design, encoding or operating software. From the electronic operating point of view modern computers have very fast processors and large memory (random access and fixed disc) capabilities, and can be fitted with high capacity colour processing cards and fast external communication systems. On the other hand, those jet printers offering

Digital image design, data encoding and printed images

161

improved production capabilities (1±3 m/min) that have been introduced, such as the Canon TPU (multihead bubblejet system), the Reggiani DReAM (Scitex/ Aprion piezo technology) and Zimmer Chromotex 2003 (with Jemtex continuous ink jet printheads) appear to be less economically attractive than lower productivity machines, such as the Mimaki and Ichinose, originally developed for the reprographics market. In general ink jet printing still continues to exploit only niche markets (sampling, strike-off, haute couture and customisation) rather than the short run/rapid response printing business for which the technology could be ideally suited. Unfortunately the `weaving shed' concept of a factory filled with relatively low cost/productivity machines does not appear to have proved sufficiently attractive to any manufacturer, with the exception of Seiren in Japan. Non-contact, white light phase body measuring systems are now available which, with suitable software and in conjunction with improved pigment jet printing and automated garment panel cutting technology, may yet see the ultimate dream of totally in-store customisation of garments realised.14 From a purely design point of view digital printing methods will continue to give a new freedom to designers who are no longer shackled with the conventions and mechanical constraints of screen printing. Thus, quite apart from the lack of colour and pattern repeat constraints and the ability to produce photorealistic effects, designers can now introduce novel shadow, moireÂ, textured, blurred and layered effects and in general greater individualism.

9.7

Sources of further information and advice

File formats http://www.jpeg.org http://partners.adobe.com/public/developer/en/tiff/TIFFphotoshop.pdf http://www.faqs.org/faqs/graphics/fileformat-faqs

Dithering techniques http://photo.epfl.ch/workshop/wks96 http://www.wellesley.edu/pmetaxas/pck50-metaxas.pdf

Colour reproduction http://www.poynton.com/PDFs/Guided_tour.pdf http://www.nmnh.si.edu/cris/techrpts/imagopts

Gamut mapping http://www.cie.co.at/publ/abst/156-04.html http:/www.colour.org/tc8-03

162

Digital printing of textiles

Books C.W. Brown and B.J. Shepherd, Graphics file formats: Reference and guide, Greenwich, Manning Publishing, 1995. R.W.G. Hunt, The reproduction of colour, 6th edn, Bognor Regis, John Wiley, 2004. A. McNamara and P. Snelling, Design and practice for printed textiles, Oxford, Oxford University Press, 1996. J.D. Murray and W. van Ryper, Encyclopedia of graphic file formats, 2nd edn, Boston, Addison-Wesley Professional, 1999.

9.8

References

1. T.L. Dawson, `The use of digital systems in textile printing', in L.W.C. Miles (ed.), Textile printing, Bradford, Society of Dyers and Colourists, 2003, 301. 2. B.E. Bayer (Eastman Kodak), US Patent 3,971,065 (1973), `Color imaging array'. 3. T.L. Dawson, `Ink-jet printing of textiles under the microscope', JSDC, 116 (2000), 52. 4. W.B. Pennebaker and J.L. Mitchell, JPEG image compression standard, New York, Van Nostrand, 1992. 5. D. Santa-Cruz, R. Grosbois and T. Ebrahimi, in T. Ebrahimi, C. Christopoulos and D. Lee (eds.), Signal processing: Image communication, New York, Elsevier, 2002, 113. 6. I. Kabir, High performance computer imaging, Greenwich, Manning Publications, 1996, 446. 7. R.W. Floyd and L. Steinberg, `An adaptive algorithm for spacial grey scale', Proc. Soc. Inf. Display, 17 (1976), 75. 8. P. Emmel and R.D. Hersch, `A unified model for color prediction of halftoned prints', J. Imag. Sci. Technol., 44 (2000), 351. 9. P. Emmel and R.D. Hersch, J. Imag. Sci. Technol., 46 (2002), 237. 10. V. Ostromoukhov, `Chromaticity gamut enhancement by heptatone multi-color printing', SPIE Proc., 1909 (1993), 139. 11. J. Morovic and M.R. Luo, `The fundamentals of gamut mapping: A survey', J. Imag. Sci. Technol., 45 (2001), 283. 12. T.L. Dawson, `Spots before the eyes: Can ink jet printers match expectations?', Color Technol., 117 (2001), 185. 13. U. Hees, M. Frechte, J. Provost, M. Kluge and J. Weiser, `Ink±textile interactions in ink jet printing ± The role of pretreatments', in T.L. Dawson and B. Glover (eds), Textile ink jet printing, Bradford, Society of Dyers and Colourists, 2004, 57. 14. D. Bruner, `An introduction to the body measurement system for mass customised clothing', http://techexchange.com/thelibrary/bmsdes.html.

10 Digital colour management T L D A W S O N , formerly of University of Manchester, UK

10.1 Introduction When digital colour processing was first introduced into the reprographics industry it soon became evident that some means was required of controlling the process by which digital data from original artwork which had been computer generated or scanned could be transferred reproducibly to paper by the printing process, other than by using a tedious trial and error procedure. Equally colours viewed on a computer monitor were often a very poor guide to those finally achieved in the printed material. The basic reason for this problem lies in the fact that each of the complex physical processes involved in perceiving, capturing, displaying and reproducing colours varies, hence the difficulty in attaining a result which is consistently acceptable to the viewer. The problem may be summarised as follows: · The observer's vision. Light (reflected from the pattern/print or transmitted from a monitor) is focused on the retina which sends electrical responses to the brain from the red, green and blue cone receptors, although it is nowadays more correct to refer to the cones having a response to long, medium and short wavelength radiation. · Digital image capture. Image is captured as a series of RGB (red, green and blue) cell responses forming a matrix (digital camera) or raster pattern (scanner). · Computer monitor display. Conventional CRT (cathode ray tube) display screens have triads of RGB phosphor dots which are activated by three modulated electron beams. A TFT LCD (thin film transistor liquid crystal display) monitor screen is composed of a matrix of individually addressable translucent cells with RGB filters and has a white backlighting system. · Digital colour printer. Uses CMYK (cyan, magenta, yellow and optionally black) printing ink primaries which, in most digital printers, are jetted onto the substrate as a `superpixel' matrix of ink spots (see Chapter 9).

164

Digital printing of textiles

In all the above cases there will probably also be differences in ambient illumination, its intensity and colour temperature required (CIE Illuminant D65 for textile and D50 for reprographic applications) and in the backgrounds against which patterns or monitor screens are viewed. Computer systems for acquiring, displaying and printing colours digitally normally employ an overall colour management system (CMS) to assist the transfer of colour data from image capture through to the final printing system, a need which was first recognised by the reprographics industry and a number of photographic companies (Kodak, Agfa, Fuji), computer hardware/software providers (Microsoft, IBM, Apple Macintosh), print machinery manufacturers (Heidelberg, LinoColor, Scitex) and suppliers of colour control equipment (Barco, X-Rite, GretagMacbeth) amongst others, each of whom have marketed software which allows cross-system transfer of data. In recent years modified versions of the reprographic CMSs have been further adapted and incorporated into textile inkjet printing control software. Table 10.1 gives some examples of CMSs of varying degrees of sophistication. Most systems require the user to carry out their own measurements with a suitable spectrophotometer whilst others offer measurement and profiling services. The main purpose of the CMS is to provide a control system by which the measured colour data of a design may be reliably and accurately transformed into output data for display on a monitor or as input to a printer, so that the appearance of these outputs reliably represents that of the input design to the observer. This involves a number of colour data transpositions which are summarised in Fig. 10.1. In addition to ensuring colour consistency between the various input and output devices and providing software to drive a range of digital printers, a CMS will often provide other useful features such as

10.1 Summary of colour data transpositions.

Digital colour management

165

Table 10.1 Colour management systems Manufacturer (Website: http://)

Product

Platform

Agfa (www.agfa.com/graphics)

ColorTune CMM

MacOS

Aleph (www.alephteam.com)

Newton

MS Windows

Chromix (www.chromix.com)

ColorValet

MS Windows/MacOS

ColorSavvy (www.colorsavvy.com)

SavvyProfile suite

MS Windows/MacOS

Color Solutions (www.color.com)

ColorBlind Prove-It

MacOS

Colorburst Systems (www.colorburstip.com)

Colorburst Pro

MS Windows

Colorvision/Datacolor (www.colorcal.com)

ProfilerPRO, DoctorPRO

MS Windows/MacOS

Ergosoft (www.ergosoftus.com)

TexPrint

MS Windows

Fujifilm (www.colorprofiling.com)

ColorKit profiler

MS Windows/MacOS

GretagMacbeth (www.gretagmacbeth.com/i1)

NetProfiler, Profilemaker

MS Windows

Heidelberg (www.heidelbergusa.com

Prinect Calibrator/Profiler

MS Windows

Kodak Polychrome (www.kpgraphics.com)

Matchprint

MacOS

Pantone (www.pantone.com)

ColorVision and ColorPlus

MS Windows/MacOS

X-rite/Monaco (www.xrite.com)

MonacoEZcolor & Profiler

MS Windows/MacOS

organising the `queue' of print jobs. The CMS may also provide means for communicating design and colour data via the Internet. It is often the case, particularly with longer print runs, that the digital design, having been sampled/proofed on a digital printer, is then screen printed with print pastes containing completely different dyes. In such cases, having already obtained spectrophotometric data on each colour in the design, an additional feature of the CMS may be to provide recipe predictions for the conventional print production. To be able to do this the system needs to be provided with

166

Digital printing of textiles

colorimetric data for a range of calibration prints for each dye which may be used. The spectrophotometer used for device calibrations can, of course, be used for this purpose. To further understand the need for colour management systems and how they operate, it is first necessary to appreciate the various ways by which any colour may be unequivocally specified and what functions the various proprietary CMSs can carry out to ensure reliable colour reproduction.

10.2 General numerical colour specifications Any colour may be specified by three coordinates that locate its position in a three-dimensional colour space, which is, however, often represented graphically in two dimensions or as a planar projection. There are a number of standard CIE (Commission Internationale de l'Eclairage) colour spaces, each varying in its overall uniformity and each having its own coordinates.1,2 Three commonly used colour spaces are determined as follows: · CIE xy colour coordinates: XYZ or xyY (usually depicted as a 2-D, x/y plot). The total range of this colour space represents the limits of human vision. · CIELAB colour coordinates: L*a*b*, a visually more uniform colour space usually displayed as a 2-D, a*/b* plot. · CIELCH colour coordinates: LCH (lightness, chroma, hue) sometimes used as an L/C plot to show the chromatic build-up of a particular colour. Figure 10.2 is an illustration of the use of the 1931 CIE x/y chromaticity diagram to illustrate that the range or gamut of shades which can be produced on a typical monitor screen and especially an HDTV (high definition television; SMPTE 240M) display is considerably wider than that which can be achieved by an inkjet printing device, particularly if the printer is using only CMY primaries. When processing device-dependent colour data it is usual to compute with 8 bits per channel even though the colours may have been measured with as many as 16 bits per channel. Thus with 8 bits each for RGB (24-bit colour) and CMYK (32-bit colour), there are 256 grey levels for each primary. This yields a theoretically possible 16.7 million (224 ) colours but human observers cannot in fact distinguish differences between all these colours. The simple relationships between RGB and CMY in terms of their digital values are: R = 255 ÿ C

G = 255 ÿ M

B = 255 ÿ Y

Thus R, G and B = 0 and C, M and Y (or K) = 255 represents black and reversing these values gives white or, in a print, the colour of the substrate. Such a simple translation from the RGB values of the colours in a design which has been captured by a scanner into CMY values for a printed image would, however, produce a very unsatisfactory reproduction of the shades of the

Digital colour management

167

10.2 An illustration of the use of the 1931 CIE x/y chromaticity diagram.

original image owing to the wide variations in the colour reproduction characteristics that always exist between any input and output device. Thus in an ink jet printer there is no simple linear relationship between the digital RGB or CMY values and the hue or intensity of the printed colour. In practice conversion from RGB to CMY values is by the use of complex polynomial equations or other non-linear transformations3 and is therefore usually achieved indirectly from colour measurement of target shades which allow a three-dimensional, reference LUT (look-up table) to be constructed. If in addition there is a requirement that any colour images displayed on a monitor should appear the same on a different display and should also accurately represent those that will eventually be achieved by printing, still further calibration and characterisation of the equipment will be required. All such controls of scanners, cameras, displays and printers require mathematical computation of varying degrees of complexity and it is the critical role of the CMS to facilitate this. Each type of colour-capture, display or reproduction device processes colour in its own particular colour space, and in order to be able to accurately transform every colour (or know that a particular colour is `out of gamut') it is necessary to devise a common, device-independent, colour space. Over the years a number of colour spaces have been proposed, some of which are quite extensive (such as Adobe sRGB) relative to those sometimes used in CMSs. Thus the sRGB (IEC1966-2.1) colour space4 which has been adopted internationally is used in the ICC Profile format specification5 which is

168

Digital printing of textiles

discussed in Chapter 11. The extent of the sRGB colour space is very similar to that of a typical computer monitor, or a standard European system TV display, illustrated in Fig. 10.2. When employing ICC Profiles the device-independent colour space is referred to as a Profile Connection Space (PCS) and the software which processes the data is a Colour Matching Module (CMM) or the Colour Engine. For their part the two major computer software suppliers, Microsoft and Apple, provide their own internal colour management controls as part of the operating systems, namely ICM (image colour management) and ColorSync, respectively, but if a third party CMS is to be used successfully these native management systems must be disabled. The extent of the sRGB colour gamut is very similar to that of a typical colour monitor whether it be a conventional CRT (cathode ray tube) or LCD (liquid crystal display) type (Fig. 10.2). The transformation of RGB values from either the monitor or scanner responses into the XYZ values of the sRGB colour space is carried out in two stages, firstly by using non-linear gamma corrections (power functions of the type R , G and B ) for each R, G and B response and then applying a linear 3  3 matrix transform to these linearised RGB values of the following general type: 2 3 2 3 2 3 X a b c R 6 7 6 7 6 7 4Y5 ˆ 4d e f 5  4G5 Z

g

h

i

B

Note that the RGB values in this matrix are `normalised', so, for example, a linearised digital value of 200 corresponds to 200/255, i.e. 0.78. In this type of matrix some of the coefficients, a to i, can be negative and XYZ colours will be out of gamut when one or more of the calculated RGB values is negative or greater than unity. If there are differences in the viewing conditions relating to the input and output (such as the reference white point of the display and that which relates to the print on textile) then a chromatic adaptation transform must be applied to the data. Chromatic adaptation and other appearance-modelling transforms usually take a similar mathematical matrix form.6 In practice the calculations can either be run in real time by the computer or be stored as LUTs containing, for example, a matrix of RGB values for a particular device which can be read off as the corresponding deviceindependent L*a*b* values. LUTs do not usually comprise the full 16.7 million colour combinations, so the software computes some colours by a process of interpolation.7

10.3 Characterising display, input and output devices In order to achieve consistent display, acquisition and reproduction of textile designs, it is essential to both calibrate and characterise the various pieces of equipment and thereafter to repeat these controls at regular intervals because of

Digital colour management

169

the changes that may occur, for example, in sensor responses, light source emissions and for printers, in the standardisation of the ink supplies and particularly the properties of the textile substrates. The three types of device require a somewhat different approach in the manner in which they are controlled.

10.3.1 Control of colour display monitors The screen of a CRT (cathode ray tube) monitor consists of tiny R, G and B phosphor spots which are caused to fluoresce when hit by the scanning electron beam, the intensity of light emitted being related to the drive voltages applied to the grid of each electron gun. These voltages in turn are generated according to digital values fed via a DAC (digital to analogue converter chip) and video amplifier on the computer's video card. The relationship between the light intensity obtained on the screen and the initial digital RGB inputs is not linear but, apart from an initial offset, follows a mathematical power relationship defined as the gamma value. The term gamma value can be confusing since it is also used in computer operating systems where the standard colour space gammas are usually set at 1.8 (MacOS) or 2.2 (MS Windows). The `native' gamma ( ) of a CRT, which usually lies between 2.5 and 3.0, is intrinsic to a particular tube and relates the gun voltage DAC level (D) to the resulting screen luminance (L), as defined by the following relationship: L ˆ KD To add to the confusion over gamma values, gamma corrections can be applied at other stages, for example when a scanner acquires an image or when data is processed by a video card. When an image is saved it is possible for the gamma which has been applied to be stored within the file, but this is possible only for certain formats such as TIF, TGA and PNG (see Chapter 9, Table 9.2). The mathematical basis for the characterisation and standardisation of a conventional monitor based on a CRT display has been well established8 but requires additional equipment and may appear at first sight to be a somewhat laborious operation. Accordingly some simpler systems have been devised, such as Adobe Gamma, Praxisoft WiziWYG, ColorWizzard and GretagMacbeth WebSync, which allow visual setting of contrast/brightness and the red/green colour signals by quick, on-screen adjustments. Similarly `generic' colour profiles for individual monitors are available from the manufacturers. It is also possible to adjust the individual RGB gamma settings using the videocard software control system but this is a very trial and error method. All such visual adjustments give at best only a limited improvement in the accuracy of the display, but in fact for some users there appears to be little need for the computer's display to be closely controlled. On the other hand, some firms employ colour communication software which ensures that their customers can

170

Digital printing of textiles

specify and approve shades digitally from on-screen appearance (using systems such as Datacolor International's ImageMaster), and for such requirements there is no alternative but to use full colorimetric characterisation of all displays used in the communication network. Initially the monitor, having been left on for a considerable time to become stable, is calibrated, before the characterisation stage. Calibration is carried out by adjusting the contrast control (setting the gun amplifier's offset so that a zero digital D value produces zero luminance of the screen) and the brightness control (which paradoxically sets the black level). The gamma level affects mid tones to the greatest degree as the black and white settings now become fixed points in the gamma curves (see Fig. 10.3). The final calibration requires that the colour temperature of the white point (when R, G and B = 255) corresponds to 6500K (Illuminant D65) which is nowadays the standard for most generalpurpose VDUs. With some of the more expensive monitors there are built-in measuring devices and associated software which can achieve these measurements automatically, whilst with other systems the spectrophotometer is interfaced directly via a USB (universal serial bus) connection to the computer, allowing direct control of the monitor display. For accurate characterisation of a general-purpose CRT or LCD monitor, it is essential to use some type of colorimeter or spectrophotometer which can measure the colour coordinates of areas of colour displayed on the screen. These are usually small portable spectrophotometers such as Spectrolino and Eye-one Pro (GretagMacbeth) and particularly the modestly priced ColorSpyder (Colorvision/ Datacolor) and MonacoOPTIX (X-rite) colorimeters that can be placed against an LCD monitor screen without distorting its extremely thin glass sandwich

10.3 Gamma response curves.

Digital colour management

171

construction. The Spyder and Optix devices utilise the output from a series of photodiodes (seven and 22 respectively) each fitted with a filter so as to measure screen luminance over a range of wavelengths when coloured test patches are displayed. In addition to the use of surface contact colorimeters it is also possible to use more sophisticated, remote-sensing spectroradiometers such as the EyeBeamer (GretagMacbeth) and the Bentham TP300. The generation of the test colour patches is controlled by the individual CMS program selected which compares the measured colour coordinates of the test colours with the input values and from this data compiles a monitor profile conforming to the ICC specification. In recent years the technical performance (with respect to spatial colour uniformity and viewing angle sensitivity) and the price level of TFT LCDs has improved considerably, and LCDs are increasingly replacing the more bulky CRT-based monitors. Initially their calibration and characterisation proved somewhat difficult. TFT LCDs are transmissive devices fitted with a white backlighting system which must ensure that when the RGB filter arrays are set equally (R = B = G = 255) the resultant white point corresponds to the D65 standard. Several workers have shown that the gamma-correction model which is satisfactory for CRT characterisation is not adequate to describe the response of LCDs, as these follow an S-shaped function curve (as indeed does the output from a printer). The colour reproduction characteristics of LCDs also differ from those of CRTs in that the white point colour temperature increases as luminance decreases, an effect known as `grey tracking'.9 This can be corrected by modifying the video look-up table (LUT) of the display device driver. There are also problems in that the chromaticity of the LC cell display changes with applied voltage, which affects the grey level, and there can also be a cross-talk effect between adjacent cells. Finally, although the performance of LCD displays has been greatly improved, there are still limitations in the screen viewing angle. Despite these problems the range of shades (colour gamut) which can be displayed on an LCD is wider than that of a CRT-based monitor and, by choosing a suitable characterisation model, almost as good colour management of LCDs can now be achieved.10 Suitable software for LCD characterisation has now been included in most CMSs. Once characterised, all display devices need checking at regular intervals to overcome time-related `drift'. Some firms do this every week as a matter of routine whilst a once-a-month control system is advisable. Scanners tend to be more stable but still require checking from time to time, as do printers, particularly as differences will arise if there are any changes in the `standard' substrates or inks being used.

10.3.2 Characterisation of input devices For the acquisition of colour design data the most commonly used device is a scanner, usually employing a CCD or CMOS sensor array with RGB filters. Less

172

Digital printing of textiles

commonly the input may be taken from a digital camera, again with an RGB array but the individual pixels are often in a Bayer mosaic arrangement with two G for every R or B cell.11 The colour data can be downloaded in this form as an RAW format file, or alternatively a demosaicing algorithm can be performed to give interpolated RGB data for every pixel. As with monitors, the scanner should first be calibrated, although these devices do tend to run very consistently over long periods of time and to maintain the settings incorporated by the manufacturer. The most direct and efficient way to characterise a scanner is to scan a special test image comprising both a grey scale and a wide range of colour patches, for each of which the colour coordinates are known or can be measured. The IT8.7/2 to 7/4 series of reflective colour targets are produced by Kodak, Fuji and Agfa to ANSI standards. For example the IT8.7/3 (ISO 12642) card has 928 colour patches and is commonly used rather than the much less extensive 20 shades of the GretagMacbeth ColorChecker card. For an extension of this range of shades (1485) the ECI 2002 (European Color Initiative) colour targets may be used. The colour coordinates of all the colours of the IT8 target as issued are closely controlled and the colour values (in terms of XYZ, L*a*b* and LCH) can be downloaded, for example from the Fuji website (http:// www.colorprofiling.com). For greater accuracy they may also be measured automatically using a small reflectance spectrophotometer such as the Spectrolino/Spectroscan (GretagMacbeth) or PULSE (X-Rite) devices. Such spectrophotometers may be used either with a software-driven, actuated support table or as a manually operated scanning device in conjunction with a guide frame which fits over the target shade card. Both types of device can complete the scanning and storage of the very extensive test patch data in only a few minutes. Using a sufficiently large number of test colours, an LUT can be constructed relating the captured RGB values to the actual tristimulus values for each shade, with interpolation for intermediate points not included in the LUT data. Clearly the greater the number of test patterns actually measured, the less need there is for interpolated data. Other, often less precise, mathematical methods of characterising both scanners and cameras have been described.12,13

10.3.3 Characterisation of printers If a suitable scanner has first been characterised, it is possible to characterise a printer indirectly by first scanning an IT8 target card using the scanner profile. This image is then printed with any generic printer profile controls switched off. With suitable software provided from within the CMM it is then possible to amend the values in the LUT for the printer and also relate these values back to L*a*b* values in the sRGB device-independent colour space. Alternatively the data for the IT8 shades can be input to the computer display (which in some

Digital colour management

173

systems may be done by scanning the IT8 target and the test print simultaneously), or the resulting image printed and the colour coordinates for each printed shade then measured spectrophotometrically. Comparing the input and output colorimetric data allows an LUT specific to that printer to be constructed. In all such characterisation work it is essential to record details of the substrates which are printed, because although some degree of control of the total amount of ink which can be applied to the fabric surface can be achieved via the printer driver software, variations in chroma, lightness and sometimes hue will certainly be evident when the substrate is changed. Similarly a different characterisation procedure must be carried out for each dye/fibre type combination and also for the same dye and fibre if, as is sometimes done for convenience, the shade specified may be either for the printed shades before as well as after the dye fixation (and possibly, washing-off) operations, a situation that has no comparison in reprographics printing. There still remain a variety of mathematical characterisation methods for printers ranging from simple first-order masking systems using linear matrix transforms to higher-order polynomial solutions based on the halftone process whereby the CMY colour spots are treated as producing colour in an additive manner according to the relative area each group covers within each pixel and the light-scattering properties of the substrate, following the colour prediction models of Neugebauer and Kubelka-Munk.14

10.4 Colour gamut and rendering intent Figure 10.4 illustrates the overall range of shades (i.e. the gamut) that can be achieved when printing with typical reactive dye based CMY inks, but in this case for clarity the representation is on a CIELAB diagram in which the colour distribution is much more uniform than in a CIE chromaticity plot. The extent of the gamut can be considerably expanded if additional inks, comprising brighter primaries such as orange, yellowish-red, reddish-blue and yellowish-green, are selected. This idea was first promoted with the Pantone Hexachrome and Heptatone colours which utilise an additional bright red, purple and green.15 Many inkjet printers for textiles can now accommodate up to seven primary shades although the printer driver software is correspondingly more complex. Irrespective of the primary shades used, but particularly when only CMY inks are involved, it is not uncommon for it to be impossible to match certain shades specified in an original design. This situation is recognised by the colour management software either when input data is converted to sRGB colour space or, somewhat less likely, when it is transformed into the printer colour space. In general the least objectionable gamut remapping is that which preserves the original hue while sacrificing lightness or saturation.16 If this occurs the user is asked for a `rendering intent' which can be specified in one of four ways, namely:

174

Digital printing of textiles

10.4 The overall range of shades that can be achieved when printing with typical reactive dye based CMY inks.

· Perceptual (Picture). The method generally recommended for photographic reproduction because it applies the same gamut compression to all images, thus maintaining the same overall relative colour rendering balance. · Relative colorimetric (Proof). Reproduces out-of-gamut colours to the nearest reproducible hue. Preserves lightness but not saturation. · Absolute colorimetric (Match). Converts out-of-gamut colours to the nearest hue but sacrifices saturation and lightness. · Saturation (Graphic). Maps the saturated primary colours in the source to those in the destination irrespective of differences in hue, saturation or lightness. As the name suggests, this intent is used mainly for business display graphics. The technology of gamut remapping using lightness- and chroma-preserving scaling functions has been studied extensively and several algorithms perform well when printing on paper.17 The author is, however, not aware of any similar comparisons having been carried out specifically on a textile substrate. Several CMSs offer the possibility of visualising both 2-D and 3-D displays of the colour gamuts that can be achieved with from three to eight primary inks.

Digital colour management

175

10.5 Colour communication In modern commercial textile environments there is an increasing need for rapid and reliable communication systems both internally (possibly between many geographical sites) and with customers, specifiers and suppliers. This is particularly the case with the trend for suppliers and major retail groups to source globally rather than locally. To facilitate this there needs to be a rapid and reliable data transmission system which should preferably be `open' (nonproprietary). Examples of colour data that may need to be transmitted are: · Large, high definition image files, particularly if in 48-bit, uncompressed (e.g. TIF, BMP) colour data format · Colorimetric measurement data which may, for example be transmitted in XML (eXtensible Markup Language) code exemplified by GretagMacbeth's CxF language for colour data transmission18 · Colour management data such as equipment characterisation and ICC profile information. Particular problems can arise when transmitting and viewing colour images over the Internet,19 where it is desirable that the sender's data can be displayed accurately by the receiver (e.g. for Internet sales of coloured garments), for without suitable monitor characterisation even the 216 so-called `web-safe' colours may not always correspond. Originally GIF-encoded, 8-bit graphics were used on many websites to assist in rapid download rates, but this has become less common with the increasing adoption of high speed broadband communication and 24-bit `Trucolor' display. However, 16-bit `Hicolor' is sometimes used for which none of the individual colours (other than black and white) correspond exactly with those of 8- or 24-bit colour. Depending on the operating and browser systems (e.g. Internet Explorer, Netscape Navigator, Windows, MacOS, Linux, etc.) in use, the software may either choose the nearest colour supported or process the design image by dithering adjacent pixels, using neighbouring colours from the available palette. This type of dithering of whole pixels (as opposed to the dithering of ink spots within a pixel) can produce very objectionable effects, particularly in a textile design with large areas of solid shade. A number of firms offer software which enables web colour management and specification (e.g. GretagMacbeth's NetProfiler and eWarna's Online Colour eXchange).

10.6 Colour reproduction performance of equipment operated with a CMM A few workers have published results of investigations to compare the repeatability and accuracy with which colours can be reproduced by digital printers,

176

Digital printing of textiles

Table 10.2 Shade reproducibility from scanner to digital print Substrate Number of test colours

Mean colour difference (
Ecmc )

Reference

Not characterised Characterised Paper Paper

24 24

16.5 9.6

6.0 2.6

Randall20 Dawson7,21

Textiles

24

N/A

6.4

Dawson7

Paper

240

29.4

2.6

Sharma and Fleming22

with and without CMS control, although most of these studies concerned prints on paper rather than on textiles. Accuracy of shade matching can be assessed using a suitable spectrophotometer and calculating the total colour difference (
E) according to the CMC(2:1) colour difference equation. In general a
E of 1 or less represents a very good degree of colour matching for a dyed fabric and possibly a value of up to 3 might be completely acceptable in a textile print.7 Some data published for the reproduction of the 24 colours on the GretagMacbeth ColorChecker by inkjet printing on paper without any device characterisation shows that reproduction is poor, but although this is definitely better after device characterisation there still seems to be room for improvement. In a more extensive series of tests using the 240-shade IT8.7/2 test card and six different CMSs, the possibility of attaining much better reproducibility, at least in paper printing, has been demonstrated (Table 10.2). Day-to-day repeatablity has been evaluated for a variety of wide digital printers on textile fabrics and was found to be very good with
Ecmc ˆ 1:0  0:6, slightly better than was found for some office paper printers with
Ecmc ˆ 1:7  0:8.7 Practical and economic considerations require that print performance should also be reliable to ensure minimum downtime. The commonest faults relate to partial or complete malfunctions of individual jets and, if two neighbouring jets are affected, this appears as a stripiness in the print, particularly when printing heavy shades at lower dpi settings. Such effects may arise simply because of the tiny drops being deflected from their normal direct path to the substrate surface. Misalignment defects can be lessened if the distance between the jets and the substrate in minimised, although this increases the likelihood of contamination from loose textile fibres. Fibres adhering to the printhead surface or partial jet blockages because of inadequate filtering of the ink supply can usually be cured by head cleaning or flushing the system. All printers have software which allows a flushing cycle and indeed this is interposed at intervals during printing (another factor which reduces output). In more difficult cases the ink supply can be changed to a special cleaning fluid.

Digital colour management

177

Thermal printers can suffer from jet blockages due to a charring effect, known as kogation, occurring in the region where the tiny, bubble-forming heaters are located, so that a build-up of insoluble matter occurs which affects jetting performance. Careful formulation of the inks minimises the effect. Extreme thermal stress can ultimately cause the electrical interconnects to the heater elements to fail and in this case the ink cartridge must be replaced. To identify jet faults, test patterns can be printed at intervals between production runs, and this can be carried out automatically on some machines such as the latest Encad and ColorSpan units.

10.7 Future trends in colour management Cosiderable progress continues to be made by manufacturers and software suppliers in providing integrated solutions which are simple to apply and which utilise small, relatively low cost, calibration/characterisation equipment. In particular, with the trend to the adoption of more extensive colour targets, the advent of simple yet speedy methods of acquiring large amounts of calibration data with minimum effort is commendable. The continuing development of `entry level' systems will enable even the smallest enterprise to achieve accurately characterised colour displays and reproducible print results without the need for highly skilled technical operatives.

10.8 Sources of further information and advice Standard RGB colour spaces http://www.w3.org/Graphics/Color/sRGB http://www.srgb.com/c55.pdf http://www. www.cl-c.com/Color%20Management.pdf

IT8.7/3 Color Target http://www.clemson.edu/printcon/downloads/series.pdf

Colour space coordinate calculations/conversions http://www.brucelindbloom.com http://www.dgcolour.co.uk

Colour management http://www.boscarol.com/pages/cms_eng http://normankoren.com/colormanagement.html

178

Digital printing of textiles

http://www.measureitrite.com http://www.tasi.ac.uk/advice/creating/colour2.html http://www.ipa.org/tech/color_management http://www.ewarna.com

Books R.M. Adams and J. Weissberg, The GATF practical guide to color management, 2nd edn, Pittsburgh, PA: GATF Press, 1998. M.D. Fairchild, Color appearance models, 2nd edn, Chichester: John Wiley, 2004. E. Gioganni and T. Madden, Digital colour management: Encoding solutions, Englewood Cliffs, NJ: Prentice Hall, 1998. P. Green, Understanding digital color, 2nd edn, Pittsburgh, PA: GATF Press, 1995. P. Green and L. MacDonald (eds), Colour engineering: Achieving device independent colour, Chichester: John Wiley, 2002. A. Sharma, Understanding color management, New York: Thomson Delmar Learning, 2003.

10.9 References 1. R. McDonald, Colour physics for industry, 2nd edn, Bradford: SDC, 1997. 2. R.S. Berns, Billmeyer and Saltzman's principles of color technology, 3rd edn, Bognor Regis: John Wiley, 2000. 3. A. Johnson, `Methods for characterising colour printers', Displays, 16 (1996), 193. 4. M. Stokes, M. Anderson, S. Chandrasekar and R. Motta, A standard default color space for the Internet ± sRGB, Ver. 1.10, Hewlett-Packard, 1996. 5. Spec. ICC.1:2001-12, International Color Consortium (www.color.org). 6. `A colour appearance model for colour management systems, CIECAM02', TC8-01, CIE159:2004. 7. T.L. Dawson, `Spots before the eyes: Can ink jet printers match expectations?', Color Technol., 117 (2001), 185. 8. R.S. Berns, `Methods for characterising CRT displays', Displays, 16 (1996), 17. 9. G. Marcu and K. Chen, `Gray tracking correction for TFT-LCDs', in IS&T/SID 10th Color Imaging Conf., Scottsdale, AZ, 2002, 272. 10. G. Sharma, `LCD displays vs. CRTs ± Color calibration and gamut considerations', Proc. IEEE, 90 (2002), 605. 11. T.L. Dawson, `Light detecting devices: their use for colour measurement and image capture', Rev. Prog. Col., 34 (2004), 72. 12. A. Johnson, `Methods for standardising colour scanners and digital cameras', Displays, 16 (1996), 183. 13. G. Sharma, `Target-less scanner color calibration', J. Imag. Sci. Technol., 44 (2000), 301. 14. P. Emmel and R.D. Hersch, `Colour calibration for colour reproduction', ISCAS 2000 ± IEEE Int. Symp. on Circuits and Systems, Geneva, V-105. 15. V. Ostromoukhov, `Chromaticity gamut enhancement by heptatone multi-color printing', in IS&T/SPIE Proc. Conf. on Device-independent Color Imaging and Imaging Systems Integration, San Jose, CA, SPIE 1905 (1993), 139.

Digital colour management

179

16. E.D. Montag and M.D. Fairchild, `Psychophysical evaluation of gamut mapping techniques using simple rendered images and artificial gamut boundaries', IEEE Trans. Image Processing, 6 (1997), 977. 17. G.J. Braun and M.D. Fairchild, `Image lightness rescaling using sigmoidal contrast enhancement functions', J. Electronic Imaging, 8 (1999), 380. 18. T. Senn, T. Braun, S. Greter and F. Lamy, US 2001004801 (GretagMacbeth, USA), 2001. 19. D. Saunders, J. Cupitt, R. Pillay and K. Martinez, `Maintaining colour accuracy in images transferred across the Internet', in L. MacDonald and R. Luo (eds), Colour imaging: Vision and technology, Chichester: John Wiley, 1999, 215. 20. D.L. Randall, `Digital images for textiles ± next generation', www.techexchange.com/ thelibrary/digitalimagingNG.html. 21. T.L. Dawson, `Ink-jet printing under the microscope', JSDC, 116 (2000), 52. 22. A. Sharma and P.D. Fleming, `Evaluating the quality of commercial ICC color management software', Proc. TAGA Tech. Conf., Asheville, NC, 2000, 336.

11 ICC Color management for digital inkjet textile printing E L O S E R and H - P T O B L E R , ErgoSoft AG, Switzerland

11.1 Introduction What is pink? To measure and to describe colors is a science of its own. Unlike for time, length or mass, there is no unit for color. In addition, the physical description of a color event needs an infinite number of values: it is the spectrum of the colored light, and since a spectrum is continuous it cannot be described by a discrete number of values. To describe the visual impression of a color event, only three values are necessary: color space is said to be three-dimensional. The international commission on illumination (CIE) has developed some well-established unit systems that can be used to describe a color,1 for example CIE XYZ, CIE Lab, CIE Lch or CIE Luv. The advantage of these color spaces is that they can describe any visible color, and that they are well defined. They allow one to define a color in absolute values and they do not depend on the device that has been used either to measure or to reproduce the color. They are called `device independent color spaces'. Color spaces like RGB or CMYK are different. They depend on the device that has been used for measuring or reproducing the color: for example, an RGB value could be the measurement result of a scanner or the input of a CRT device, while a CMYK value is usually a recipe that is printed with, for example, an ink jet printer.2 The job of color management is to translate between the different color spaces: for example, it takes an RGB value from a scanner and calculates an RGB value that will display exactly the same color on a CRT, or it calculates the CMYK value that can be printed on an ink jet device. While typical RGB devices such as scanners, digital cameras, CRTs or LCDs can be characterized by quite simple formulas, ink jet printing devices are very difficult to characterize. The only way to describe the behavior of an ink jet printer very accurately

1. From now on the expression `color' shall be used to mean `visual impression of a color event'. 2. Also RGB color spaces can be device independent: e.g. artificial RGB color spaces like ECIRGB, sRGB or AdobeRGB are not related to a special device.

ICC Color management for digital inkjet textile printing

181

is to use big tables, where for many different input colors the appropriate output colors are listed. These formulas or tables are called profiles.

11.2 Overview of textile colors and common color spaces It is true that a mixture of the three ideal colors cyan_ideal, magenta_ideal, and yellow_ideal, will not cover the whole gamut of visible colors, but it will cover the maximum gamut that can be reached by a mixture of only three colors. The problem is that even today's chemistry is not able to deliver colors that exhibit spectra like the ideal colors: all available colors are far from being ideal. This becomes even more difficult if you want to print not only with ordinary dye inks on glossy paper, but on fabrics with some reactive, disperse, acid, sublimation or other textile inks: the gamut becomes quite small (Fig. 11.1). In addition to this, the colors that are normally used in textile design are quite hard to achieve with a mixture of just three or four3 colors: for example, navy blue is a very common color for textile design and normally you cannot be reached by a combination of normal cyan, magenta, yellow and black (CMYK). If one asks about colors that are hard to get by standard CMYK, the most frequent answer would be saturated red, green or blue. These are only the most impressive colors, but also a dark brown or ocher can be very difficult to print. The only way out is to add several special colors to your textile ink jet printer: for example, Hexachrome uses orange and green as additional colors, or you could add red, green and blue to CMYK. There are many different settings and it's hard to decide which is the best for a special purpose. We will give some advice later in this chapter. There is another motivation to use even more inks in an ink jet printer: usually even the smallest droplets of black, cyan or magenta will be seen on a substrate as single dots when printing light color shades. To prevent this, these color shades are printed only with the light versions of the inks: light cyan, light magenta and light black. While calibrating an ink jet printer for paper is generally fairly simple, doing so for textiles is a bit more complicated. Having completed a printout on paper, you just have to wait for the inks to dry before you can start to measure the results.4 Textile printing, however, usually needs some post-treatment, such

3. Since just three colors (CMY) don't allow the printing of dark color shades, almost every common printing process uses at least four colors where black (K for blacK) is added: it is called the CMYK process. Another advantage of CMYK to CMY is that you can save ink: this is not only a financial aspect (K is cheaper than even just a single C, M or Y, and cheaper still than the sum of all three, C+M+Y), but it's also a necessity because some printing substrates don't allow printing with up to 300% of a single ink (100% cyan + 100% magenta + 100% yellow). 4. To get more precise results, you should wait not only until the printout has dried completely, but also some additional hours because during this time the colors (sometimes) change slightly.

182

Digital printing of textiles

11.1 Spectra of different color sets for CMYK printing: inks for textile or paper, and ideal CMYK inks. Black and yellow inks, for both textile and paper, show quite ideal behavior. Especially for magenta and cyan, the textile inks are far away from the ideal spectra. Epson Ultrachrome Inks on paper, and DyStar Jettex R reactive inks on cotton, have been used as samples.

as transferring ink from paper to textile by a sublimation process, or steaming, washing and drying for a reactive ink process. It is not until this final step that the colors are valid and might be used for calibration purposes. Therefore, efficient software for textile printing should not need too many steps with many printouts, which have to be done one after another to do a complete printer calibration. Otherwise you will have to repeat the whole workflow (printing, post-treatment and measuring) too often and this will take too much time.

11.3 ICC basics Given an environment of 10 different input devices (scanners, etc.) and 10 different output devices (printers, monitors, etc.), either you would need 100 different translation tables (10  10) to translate a color value from any input to a color value of any output device (`closed loop color management'), or you

ICC Color management for digital inkjet textile printing

183

would use ICC color management instead (Ostromoukov et al., 1994). ICC color management would need only 10 input profiles and 10 output profiles (20 profiles altogether). The idea behind it is to divide the translation from input to output color space into two steps: 1. 2.

Translate input color value to a device-independent color value. Translate device-independent color value to output color value.

The ICC standard uses either CIE-Lab or CIE-XYZ values as deviceindependent color values. This color space is called the `profile connection space' (PCS). ICC color management can be divided into two tasks: 1.

2.

Preparation: build a profile for any input device and any output device. This is done by using a calibrated spectrophotometer (e.g. devices of Gretag, Xrite, Avantes, etc.) and `profiling' software (e.g. ErgoSoft ColorGPS, Monaco Profile Creator, Gretag Profile Maker, etc.). Application: select the appropriate input profile and output profile for every job (printing, or viewing on a monitor). The calculation, i.e. combination of the two profiles and calculation of the correct output color value for every input color value, is done by the so-called color management engine, CME (sometimes also called color management module, CMM, or color management system, CMS), which can be either part of the RIP software or part of the operating system.

The second task is less difficult and much better defined. The CME takes an input and an output profile and calculates a transfer function that delivers for any incoming color value the output color value. Even though the transfer function that has to be created by the CME is not defined absolutely by the ICC standard and in consequence some differences between several CMEs may exist, these differences are quite small and they can be reduced further by using ICC profiles with higher accuracy, i.e. larger profiles. For proofing or simulation, applications such as the simulation of a second output device can be included to the transfer function. Until now, we have only talked about exact color matching. Actually, this is not the real objective in many cases: one has to be aware of the possibility of different gamuts of input and output color space, and that, for example, a color taken with a camera could not be printed with the ink jet, or the pure white colors are different in input and output. Depending on the subject that has to be printed, the real objective for color translation might be different. Therefore the ICC standard offers several `rendering intents' the user can choose from (see Table 11.1).

184

Digital printing of textiles

Table 11.1 ICC standard, rendering intents Rendering intent

Description

Typical application

Absolute colorimetric

100% exact color match (colors out of gamut can lose contrast, color gradients might end in posterization)

Exact color reproduction, preview, proofing (including simulation of media white)

Relative colorimetric

Exact color match, but adaptation of media white (colors out of gamut can lose contrast, color gradients might end in posterization)

Simulation, color reproduction, preview, proofing (excluding simulation of media white)

Perceptual

Input colors are compressed; also colors out of gamut offer contrast, smooth color gradients, no posterization

Reproduction of, for example, photographs on a printer with smaller gamut

Saturation

The contrast of the input colors is Reproduction of, for increased: very pure and saturated example, business graphics colors can be reached

11.4 ICC advantages and disadvantages 11.4.1 Advantages The major benefit of color management is the adaptation of the colors of any input device and any output device. You are able to print out exactly the same colors you have previously measured with a scanner. The advantage of ICC color management is to do this not only at a single place with just one pair of devices, but to be able to print out exactly the same colors that somebody else has captured in another part of the world: color values in graphics become standardized. You just have to couple your graphic with the corresponding profile and everybody, everywhere, with a profiled printer can print it out in the correct way. ICC color management allows you to define colors in your graphics with standardized values, in the same way as you can define, for example, lengths with standardized values such as centimeters. The ICC standard supports profiling of many different devices such as any kind of printer or printing machine, monitors, scanners, digital cameras, image setters, etc. The other advantage of ICC color management is the possibility of simulating the behavior of one output device on another output device. For example, you could use your ink jet printer to exactly simulate the output of an offset press, or the output of a textile screen printing machine. This simulation is called proofing. A less expensive method of proofing is soft proofing, where a calibrated monitor is used instead of a printer as output device. This can also be done using

ICC Color management for digital inkjet textile printing

185

ICC color management. However, it is obvious that for proofing applications the gamut of the proofing device must include the gamut of the simulated device. If the media of proofing and simulation devices are different, the media color of the simulation device can be simulated (by using absolute colorimetric rendering intent) or ignored (by using relative colorimetric rendering intent).

11.4.2 Disadvantages To convert a color value is not as simple as doing conversions like Celsius to Fahrenheit or meters to inches. Color conversions are usually non-linear and can only be described by multiple values that are organized in tables. Since color spaces are at least three-dimensional, the tables usually must have at least three dimensions. A reasonably accurate description of the color non-linearity would need around 30 different values per dimension, i.e. 30  30  30 ˆ 27,000 values. For different rendering intents, different multidimensional tables are necessary, and since a profile must be able to do the conversion in both directions, one has to apply another factor of 2. This is the reason why ICC profiles easily become quite big. This gets even more difficult for CMYK colors, where four dimensions are necessary, and it becomes a forbidding task for more than four colors. This is an important aspect because it can become necessary to add an appropriate ICC profile to every graphic or image. The ICC standard provides only one perceptual rendering intent. As previously described, the perceptual rendering intent should offer the appropriate contrast compression for a special output device. But this contrast compression depends not only on the output device, but also on the actual input profile, or rather on the actual input data, i.e. only one perceptual rendering intent is too coarse. The ICC standard does not offer dedicated tools to reduce or control metamerism effects or fluorescent colors. Especially, optical brighteners which are used in almost every printing material may cause color shifts. Usually ICC output profiles are huge multidimensional tables with discrete color recipes at each table position. For values between table positions some interpolation techniques are used, so it may happen that some recipes will not appear anywhere. In other words, profiling an output device will never offer a bigger gamut; on the contrary, in some cases it may reduce a gamut.

11.5 Requirements and problems for ICC profiling There are many wants and requirements for ICC profiles or rather the profiling software. 1. Color precision. Certainly the major task is to offer perfect color matching. Color management must be able to deliver exactly the same output color for a desired input color. While this is the objective for all colors when

186

2. 3. 4.

5.

6.

7.

8. 9.

Digital printing of textiles colorimetric rendering intent is used, it is also the objective for certain colors when perceptual rendering intent is used. In the latter case, especially neutral gray colors or human skin colors have to be matched very precisely because the human eye is very sensitive to color shifts in these shades. Output profiles should take advantage of the whole gamut of the output device. Color gradients must be reproduced without any steps, especially for the perceptual rendering intent. The profiling software must offer opportunities to limit the amount of ink that is used in an output profile: for example, since one is highly unlikely to find a medium that is capable of taking up to 400% ink (100% cyan + 100% magenta + 100% yellow + 100% black) the ink usage must be reduced. Usually a paper can handle up to 300%, more or less. Different ways to control the black should be offered (black generation models): sometimes it might be preferable to print light shades of gray using cyan, magenta and yellow instead of pure black, because single black dots can disturb more than dots of cyan or magenta do, but the contrary may be preferable in other cases, such as when a light black ink is present or the black dots are to small to disturb. This would save ink, reduce color shift in neutral gray tones and reduce metamerism effects: it is called GCR (gray component replacement). Output profiles should contain not only tables to do color conversion in the direction of the output color space but also tables in the opposite direction and preview tables. This provides the possibility to do a backward translation of the output data or a simulation of the output device. In addition the profile should contain a gamut table that describes the gamut of the output device ± which colors can be reproduced and which cannot. Profiling software, especially when used for digital textile printing, must be able to handle more than just CMYK. It must be able to deal with additional red, green, blue, navy, orange, golden yellow or any other special color to generate profiles with bigger gamut. Ink should not be wasted: if a color can be reproduced by two different color recipes, the recipe that uses less ink should be used. Since color reproduction using subtractive color mixing (e.g. ink jet technology or other CMYK devices) is a very non-linear process compared to additive color mixing (e.g. RGB monitors), prediction of the colors is very difficult. The only way to do this is to print out big calibration charts with many different color recipes and measure the result with a spectrophotometer. The more intelligence is spent on the profiling software, the better the color prediction can be made and the number of calibration recipes reduced. This has serious impact on productivity, and one profiling software can get the same output quality with just 300 measurements where a competitive product needs more than 2000 measurements! Also the

ICC Color management for digital inkjet textile printing

187

method of interpolation between single measurements highly influences the precision of the color calculations. Therefore, smart profiling software does not use just simple linear interpolation. 10. In output profiles color recipes must be defined not just for colors inside the gamut, but also for any color outside the gamut. To do this, the color outside the gamut must be mapped to a color inside the gamut. Usually this will be a color with the same hue, the same brightness but lower saturation. Good profiling softwares provide more sophisticated `gamut mapping', and allow some configuration of the mapping algorithm. Actually, it is not a simple task to enforce a (subjectively) constant hue, because the value for hue in the Lab or Lch color space is only an approximation of the perceived hue. If the gamut mapping algorithm is looking just at the Lab/Lch hue, blue colors may become somewhat purple and red colors may become orange. 11. Since profiling is not a simple task, many users are not willing or able to understand all the details, so that profiling software should offer different automatic modes to make it easy enough to use.

11.6 Current technologies Many companies provide software for creating ICC profiles. The most famous in the market are Logo, that unified later with GretagMacbeth (Gretag Profile Maker), Heidelberg (PrintOpen), Color Solutions (BasICColor), Monaco (now Xrite) (Profile Creator), PraxiSoft (WiziWyg), and ErgoSoft (ColorProf and ColorGPS). Only ErgoSoft provides not just a single profiling software but also a complete software RIP (TexPrint). TexPrint is a powerful software that allows efficient printing on large-format ink jet printers, with many extensions specialized for digital textile printing. It offers a very user-friendly interface to intuitively create print jobs, direct control of many large-format ink jet printers, an integrated halftoning engine, sophisticated linearization, complete color management, PostScript interpreter and much more. The integrated profiling software (ColorGPS) is cutting edge and delivers superior profile quality at minimal effort for calibration. It is the unique software that is capable of generating profiles for up to 12 different output colors that are in demand, particularly for textile printing. The combination of RIP and profiling software allows extensive data exchange between the two parts of the software and therefore provides serious efficiency benefits for the user. We focus on this solution for digital textile printing as a reference and limit our further discussion to this software package.

11.6.1 Print environments Inkjet printing is a very sensitive process that depends on many factors, for example printer type and model as well as the actual device, type of paper or

188

Digital printing of textiles

other media, ink setting, current resolution, halftoning method, printing speed, output quality, pre- and post-treatment of the media, output method and many more. If one of these factors is changed the result will or can be different. To control all these factors, print environments are used. A print environment gives detailed information about the whole configuration of your printing system. If different configurations are used, different print environments must be created, for example if you are using two kinds of paper whose only difference is their weight (grammage). To get the best results, each of these print environments must be calibrated. Sometimes this might be too time-consuming and you may share the same calibration among several print environments. To do so may be adequate if the only difference would be a small discrepancy of the grammage of the material but it would be a sin if you change from a plain paper to a glossy paper. Digital printing on textiles compared to digital printing on paper usually takes much more time. Printing on paper is quite simple but printing on textiles frequently needs special treatments before or following the printing process, such as sublimation transfer or steaming and washing the fabric. While simplicity of the calibration process for paper printing is nice to have, in the case of textile printing it's a must: the most efficient way would be to combine all the calibrations in one single calibration task. Unfortunately this is not feasible, because the quality of the whole calibration would be too poor. A two-step process will be a good compromise between calibration quality on the one hand and usability on the other. In a first step the printing process is linearized and in the following step the profile is created. TexPrint, which offers powerful support for many different print environments, uses this two-step calibration process.

11.6.2 Linearization (calibration of single colors) Usually, halftoning techniques for ink jet printers will give quite non-linear results. This gets more difficult in the case of the common frequency-modulated halftoning methods such as error diffusion or stochastic screening. The new ink jet print heads supporting of variable dot sizes as well as light ink usage will further increase non-linear behavior. Therefore serious linearization is essential. TexPrint offers an extensive calibration tool in the form of a wizard. The user is asked about the quality he requires, that is, the effort he is willing to put in, then a calibration chart is printed. Afterwards, this chart has to be measured with an integrated measurement tool, which supports all common color measurement devices. The results are displayed both graphically and in a text list (see Fig. 11.2) where they can be controlled and, if need be, changed. In addition, the ink usage can be limited to a certain level. This should be done if further ink usage would not result in a further notable increase in optical density, i.e. the density measurement curve runs against a saturation level. It should be used only to prevent wasting of ink for a single channel. If problems with the sum of ink of

ICC Color management for digital inkjet textile printing

189

11.2 Measurement results in linearization wizard.

all channels are present, these problems should be dealt with while creating the ICC profiles, when the mixtures of the different inks are calculated. To describe the result of a density measurement, two different measuring units are common: 1.

2.

Area coverage: the ratio of an area that is partially filled with 100% of color; e.g. an area coverage of 50% means the same density that you would get by printing a pattern where exactly 50% of the area is covered with full color. Optical density: the negative decade logarithm of the re-emission ratio.

The two measuring units are linked by the formula of Murray Davies (Yule, 1967): 1 ÿ 10ÿD 1 ÿ 10ÿDfull where a is area coverage, D is optical density, and Dfull is optical density of full color. Although optical density is the unit that is normally provided by the aˆ

190

Digital printing of textiles

measurement devices, area coverage is widespread among experts. This has its origin in the (offset) printing industry. Usually when we talk about linearization of the printing device, our goal is not to get perfect linear behavior, i.e. a color value of 50% should not result in an area coverage of 50%! In former times, especially in the case of newspaper or offset printing, even a well-calibrated printing machine was not able to offer perfect linear behavior, but it was able to provide well-defined dot gain behavior. The effect of dot gain is a curve above the perfect linear curve with a smooth shape. The value of dot gain is the difference between this curve and the linear curve: for example, a dot gain of 20% would mean that an input of 50% would lead to an output of 50% ‡ 20% ˆ 70% area coverage. Since the kind of paper used for printing greatly affects the dot gain of the printing process, there is no common standard value for the dot gain parameter. A value of 20% can be recommended as somewhere between newspaper printing and glossy printing. It also results in a constant visual contrast over the whole range: a step from 10% to 20% in color value would lead to similar contrast as a step from 80% to 90%. Nevertheless the calibration wizard of TexPrint allows extensive definition of the linearization target (see Fig. 11.3). This is only useful and necessary if you work without ICC profiles, and the color space of your design application would be the color space of your ink jet device. We do not recommend this and therefore it will not be discussed in detail. But there is another feature provided by the linearization target definition in TexPrint that is very useful: you can export and import linearization targets. Using this feature you align a whole set of printers just by exporting the target of the weakest printer and importing this to any other. By doing so, all printers will behave exactly the same.

11.6.3 ICC profile generation (calibration of mixed colors) In the previous paragraph the calibration of each single ink channel was described. Now we'll describe how mixtures of inks can be calibrated. This is done by ColorGPS, the ICC profiling software of the TexPrint suite of ErgoSoft. Although ColorGPS is very powerful software with lots of adjustments, it is simple to use and will find by itself highly useful presets for virtually all the parameters. Only the quality the user wants to get, or rather the effort for calibration he is willing to spend on the one side, and the amount of ink that the printing media can handle on the other side, has to be defined manually. With this information a calibration chart is calculated. The calibration chart is optimized to reduce the number of calibration patches to a minimum. Without this optimization the number of patches needed for the calibration will increase enormously with the number of inks: to get a good color calibration it's not sufficient to measure only recipes where the different inks are printed with either 0% or 100%. Mixing colors of an ink jet device is a very nonlinear process and therefore levels between 0% and 100% have to be tested as

ICC Color management for digital inkjet textile printing

191

11.3 Target definition in linearization wizard.

well, for example 0%, 25%, 50%, 75%, 100%, i.e. five levels. In the case of three different inks this would result in 125 (53) patches, or for CMYK 625 patches. This can be handled, but how about 400,000 patches in the case of eight colors? Either the accuracy is reduced (check out just 0%, 33%, 66%, 100%, or only 0%, 50%, 100%), or one has to find other rules, as ColorGPS is doing. The idea behind those rules of ColorGPS is that not every combination of ink will be tested, depending on the current ink set. To explain this, consider a setup of CMYK plus two additional inks. Sometimes it is useless to combine both additional inks in a recipe: if orange and green were the additional inks, a combination of them would result in a gray color that can be printed using just CMYK. But also the opposite can happen: if red and orange were used as additional inks, a mixture of orange and red could result in a color unattainable by any other ink combination, and thus a combination of these two inks is reasonable. ColorGPS is able to decide, without any user interaction, which ink combination should be used. It extracts the color information needed for each ink from the linearization measurement that has been done previously. The

192

Digital printing of textiles

reduction of calibration effort is enormous: for example, for an eight-color profile ColorGPS gets an even better result with just 300 patches where other profilers need more than 2000 patches. The calculation of the optimal calibration chart is done very quickly, usually within one or two seconds. Afterwards the calibration chart can be printed and finally measured with the integrated measurement tool that supports all common measurement devices. Having done the color calibration, the ICC profile can be calculated. Depending on the precision, namely the size of the profile, this may take a while (up to some minutes). You can save the profile, test it and use it to check whether it's acceptable or not. It may happen that for some colors too much ink is used. If so, the ink limit of ColorGPS can be reduced, and without redoing the calibration, a new profile can be calculated. Only if the change of the ink limit is serious (e.g. from 300 down to 180) should you do a new calibration, because otherwise the accuracy would become too poor. It is also possible to increase the ink limit after a color calibration has been executed, but the increase should be moderate (not exceeding 50%). ColorGPS is able to accept even extreme values for the ink limit: for example, it can manage a limit of 80% as well as a limit of 600%, so that there is no need to limit a single ink channel within the linearization as there would be when other profilers are used, where sometimes ink limits under 200% are not supported. ColorGPS provides another amazing advantage: the calculation of the final ink recipes that are used in the ICC profile is done completely automatically. The user does not have to specify rules as to how (additional) inks should be used. ColorGPS finds out whether, for example, an additional blue ink might be used for colors that are darker, or brighter, or more saturated than cyan plus magenta, or whether the additional blue ink is covered by the gamut of the other inks and therefore can be omitted. The way the different inks are utilized is defined by an algorithm that results in smooth gradients of the different inks. Changing from one ink to another for neighboring target colors is inhibited as far as possible, because this would always result in color steps whenever the profile is applied. A short explanation of how the color steps arise is given here. As already mentioned in a previous section, an ICC output profile is made up of a huge three-dimensional table and two relevant colors might be located at neighboring positions in this table. Let's assume that one ink would be `golden yellow' (g.y.) and the other would be `orange yellow' (o.y.). Let's further assume that somewhere in the table a recipe would consist of (nearly) 100% of `golden yellow' and the recipe at a neighboring location would consist of (nearly) 100% of `orange yellow'. When this profile is applied, and the recipe for a color exactly in the middle of these locations is calculated, a recipe like 50% `golden yellow' plus 50% `orange yellow' would be the result, because this is the mean of 100% g.y. + 0% o.y., and 0% g.y. + 100% o.y. The problem about this `intermediate' recipe is that it will look considerably different! A recipe consisting of two inks,

ICC Color management for digital inkjet textile printing

193

each at 50%, will not cover 100% of the medium. The reality would be that 25% of the medium would be covered with both inks, 25% with one and 25% with the other ink, and 25% would remain uncovered! This is a long way from the intermediate color. If you find this difficult to imagine, just think about using black as the first ink and also as the second ink. The expected intermediate color value would be black, but the reality is the mixture of 25% double black, which is black just as well, 50% (single) black, and 25% white, i.e. 75% black + 25% white, giving gray! Sometimes even seriously different recipes for neighboring colors cannot be avoided, since actually they are needed! This can happen when using inks that have almost the same color and each of them is expanding the gamut of the others, i.e. none of the inks can be omitted. In consequence there is a trade-off between big gamut requiring many different inks, on the one side, and high accuracy with smooth color gradients, where only few colors can be used, on the other side. Using more inks neither increases the accuracy of a profile nor improves the smoothness of the colors. The only reason to increase the number of inks may be to enlarge the gamut. The following rule should be considered: use an ICC profile with fewer inks as long as it covers the gamut you need, and choose a profile with many inks only if really needed! Calculating a profile is done by ColorGPS in two steps. First, for every color within the gamut an ink recipe that will match the color is calculated. Second, any color outside the gamut is mapped to a color within. Theory about gamut mapping can be found in Morovic (1998) and Morovic and Luo (2001). The mapping is controlled by two methods: first, changing the contrast, and second, the `real mapping'. Changing the contrast will change the `pretended gamut'. If the contrast for the colors within the gamut is reduced, the pretended gamut increases, i.e. fewer colors remain outside. If the contrast is increased, the same input colors become more saturated or darker, the edge of the gamut will be reached earlier and therefore the pretended gamut will become smaller. The contrast calculations can be described as a kind of scaling. Scaling can be done in the direction of the L-axis (for contrast in the luminance direction) and the Caxis (for the saturation direction). Afterwards, the `real gamut mapping' will replace any color outside the gamut by an appropriate color inside the gamut. This replacement will change depending on the character of the original color: hue, saturation or lightness will be affected. For the colorimetric rendering intent all three values are equally important and therefore the weighting factors for each deviation should be equal. In the case of perceptual or saturation rendering intent usually they will be different: for example, to preserve hue is more important than to preserve lightness or saturation. Although the standard setting leads to fine results, all the parameters that control the gamut mapping may be changed to fulfill special needs (for an example see Fig. 11.4). Finally a tip for daily work: you should develop a system to define names for profiles or other calibration files. Include all the necessary details, like printer

194

Digital printing of textiles

11.4 Parameters to control gamut mapping of colors.

model, resolution, quality, media, etc., in the file names. You will have to invent some abbreviations, otherwise the names become too long. Without this information at a glance, efficient work with many different profiles will not be possible.

11.7 Results Using ICC-based color management helps you to control the color output of your ink jet device. To determine the quality of the color management is not a simple task and a benchmark must contain more than just a single discipline. At least three of the following tasks should be checked: 1. 2.

Precision of color matching Smoothness of color gradients

ICC Color management for digital inkjet textile printing 3. 4.

195

Largeness of the gamut Suitability of gamut mapping.

For tasks 1 and 3 an objective measurement can be defined. For the other tasks this will hardly be possible and they have to be judged subjectively. A benchmark for the color matching may be the color difference between target and measurement of some test colors after printing. The difference can be scaled in units of
E (i.e. Euclidean distance for the two Lab values) or in some derivatives like
E94 or
ECMC (Clarke et al., 1984). The test patches should be within the gamut of the target device because treatment of out-of-gamut colors is mainly a concern of gamut mapping and should not be judged by a simple
E value. The average and the maximum color difference are very significant figures and are appreciated as a benchmark. The largeness of the gamut can be determined using the gamut tables of the profile. This can be done very easily, but since there is no well-defined rule how the gamut tables have to be calculated by the profiling software, they can be used just to compare profiles created by a single profiling software, otherwise it's like comparing apples and oranges. For subjective testing, appropriate test images are needed. They must cover the regions of interest. Since color spaces are three-dimensional and an image is always restricted to two dimensions, it can cover only a subset of the whole color space. Furthermore, a very common mistake is to use RGB images for testing where sometimes sRGB is the color space. sRGB may provide a gamut with a larger volume, but an actual ink jet device can also produce colors that are not covered by the sRGB gamut anyway! Another disadvantage of common RGB color spaces is that their most saturated colors (red, green, blue, cyan, magenta or yellow) normally are quite different from the corresponding colors of the ink jet. Usually, the most saturated colors of an ink jet are darker. A good advice is to build up a generic CMYK profile where the Lab values for the different inks can be defined manually. Use values that are typical for cyan, magenta and black, or, to make sure that a larger gamut will be provided by this profile, you may use Lab values that are (slightly) more saturated. Use such a profile as the working color space within your design application to draw and generate test images with interesting and suitable test colors and color gradients. To give you an example for an ICC profile application for digital textile printing, a comparison of six and 11 color setups is shown here. Due to limited space we will reduce the discussion to a gamut comparison of the two setups. For the experiment, DyStar Jettex R reactive inks on cotton were applied. Since the gamut of a simple four-color setup is even poorer, we used the six-color as a simple setup and the whole 11-color as an advanced setup. The six-color setup consisted of `Yellow 5G', `Orange RN', `Red 4B', `Blue 3R', `Turquoise GM' and `Black BN', and for the 11-color setup `Golden Yellow GR', `Red FB', `Navy 4R', `Green 2GM' and `Brown 2R' were used in addition. In Figs 11.5

196

Digital printing of textiles

11.5 Gamut pictures for six-color (gray outline) and eleven-color (white outline) setup. Shown are planes of the Lab color space for constant values of L (on every picture yellow is at the top and blue at the bottom, green on left and red on right-hand side).

and 11.6 different planes of the Lab color space (L = 100, L = 90, . . ., L = 0) with the outlines of the gamuts are shown. It can be seen that even inks with low color saturation (like navy or brown) will increase the gamut: at their level of lightness, they offer the highest saturation. The effect of navy can be seen best at level L = 20, and the effect of brown at level L = 40 (upper right corner of the gamut). One should bear in mind that especially these two colors play an important role in textile design, mainly for clothing.

11.8 Conclusion and future trends ICC-based color management is the single well-established standard for color management. There may be some other solutions, but usually they are proprietary. Furthermore, in most cases they do not allow separate calibration of input and output devices and will offer just closed loop calibration, where a combination of input and output is calibrated. Almost every important graphic software supports ICC color management. Since color management is not just a trivial task, neither for the user nor for the software developer, there are still many applications where colors can be used but color management is not offered

ICC Color management for digital inkjet textile printing

197

11.6 Gamut pictures for six-color (gray outline) and eleven-color (white outline) setup. Shown are planes of the Lab color space for constant values of L (on every picture yellow is at the top and blue at the bottom, green on left and red on right-hand side).

at all, such as office software. Probably in future every software that deals with color will support ICC color management. Not only quantity, but also quality of ICC color management will be subject to change: today there is only one ink jet printer that is able to support up to 12 different inks, and ColorGPS is the single profiling software that can handle up to 12 inks. Since textile printing has a strong need for many saturated colors that cannot be completely reproduced by a four-, six- or even eight-color process, it is very likely that in the future also textile ink jet setups with up to 12 colors will become more popular. In addition, the quality of interpolation for color calibration might be enhanced to increase color precision and to reduce the effort of calibration in future. ICC color management works well and now evidence can be seen that something else might replace the ICC standard some day.

11.9 Sources of further information and advice Fraser, Bruce, Real World Color Management: one of the best books about color management Sharma, Gaurav, Digital Color Imaging Handbook: profound technical information

198

Digital printing of textiles

http://www.color.org/ International Color Consortium, e.g. standardization of ICC profiles http://www.eci.org/eci/en/ European Color Initiative, e.g. application of ICC standard http://www.boscarol.com/pages/cms_eng/ Introduction to color management http://www.coloraid.de/ Tool for color management (e.g. public domain color management system for MS-Windows and Linux) http://www.brucelindbloom.com/ Information about different color spaces, many color calculators, test images. Highly recommended! http://www.cie.co.at/cie/ Homepage of the international commission on illumination (Commission Internationale de l'Eclairage): responsible for standardization of color measurement http://www.colorsystem.com/ Virtual color museum http://www.xrite.com/ X rite provides hardware and software for color calibration http://www.gretagmacbeth.com/ GMB, Gretagmacbeth, provides hardware and software for color calibration http://www.ergosoft.ch/ ErgoSoft AG develops profiling and RIP software

11.10 References Clarke, F. J. J., MacDonald, R. and Rigg, B. (1984), Modification of the JPC79 colourdifference formula, Journal of the Society of Dyers and Colourists, 100: 117. Morovic, J. (1998), To develop a universal gamut mapping algorithm, Ph.D. Thesis, University of Derby, UK, 1998. Morovic, J. and Luo, M. R. (2001), The fundamentals of gamut mapping: a survey, Journal of Imaging Science and Technology, 45(3): 283±290. Ostromoukhov, V., Hersch, R. D., Peraire, C., Emmel, P. and Amidror, I. (1994), Two approaches in scanner±printer calibration: colorimetric space-based vs. `closedloop', IS&T/SPIE International Symposium on Electronic Imaging: Science & Technology, SPIE, Vol. 2170, San Jose, California, USA, 133±142. Yule, J. A. C. (1967), Principles of Color Reproduction, John Wiley & Sons, New York, 212.

Part III Digital printing coloration

12 Substrate preparation for ink-jet printing C H A W K Y A R D , University of Manchester, UK

12.1 Introduction Pre-treatment of textiles in preparation for ink-jet printing is carried out because inclusion of auxiliary chemicals and thickeners into the low viscosity ink has proved troublesome. Thus the methodology is akin to `two-phase' conventional printing as opposed to the `all-in' approach. In the latter case all the dyes, chemicals and thickeners required are included in the print paste, whereas in the former some of the ingredients, particularly chemicals, are applied before, or after, printing. Although ink-jet printing of textiles has been established since about 1987, it has yet to account for a significant proportion of the total volume of printed textile production. The main reason for this is the low productivity of the printing equipment that has been available, but the expense and complexity of having to pre-treat the fabric prior to printing is also a stumbling block. Having applied thickening agents and chemicals to the fabric to aid fixation of dyes and prevent undue wicking and penetration, these, along with unfixed dye, have to be removed after the fixation step. This is in contrast to transfer or sublimation printing, where the thickener and chemicals remain on the paper after the volatile disperse dyes have diffused into the textile during printing. Consequently no further fixation or wash-off is required. Jet printers for printing directly onto fabric can also be used to print transfer paper, as the machines were originally designed for the graphics industry. Transfer of the design onto rolls of fabric may then be carried out in a continuous manner using a cylinder and blanket-type machine, known as a transfer calender. The graphics printers, which have been adapted for ink-jet printing of textiles, are discontinuous. This point will be elaborated on later. While transfer printing is an attractive alternative to ink-jet printing for polyester substrates it is not viable for natural fabrics such as cotton, wool or silk, as disperse dyes are not substantive to them. Historically there have been attempts to transfer print these fibres, particularly by Dawson International and Tootals. Substantive dyes, such as reactive or acid dyes, were printed onto paper,

202

Digital printing of textiles

but thickener and chemicals were applied to the substrate. Some moisture was also required to ensure the dyes dissolved. The Fastran Process1 was developed by Dawson International. It was a batchwise process used mainly for printing pre-treated woollen garments with acid dyes. Tootals built the DewPrint machine to carry out the damp transfer process on a continuous basis.2 The machine never gained any commercial acceptance, however, as printing a damp substrate was difficult. Ink-jet printing of textiles has generally been carried out using pre-treated materials, with the exception of carpets. The Millitron carpet printer3 enables nylon or wool carpets to be printed on a continuous basis using low viscosity acid dye inks. The carpet is often pre-treated with an aqueous solution of surfactant which is hydro-extracted with a suction slot prior to printing. This process has been successfully used for over 30 years, and is still the best example of continuous jet printing on a commercial scale. On a smaller scale, the Zimmer Chromojet machine, used mainly for printing carpet tiles, and the more recent Chromotex fabric printer have also met with some success. Zimmer now have a continuous `beam' Chromotex printer for carpets. Once more, the ink contains dyes, thickener and chemicals in the `all-in' style. When printing cotton the choice has generally been between reactive dyes and pigments. The pigment printing process is simpler, as it involves three main stages (print, dry, bake/cure), whereas reactive printing has two extra processes (print, dry, steam, wash-off, dry). Pigment printing is therefore a more economical procedure, and for that reason is favoured in most countries. Polyester/cotton presents more difficult technical problems than 100% cotton when printing with dyes, and although these have been overcome,4 pigment printing is overwhelmingly the choice for this blend. However, when ink-jet printing emerged, reactive dye inks were made available in the first instance. Jet printing with pigments has proved to be technically difficult and the development of systems has taken much longer. The problems for pigment systems were exacerbated because the graphics printing machines first adapted for jet printing were of the thermal drop-on-demand type, the Encad machine being the most popular. In this type of printer the ink is heated and a bubble of vapour forces the droplet out of the nozzle. Pigment inks, especially if they contained the binder, were not compatible with this type of machine. Reactive printing by the `all-in' method is the normal approach for screen printing, but for jet printing it has certain dangers. These will be dealt with later. As a result the jet printing of cotton, wool and silk has generally been carried out by the `two-phase' method, the ink containing only purified dyes, the thickener and chemicals being applied to the substrate in a pre-treatment. Although the quality of the resulting prints is excellent, the extra expense of pre-treating the fabric by a pad/dry process makes the process uneconomical for anything but short runs. Accountants also frown on the concept of holding stocks of expensive pre-treated fabric.

Substrate preparation for ink-jet printing

203

Over the past few years one of the barriers to successful ink-jet printing of pigments has disappeared with the introduction of the piezo type of printer, since no heating of the ink takes place. It was clear from the start that it would be of great commercial benefit if a pigment-based printing system, which did not require a fabric pre-treatment, were to become available. Early attempts were unsuccessful as the pigments produced much duller shades than could be achieved with dyes, and there was a tendency for nozzles to block, in other words the `runnability' was poor. The inks tended to dry out in the nozzle, and in an attempt to avoid this happening, some producers installed air humidifiers in the vicinity of the printer. Companies such as Ciba and BASF have improved their pigment inks over a long development cycle, and these drawbacks have been corrected to a considerable extent.

12.1.1 Reasons for pre-treatment The main reasons for separating the dyes from thickeners and other chemicals and applying them separately to the fabric are as follows. · `All-in' inks are less stable and have lower storage stability, e.g. reactive dyes are more likely to hydrolyse when alkali is present in the ink. · Chemicals in the ink cause corrosion of jet nozzles; the deleterious effect of sodium chloride on steel surfaces is well known, for instance; inks for use in `charged drop' continuous printers should have low electrical conductivity. · Thickeners in the ink often do not have the desired rheological properties. · Some chemicals can be utilised in pre-treated fabric but would cause stability problems in the ink; e.g. sodium carbonate as alkali for reactive dye fixation is acceptable on the fabric but not in the ink. · The presence of large amounts of salts in aqueous inks reduces the solubility of the dyes; concentrated inks are required in jet printing due to the small droplet size. The point about the rheology of the ink requires amplification. Generally low viscosity inks are required at the printing stage, but once the ink is deposited on the substrate, higher viscosity is necessary to prevent lateral spread. This suggests the use of highly pseudoplastic thickeners.5 Unfortunately, this can cause `tailing' and satellite formation during printing.6 During printing, inks with Newtonian flow properties perform better. A further advantage of applying thickeners and chemicals separately from the dyes is that it allows the wettability and penetration properties of the fabric to be adjusted. Wettability is the lateral spread of liquid when it impinges on a fibrous matrix, and is caused by the capillary forces that are present in the narrow interstices between fibres and yarns. In the past this was assessed with the aid of a burette containing the liquid (water or sugar solution) and stopwatch.7 The fabric was held in an embroidery ring and the time taken for a drop of liquid to

204

Digital printing of textiles

spread out on the surface was measured manually. More recently this procedure has been automated.8

12.2 Ink systems 12.2.1 Reactive dyes Reactive inks are generally used for cellulosic substrates, where they produce bright shades with good fastness to washing and light. The dyes form stable, covalent bonds with the fibre under alkaline conditions. Wool, silk and nylon may also be printed with them. There are several types of reactive dye, not all of which are suitable for printing purposes. The factors involved are (i) reactivity and (ii) substantivity. 1.

2.

Reactivity. The main classes of reactive dye9 are shown in Table 12.1. Dichlorotriazines are the most reactive, and cannot be used in printing because their storage stability in inks and print pastes is too low. Monochlorotriazine (MCT) dyes are less reactive and are the most common choice for ink-jet printing inks. Substantivity. Dyes intended for printing should have low substantivity to the substrate. In this respect they are akin to dyes used for continuous dyeing, as opposed to those used in exhaustion dyeing, which should have high substantivity. The way low substantivity is achieved is to build into the dye structure bulky side groups, which sterically hinder the dye molecules from approaching dye sites on the fibre.

Reactive dyes in liquid form, as supplied for conventional screen printing, are not suitable for ink-jet printing. Commercial dyes contain quite high levels of salt (sodium sulphate or sodium chloride) and this can cause corrosion of jet nozzles. The water solubility of the dyes is also reduced. Therefore, the majority of the salt has to be removed in reactive dye ink formulations. This is generally achieved by the process of reverse osmosis (RO), in which a semi-permeable membrane is employed to separate the ionic species. Table 12.1 Major types of reactive dyes9 Reactive group Dichlorotriazine Difluorochloropyrimidine Dichloroquinoxaline Monofluorotriazine Vinyl sulphone Monochlorotriazine (MCT) Dichloro- and trichloro-pyrimidine * 1 ± most reactive, 6 ± least reactive.

Relative reactivity* 1 2 3 3 4 5 6

Substrate preparation for ink-jet printing

205

Even with reduced salt levels, dye solubility in reactive dye inks can be a problem. As a result, manufacturers have turned to the use of lithium as a partial replacement for sodium as the cation associated with the anionic, sulphonated dyes.10 Apart from the dye, the inks contain hygroscopic (or hydrotropic) agents such as diethylene glycol, propylene glycol, and diethyleneglycol monobutyl ether, to avoid drying out of the ink in the nozzles, surfactants and phosphoric acid-based buffers.11 These modifications partly explain the high cost of the inks, which, weight for weight, are more expensive than gold. Fixation of reactive prints on cellulosic substrates is by atmospheric steaming for 10 minutes at 102ëC, followed by a wash-off.

12.2.2 Acid dyes These water-soluble anionic dyes are the common choice for printing wool, silk and nylon. They do not react with the fibre to form covalent bonds, but instead are attracted to positively charged dye sites on the fibre. The shades are often deeper and brighter than are achievable with reactive dyes. Similar measures have to be taken to purify the dyes in ink formulations as are mentioned above for reactive inks. Hygroscopic agents, such as glycerine, diethylene glycol and triethylene glycol monobutyl ether, and a surfactant are also commonly included in the ink.12 Fixation after printing is by atmospheric steaming at 101±103ëC for 30±40 minutes, followed by washing and drying.

12.2.3 Disperse dyes These dyes are applied to synthetic-fibre textiles as finely dispersed particles as they have very limited water-solubility. Inks were developed later than the water-soluble dye inks because of difficulties experienced in producing stable dispersions.13 Ink manufacturers tend to produce two sets of disperse dye inks, one for direct printing onto polyester, and the other for transfer printing. The latter contain lower molecular weight, more volatile, dyes than the former. Fastness requirements may dictate the use of higher molecular weight dyes, such as in certain apparel end-uses and for automotive seat-cover materials, where light-fastness is critical. However, the transfer printing inks can be printed directly onto polyester fabric rather than paper, and there are advantages in doing this, as the dyes require less time or a lower temperature for the subsequent fixation step than the higher molecular weight ones. The high definition and tonework made possible in transfer printing is matched by the results achieved when jet printing directly onto fabric, and the extra expense of the paper and a second printing operation is avoided. Once again, the ink formulations have reduced salt content. Fixation of prints is achieved by high temperature (HT) steaming at 170±180ëC, or by dry heat at 190±200ëC, and is followed by a wash-off cycle, including a reduction clear.

206

Digital printing of textiles

12.2.4 Pigments Pigments are insoluble in water and are applied to textiles as finely divided dispersions, in a similar manner to disperse dyes. Methods of producing submicron particle size have been developed over recent years. They have no affinity for textile substrates, and are adhered to their outer surfaces by means of a polymeric, self-cross-linking binder. The fixation step after printing is by dry heat curing. Jet printing with pigments contained in the ink is fraught with difficulties, due to the risk of the particles aggregating and causing nozzle blockages. When the binder is also included in the ink, the risk of blockages can only increase. This also increases the viscosity of the ink to levels that may be too high for some printing machines.14 Under these circumstances it has been found necessary to apply the binder as a post-treatment (after printing). It is for these reasons that jet printing with pigments has taken such a long time to develop. A detailed study involving four different types of pigment dispersion (polymer dispersion, surfactant dispersion, microencapsulation and surface modification by hydrophilic functional groups) has been carried out.15 Binder was included in the inks and the optimum ratio of pigment to binder determined (1:2 for all inks). The inks formulated on surfactant dispersion showed good compatibility, printability and textile properties. Pigment systems require a dry-heat curing or fixation stage after printing. There has been much speculation, however, on the potential for ultravioletcurable systems in textile jet printing. These solvent-based inks have been in use in the graphics industry for some time. One such system, which does not require a pre-treatment for the cloth being printed, has recently been patented.16

12.2.5 Inks for paper printing Water-soluble dyes such as direct dyes and food dyes are usually used in ink-jet inks for paper printers. In the early days this meant that the prints were subject to smudging during later handling. The problem was solved by making modifications to the dye structure. Sodium sulphonate (±SO3ÿ Na‡) groups, which provide water-solubility, were initially converted to sodium carboxylate (±COOÿ Na‡) groups. These interact with the acidic size on the paper to produce ±COOH groups which cause the dye to become insoluble.17 It was later found that the ammonium salt of the acid produced even better results, since on warming ammonia is released and conversion to the insoluble carboxylic acid is virtually complete.

12.3 Fabric pre-treatments All textile substrates destined for colouration, whether by dyeing or printing, require preparation to make them clean and receptive to aqueous treatments. This is the case whether the fibres involved are natural or synthetic, although

Substrate preparation for ink-jet printing

207

natural fibre substrates require a more thorough and expensive processing route. For example, a typical sequence for woven cotton fabric is as follows: Singe > desize > scour > bleach > mercerise > dry > weft-straighten > batch The order of these processes may vary slightly and some of the processes can be combined. It is during the hot alkaline scouring stage that the hydrophobic outer layer of the cotton fibres is removed. Mercerisation with concentrated caustic soda solution improves dye yield and lustre for cotton fabrics. A less severe causticisation process is used for viscose rayon and lyocell fabrics. The batched fabric should then be checked for wettability. If this is satisfactory it is ready for dyeing or printing (if printing, the fabric is often labelled `P for P' ± prepared for print). Fabric absorbency is important in screen printing because, if the fabric is water-repellent, the first colour printed will not be absorbed before the fabric reaches the second screen, where the first colour will be smudged. However, should the fabric be very absorbent the print paste may penetrate too far through to the back, resulting in poor colour yield and subdued colours. A compromise is therefore required. The same applies to fabric intended for jet printing. Woven fabrics contain size, knitted ones have knitting oils, non-wovens contain lubricants as do carpets, and all these must be removed.

12.4 Pre-treatments for ink-jet printing 12.4.1 Cotton and cellulosics Cotton, viscose rayon and lyocell fabrics are normally jet printed with reactive dyes by the two-phase method, i.e. the fabric is pre-treated with thickener and alkali, while the ink contains the dye. The pre-treatment liquor is normally applied with the aid of a pad mangle, though it could be screen printed. In either case the fabric must be dried to about 5±7% moisture content before printing. The main constituents of the aqueous liquor are usually thickener, alkali and urea. A typical formulation is given in Table 12.2. The nature and role of each of these constituents will be discussed.

Table 12.2 Pre-treatment for reactive dyes (MCT type) on cotton 100 g/L 100 g/L 20±30 g/L

Medium viscosity sodium alginate, e.g. 6% Lamitex M5 Urea Sodium carbonate

Pad (approx. 75% pick-up) ± Dry at 120ëC or below* For viscose rayon increase urea to 200 g/L and add 10 g/L Lyoprint RG (Ciba) * The reason for the low drying temperature is that urea is unstable at higher temperatures and these cause high levels of fumes at the stenter exit.

208

Digital printing of textiles

Thickener Thickeners are employed in printing to preserve the sharpness of edges and outlines by countering the natural wicking effect of the substrate. In addition they hold moisture to enable dyes and chemicals to dissolve and enter the fibres during the steaming stage after printing and drying. They also modify the flow properties (rheology) of the ink or print paste. The thickening agent should not react with either the dye or other chemicals present because, if they do, an insoluble product usually results. This does not wash off and the fabric becomes stiff. Natural product-based thickeners are carbohydrates and contain many hydroxyl groups, which is why they absorb moisture so well. However, this also means that they react with reactive dyes in a similar fashion to their reaction with cellulose, another polysaccharide. In practice this restricts the choice of thickener to one type only, sodium alginate. This product is derived from brown seaweed and is polyanionic. It is this property that prevents the anionic reactive dyes from reacting with the thickener, since both have negative charges and so repel each other. Various grades of sodium alginate are produced, ranging from low to high molecular weight. The low molecular weight grades produce high solids content thickener solutions with fairly Newtonian flow properties, whereas the high molecular weight grades produce low solids content pastes or inks with highly pseudoplastic (shear thinning) flow properties. When the thickening agent is being applied to fabric prior to printing, as in this case, the flow properties are not as important as the solids content. A higher weight of thickener on the fabric assists the preservation of edges and outlines in the printed design during drying and steaming. When the thickener is included in the aqueous component of an oil-in-water emulsion and then dried the thickener layer takes on a microporous, sponge-like structure.18 This produces jet prints with superior definition.19 There will clearly be environmental implications, however, as the hydrocarbon phase of the emulsion will evaporate into the atmosphere. Alkali Reactive dyes react with cellulose under alkaline conditions to form covalent bonds between fibre and dye. There are various classes of reactive dyes, monochlorotriazine (MCT), vinyl sulphone etc., and these require different strengths of alkali for optimum fixation. Sodium bicarbonate is generally recommended for `all-in' pastes and inks, as it causes least hydrolysis of the dye on storage, but the stronger (and cheaper) alkali, sodium carbonate, is satisfactory for most pre-treatment purposes. Some padding formulation recommendations replace the sodium carbonate in Table 12.2 with 25 g/L sodium bicarbonate.

Substrate preparation for ink-jet printing

209

Urea Urea is a very common constituent of print pastes as it acts as both dye solvent and hygroscopic agent (or humectant). Concerns have been expressed about the detrimental environmental effect of discharging high levels of nitrogen into waste water,20 but despite this urea remains the common choice. Many padding recipes for viscose rayon increase the level of urea to 200 g/L. Other auxiliaries Another commonly incorporated constituent of the pad liquor is sodium meta nitrobenzene sulphonate (Ludigol, BASF, 25 g/L), a mild oxidising agent, which is included to avoid the risk of reduction, and hence decolourisation, of the dye during steaming. The inclusion of hydrophilic non-ionic polymers, such as polyoxyethylene diisopropyl ether,21 ethylene oxide/propylene oxide random copolymers22 or polyoxyethylene lauryl ether23 to replace the alginate thickener, has been mentioned in patents, as has the use of hydroxyalkyl imine derivatives.24 Fluorine-containing water repellents are also claimed to improve colour yields.25 Coating formulations for paper intended for ink-jet printing often contain fillers such as silica gel, and this has proved to increase colour yield and reduce bleeding when applied to cellulosic textiles.26,27 Some references in the patent literature mention the use of cationic agents, such as polyvinyl pyrrolidone derivatives, in the pre-treatment formulation,28,29 but this must be viewed with caution. Such agents may increase the colour yield of reactive prints, but there is a danger that, during the wash-off, unfixed anionic dye and hydrolysed dye will stain unprinted grounds. The requirement of a pre-treatment for cellulosic fabrics has been avoided by including in the ink a durable press finishing agent.30 Both reactive and acid dyes were examined. An alkaline catalyst was not required for reactive printing. Cationisation of cotton There has been a great deal of research into the cationisation of cotton and other cellulosics,31 and Chapter 16 of this book is dedicated to the subject. This process, which enhances the dyeability of the fibre with anionic dyes, involves the chemical reaction of cationic reactive agents with cellulose. Typically these are quaternary ammonium compounds with reactive groups such as epoxy groups attached.32 Among the factors which should be considered before such a process is carried out are: · The cost of the cationic reagent · The toxicity of the reagent · The total cost of the process, including drying

210

Digital printing of textiles Table 12.3 Pre-treatment for silk prior to reactive ink-jet printing 10 g/L 100 g/L 100 g/L 30 g/L

Lyoprint RG (Ciba) Alginate medium viscosity, e.g. 6% Lamitex M5 Urea Sodium bicarbonate

Pad (approx. 75% pick-up) ± dry at low temperature, 100±120ëC.

· Whether the higher dye fixation level normally achieved is sufficient to prevent staining of the print during the wash-off phase.

12.4.2 Wool and silk There is a choice to make when printing wool or silk, as either acid dye or reactive dye inks may be used. As explained above, acid dyes provide colours of greater depth and chroma, but reactive prints are perfectly acceptable for most purposes. When a company is printing mainly cellulosic materials with reactive dyes it would make economic sense to use them on silk too, in order to save holding stocks of two ranges of inks. It is very unusual, though, to print wool with reactive dyes. The pre-treatment formulation for reactive dyes on silk is similar to the recommendations for cotton, except that it is safer to use sodium bicarbonate than sodium carbonate (Table 12.3). The pre-treatment for acid dyes on wool and silk is quite different from that shown in Table 12.3, and there are a number of alternatives suggested by the ink manufacturers. The three main constituents are thickener, urea and an acid or latent acid (a substance that breaks down during fixation of the print to release acid): see below. A typical padding recipe is shown in Table 12.4. Thickener The preferred thickening agents for printing with acid dyes are of the mannogalactan type. These are stable to the acidic conditions which are required at the fixation stage. Sodium alginate should not be used, as insoluble alginic acid is formed in the presence of acids. The two main sources of these thickeners are guar gum and locust bean gum. They differ only in the ratio of mannose to galactose units in the polysaccharide structure.33 Table 12.4 Pre-treatment for wool and silk prior to acid dye ink-jet printing 150 g/L 100 g/L 50 g/L

Guar gum thickener, e.g. Meyprogum NP 8 (8% solution) Urea Ammonium tartrate solution (1 part water to 2 parts ammonium tartrate)

Pad (approx. 75% pick-up) ± dry at low temperature, 100ëC or below.

Substrate preparation for ink-jet printing

211

Acid or latent acid Acidic conditions during the steaming of the printed fabric are necessary to protonate amino groups in the wool or silk. The incorporation of mineral acids such as sulphuric or hydrochloric acid is not recommended, as these will damage the fabric, but organic acids such as citric acid (20 g/L) have been recommended for silk and polyamide substrates. Most formulations, however, include a latent acid. These are generally ammonium salts which decompose during steaming to release ammonia and leave the parent acid. The cheapest option is ammonium sulphate (20 g/L), but there is a small risk, as sulphuric acid may attack some dyes or damage the fabric. The ammonium salt of an organic acid is safer, hence the inclusion of ammonium tartrate in Table 12.4.

12.4.3 Nylon This is usually printed with acid dyes, using a similar pre-treatment to that shown in Table 12.4. Some experimentation with varying amounts of urea and acid may be necessary to achieve optimum results. A novel approach, in which polyamide substrates are treated with a solution of a bisphenol derivative34 for 30 minutes at 90ëC, washed, dried, and then jet printed with acid dye ink, is claimed to provide high colour yield and good colour fastness. The incorporation of cationic compounds in a pre-treatment for polyamide substrates has also been patented.35 Another patent mentions the inclusion of a reagent, Naziridinyl-N0 -stearyl urea, to reduce bleeding of the subsequent print.36 Wool, silk and nylon can also be jet printed with inks containing reactive dyes.37 In that case sodium alginate is preferred as migration inhibitor/hydrotropic agent rather than mannogalactan types for the reasons previously given.

12.4.4 Polyester Polyester is printed with disperse dye inks. Disperse dyes require temperatures above the glass transition of the polymer for them to diffuse inside the fibre. There is no great risk of reaction between the dye and the thickening agent, but the latter should be stable to the high fixation temperatures. The usual choice is sodium alginate. This acts as a migration inhibitor during drying and steaming. Synthetic thickeners have also been included as alternatives to sodium alginate. The same bleeding preventer mentioned above for nylon has also been included in a pre-treatment for polyester textiles.36 For once, urea does not appear in many pre-treatment formulation recommendations, although there are exceptions.38,39 The inclusion of urea is likely to cause fumes and pollution problems during the high temperature, dye fixation stage. Instead, the ink manufacturers have developed their own formulations, and they should be asked for their latest recommendations. Table 12.5 provides a guide.

212

Digital printing of textiles Table 12.5 Pre-treatment for polyester printed with disperse dye ink 10 g/L 100 g/L

Cibatex AR (Ciba) Sodium alginate medium viscosity, e.g. 6% Lamitex M5

Pad (70% pick-up) ± dry.

Other thickeners may be used instead of sodium alginate. One patent40 includes a mixture of polyacrylic acid, sodium salt (1%) and polyvinyl alcohol (2%) in a pre-treatment formulation; another includes -tocopherol.41 A further Japanese patent describes a different pre-treatment for polyester.42 The formulation contains carboxymethyl starch, sodium chlorate and polyoxyethylene alkyl ether. Cationic polymers have also been patented for use in pre-treatment formulations for polyester fabrics.43 In this case there will be an attraction for the anionic dispersing agent in the disperse dye ink. An interesting variation in this approach has also been patented.44 In this method a polyester fabric is pretreated with barium chloride solution and dried. The ink contains disperse dye with an anionic dispersing agent with the structure shown in Fig. 12.1. The divalent barium cation presumably replaces the monovalent sodium ion to produce a less hydrophilic, insoluble compound. It is claimed that this procedure reduced the spread of printed drops considerably.

12.4.5 Polyester/cellulose blends The Japanese company Seiren have patented a method for printing polyester cotton blend fabrics with ink containing disperse and reactive dyes.45 A pretreatment formulation including a cellulose-reactive compound, such as isonicotinic acid, a water-soluble polymer and an inactive non-water-soluble oil is mentioned.

12.4.6 Pre-treatments for pigment printing Some pigment systems for ink-jet printing require a fabric pre-treatment. The binder may be applied to the fabric and dried before it is printed instead of being included in the ink.46 Another system involves pre-treating fabric with a solution containing multivalent metal salts (e.g. magnesium nitrate). The ink contains

12.1 Anionic dispersing agent.

Substrate preparation for ink-jet printing

213

pigments that are coated (or encapsulated) with resins (e.g. styrene acrylic resins) coloured by oil-soluble dyes.47

12.4.7 Paper for transfer printing The paper for transfer printing is used simply as a carrier for the printed image, and has little or no use after transfer of the design onto the substrate. Therefore the less it costs the better, but it should not be so thin that dye vapour easily passes through it onto the heated surface behind it during transfer. Highly calendered paper based on a kraft pulp is generally favoured.48 Inorganic fillers, such as finely divided calcium carbonate, china clay or titanium dioxide, are also likely ingredients. It is also advisable to apply a coating or size to the side to be printed. This serves to provide a smoother printing surface, and one which is impermeable to disperse dye vapour. Sodium alginate, for example, is impermeable18 and so is a good choice as a size for transfer paper. Starch products, such as sodium carboxymethyl starch, on the other hand, have a granular structure, which allows free passage of disperse dye vapour.

12.5 Post-treatments When the pre-treated fabric has been dried and then jet printed there is usually little need to provide a drying station to dry the print, as the printing process is so slow. By the time the fabric is batched on a roll it has dried by exposure to the warm atmosphere in the room. If necessary supplementary heating is applied. Sometimes the printed fabric is only required for photographs in a catalogue, in which case it may not be necessary to fix the print. However, in most instances fixation and washing will be necessary. This not only ensures that the full fastness properties of the dyes are realised, but also brightens and alters the colours significantly.

12.5.1 Fixation Steaming is the process normally used to fix printed textiles. Reactive and acid dyes are steamed under atmospheric pressure at just over 100ëC. During the process steam condenses on the fabric and is absorbed by the thickener and hygroscopic agents in the printed areas. Dyes and chemicals dissolve and form extremely concentrated dyebaths within the thickener film. As a result of the extremely low liquor ratio (approximately 1:1) fixation is much more rapid than in exhaustion dyeing. High temperature steam is necessary for the fixation of disperse dyes on polyester. The Tg of polyester in steam is lower than it is in dry air, and fixation is more efficient. Usually the steam is heated to 170±180ëC at atmospheric pressure, but sometimes pressure steaming at 130±150ëC is used. Pigment prints are cured using hot air in a stenter or a roller baker.

214

Digital printing of textiles

12.5.2 Washing-off Washing the fixed jet printed fabric is likely to be carried out on a batchwise basis, since the lengths printed are usually quite short. The process, whether batchwise or continuous, takes place in several stages, the first of which should be a cold rinse. The reason for this is that thickener, auxiliaries and loose dye should be removed under conditions where the dye is unlikely to stain white or unprinted ground shade areas. The risk of this happening may be reduced further by the inclusion of reagents that hold the unfixed dye in the bath. For instance, inclusion of 1 g/L sodium carbonate in the first bath when washing off acid dyes prevents protonation of amino groups on nylon, wool or silk, thus removing any potential dye sites. The main requirement after the first cold rinse when washing off reactive dye prints is to ensure that the temperature of the hot wash reaches a minimum of 95ëC, otherwise hydrolysed dye may not be removed. Fabrics printed with disperse dyes may require a reduction clear with alkali and sodium dithionite (Na2S2O4) in order to remove surface dye. The temperature of the bath for this process should not exceed 70ëC as sodium dithionite (also known as sodium hydrosulphite or `hydros') is unstable at higher temperatures.

12.6 Jet printing machines These have been discussed in Part I of this book, and reviewed by Dawson.49 A survey of the machines exhibited at the 2003 ITMA show in Birmingham has also been published.50 Most of the machines sold for printing textiles are modifications of graphics printers. In the early days the most popular machines were bubble jet (thermal drop on demand) printers, but more recently piezo printers, many of them made by the Japanese company Mimaki, have taken over. Stork also introduced a machine based on the continuous (charged drop) principle,51 but the high cost of the machine reduced its appeal.

12.7 Limitations Nearly all the fabric jet printing machines marketed up to the present use a print head that scans across the fabric while it is stationary. The fabric then indexes forward while the head is stationary at the side. In this respect they resemble desk-top paper printers and flat-screen fabric printing machines. Jet printers would be much more productive if they utilised print heads which were stationary, but covered the full width of the substrate.52 Printing would then be fully continuous, i.e. the fabric would move steadily while it was being printed, as is the case with the Millitron and Chromotex carpet printers, and in rotary screen printing. Productivity could be still further enhanced if a machine were designed that included an applicator for the pre-treatment formulation coupled

Substrate preparation for ink-jet printing

215

with a drying station, in line with the print heads. This design was suggested in 1997 by Hawkyard,52 and has been the subject of a patent application more recently.53

12.8 Future trends The breakthrough for jet printing textiles on a large scale is likely to be as a result of the introduction of a pigment printing system that is superior to those presently available, and one that is lower in price. However, the use of multifunctional dispersing agents (MFDA) in pigment ink formulations 14 may herald the long-awaited breakthrough. These dispersing agents contain polymeric segments within their structure, and so obviate the use of traditional binders. The fact that pigment printing does not require a fabric pre-treatment is a massive incentive, as is the simple fixation step and the lack of a wash-off. The cost of reactive dye inks is also prohibitive. While the trend for shorter and shorter run lengths continues,54 the prospects for jet printing improve. When run lengths are short it would probably be better to employ a bank of several graphics-style printers, as have been used by the Japanese company Seiren, rather than one large and expensive continuous printer.

12.9 Bibliography Dawson T L, `Jet printing', Rev. Prog. Col. 1992, 22, 22±31. Dawson T L and Ellis H, `Will inkjets ever replace screens for textile printing?', JSDC, 1994, 110, 331±337. Dawson T L and Glover B (eds), Textile Ink Jet Printing, Bradford, UK, SDC, 2004. Dawson T L and Hawkyard C J, `A new millennium of textile printing', Rev. Prog. Col., 2000, 30, 7±19. Hawkyard C J, `Fit to print', Textile Tech. Int., 1996, 223±227. Hawkyard C J, `Digital textile printing ± ready for take-off', Textile Tech. Int., 2000, 56± 57. Kulube H M and Hawkyard C J, `Fabric pretreatments and inks for textile ink-jet printing', Int. Text. Bull., Dyeing, Printing, Finishing, 1996, 3, 4±15. Miles L W C, Textile Printing, 2nd edn, Bradford, UK, SDC, 1994. www.storktextile.com

12.10 References 1. 2. 3. 4.

Dawson (Holdings), BP 1,284,824 (1971). Wild K, JSDC, 1977, 93, 185. Milliken, USP 3,969,779 (1974). Provost J R and Connor H G, `The printing of polyester/cellulose blends ± a new approach', JSDC, 1987, 103, 437±442. 5. Dawson T L and Ellis H, `Will inkjets ever replace screens for textile printing?',

216

Digital printing of textiles

JSDC, 1994, 110, 331±337. 6. Babaei Lavasani, M R and Hawkyard C J, `Drop formation in ink-jet printing of textiles with solenoid valves', Melliand Int., 2000, 6, 152±155. 7. British Standard 4554: 1970 `Method of Test for Wettability of Textile Fabrics'. 8. Hawkyard C J, Babaei Lavasani M R, Khaled K K and Singh P, `An automated test method for wettability', Proc. 1st Autex Conf., 2001, Vol. 1, Pavoa de Varzim, Portugal, 344±350. Apparatus available from SDL International. 9. Madaras G W, Parish G J and Shore J, Batchwise Dyeing of Woven Cellulosic Fabrics: A Practical Guide, Bradford, UK, SDC, 1993. 10. Provost J R and Aston S O, ICI, GB 2252335, 1992. 11. Seiko Epson, JP 2004162247, 2004. 12. Seiko Epson, JP 2004210806, 2004. 13. Hauser P J and Buehler N, Image Sci. Tech., 1991, 35, 179. 14. Hees U, Freche M, Provost J R, Kluge M and Weiser J, `Textile ink jet printing with pigment inks', in Textile Ink Jet Printing, ed. Dawson T L and Glover B, Bradford, UK, SDC, 2004, 57±63. 15. Sapchookul L, Shirota K, Noguchi H and Kiatkamjornwong S, Surface Coatings Int., Part A, Coatings Journal, 2003, 86(A10), 403±410. 16. Seiren, JP 2004306469, 2004; USP 2004201660, 2004. 17. Salters Advanced Chemistry, `Chemical Storylines', Oxford, Heinemann, 1994, 225±226. 18. Hawkyard C J, `The release of disperse dyes from thickener films during thermal processes', JSDC, 1981, 97, 213±219. 19. Toray, WO 9208840, 1992. 20. Provost J R, `Effluent improvement by source reduction of chemicals used in textile printing', JSDC, 1992, 108, 260±264. 21. Seiko Epson, JP 2004143621, 2004. 22. Meisei, JP 2004115953, 2004. 23. Toyo Boseki, JP 06200484, 1994. 24. Scitex Digital Printing, EP 828024,1998. 25. Kanebo, JP 06146178, 1994. 26. Toray, JP 02300377, 1990. 27. Dawson T L, `Spots before the eyes ± can ink jet printers match expectations', Colouration Tech., 2001, 117, 185±192. 28. BASF, DE10244998, 2004; WO 2004031473, 2004. 29. Canon, JP 08120576, 1996. 30. Yang Y, Li S and Stewart N, `One-step inkjet printing and durable press finishing', AATCC Review, 2003, 3(3), 29±31. 31. Lewis D M and McIlroy K A, `The chemical modification of cellulosic fibres to enhance dyeability', Rev. Prog. Col., 1997, 27, 5±17. 32. Kanik M, Hauser P J, Parrillo-Chapman L and Donaldson A, `Effect of cationisation on inkjet printing properties of cotton fabrics', AATCC Review, 2004, 4(6), 22±25. 33. Miles L W C, Textile Printing, 2nd edn, Bradford, UK, SDC, 1994, Chapter 7, 250± 252. 34. Seiren, JP 2004131919, 2004. 35. Avecia, WO 9955955, 1999. 36. Taoka Chemical Co., JP 11302987, 1999. 37. Toray, JP 0926842, 1997.

Substrate preparation for ink-jet printing 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50.

51. 52. 53. 54.

217

Canon, JP 10025671, 1998. Canon, JP 08127981, 1996. Seiren, JP 2001295186, 2001. Canon, EP 992623, 2000. Dairiki, JP2003003385, 2003. Toray, JP62170591, 1987. Toray, JP61231288, 1986. Seiren, US 2002025379, 2002; EP 1188856, 2002; JP 2002 161486, 2002. Ciba Specialty Chemicals, WO2000003081, 2000. Seiko Epson, JP 2003055886, 2003. Rattee I D, Textile Printing, 2nd edn, ed. Miles L W C, Bradford, UK, SDC, 1994, Chapter 3, 62±64. Dawson T L, `Jet printing', Rev. Prog. Col. 1992, 22, 22±31. Glover B, `The latest technology developments in ink jet printing from ITMA 2003', in Textile Ink Jet Printing, ed. Dawson T L and Glover B, Bradford, UK, SDC, 2004, 13±29. Aston S O, Provost J R and Masselink H, `Jet printing with reactive dyes', JSDC, 1993, 109, 147±152 . Hawkyard C J, `Bubble-jet ± a serious challenge to conventional printing?', Int. Dyer, 1997, January, 17±18. Inktec Co., WO 2004085739, 2004. Teunissen A, Kruize M and Tillmanns M, Developments in the Textile Printing Industry 2002, Stork Textile Printing Group, Communications Dept., Boxmeer, Netherlands, 2002, 6.

13 Pigmented ink formulation Z F U , Rohm and Haas Company, Philadelphia, USA

13.1 Introduction Ink jet digital printing of textiles is an emerging market that presents both tremendous opportunities and significant challenges. The worldwide production of printed fabric is approximately 34 billion square yards and is dominated by rotary screen printing. During the past eight years or so the textile industry, the printer OEMs, designers, entrepreneurs and the raw material and ink suppliers have expressed great interest in digital printing technologies for design, sampling, and short run production. In other chapters of this book textile ink, particularly dye-based, ink jet print head, color management software, fabric handling equipment and fabric pre- and post-processing technologies have been discussed in detail. In this chapter, I will be focusing on polymer binder containing pigmented aqueous textile inks. I will provide general guidance on pigmented ink jet ink formulation for digital textile printing, including overall considerations, pigment dispersion selection, binder selection, surfactant selection, co-solvent and humectant selection and rheology modifier selection. I will also provide the relevant properties to test and test methods for pigmented ink development, including (1) basic physical properties, (2) ink and print head interaction properties, such as ink and print head compatibility, start-up, jetting and printing, (3) ink and media interaction properties and image quality, (4) durability and permanence of printed images such as cure conditions, wash, crock and dry cleaning, and (5) safety for both the handling of the ink itself and the printed fabric and garments. I will briefly discuss some optional pre- and post-treatments for pigmented digital textile printing and emphasize the importance of white ink for dark color T-shirt printing. I will conclude the chapter with further information and advice on some books to consult, major trade/professional bodies, research and interest groups and websites.

Pigmented ink formulation

219

13.2 Overview Initial development in ink jet textile printing focused primarily on dye-based inks such as acid dyes for silk, nylon, and wool, disperse dyes for polyester, and reactive dyes for cotton and rayon. In order to achieve adequate wash properties, they all require complicated and somewhat environmental unfriendly pre- and post-treatments. To some degree, these pre- and post-treatments defeat the original purpose of digital printing in term of ease and fast customization. On the other hand, pigment-based inks are more versatile in terms of fiber types and require only simple dry heat post-treatment (Table 13.1). Traditional pigment textile printing with emulsion-based textile binders enjoys more than 50% of the total printed textile market. However, development of textile pigment inks with emulsion-based textile binders for ink jet printing is extremely challenging due to ink stability and jetting reliability (drying and nozzle clogging) issues, especially for low viscosity print-heads. Table 13.2 summarizes a view held by some key players in the industry. The unconventional textile binder on the third row in Table 13.2 can be any new polymeric binder developed for pigmented ink jet inks, which may include ink medium soluble but water insoluble random and block copolymers or dispersant, binder and other property combined cross-linkable multi-functional agents.

Table 13.1 Comparisons of ink chemistry Ink type

Fiber types

Pre-treatment

Post-treatment

Acid Disperse

Silk, nylon, wool Polyester

Acid donor Thickener

Reactive Pigment

Cotton, rayon All, best on cotton, polyblends

Alkali Not required

Steam and wash High-temperature steam and wash Steam and wash Dry heat

Table 13.2 Print head type and viscosity versus ink formulation Piezo head, low viscosity

Piezo head, high viscosity

Thermal print head

Pigment ink without textile binder

Yes

Yes

Yes

Pigment ink with conventional textile binder

No

Yes

No

Pigment ink with unconventional textile binder

Yes

Yes

No

220

Digital printing of textiles

Pigment inks for digital textile printing present different challenges than dyebased textile inks. Pigment inks contain 50 to 150 nm crystalline particles as a colloidal system while dye inks are uniform dye molecule solutions. The permanence properties such as wash and crock come from acid (on the dye molecule) and base (on the fabric) interaction for acid dyes, from the dye solubilization in the polyester fibers for dispersed dyes and chemical reaction between the dye and the fabric for the reactive dyes. On the other hand, for pigment inks, the acid±base interaction and/or chemical reaction between pigment particle and the fabric, if any, are simply not enough and a soft, low glass transition temperature (Tg ) polymeric binder is required to achieve adequate permanence properties such as wash and crock. The challenges associated with formulating soft low Tg polymeric binder containing pigmented textile ink are as follows: · To achieve adequate (1 to 2 years) stability and shelf lifetime with regard to sedimentation, homo- and hetero-coagulation and phase separation. · To maintain the low viscosity and formulation space for jetting reliability and at the same time to load enough binder for wash and crock resistance, especially for low viscosity print heads. · To keep the soft and low Tg binder from clogging the nozzles. The above difficulties and relatively lower color density, less vibrant colors and smaller color gamut than dyes are probably the reasons for the slow development and relatively low percentage usage of pigmented textile inks for ink jet printing. This may also be associated with the choice of fabric type in current digital textile printing. Since the end use of digitally printed fabrics or assembled garments should not be very different from that of conventionally printed, there is no reason to believe that the pigmented inks for digital textile printing will not reach the same percentage (50%) as in the conventional textile printing industry. I would predict the percentage may be even higher due to the fact that it does not require pre-treatment or complicated post-treatment, which has great synergy with digital printing for simple, easy and fast customization. Easy set-up, fast customization and photographic image quality are the major advantages for digital textile printing, and slow speed and relatively high cost per unit printed area are the major disadvantages. Pre-assembled garments such as T-shirts and ready and easy to assemble items such as flags, scarves, banners, backdrops, etc., are the best places to take advantage of digital printing while minimizing the impact of the above disadvantages of digital printing. These niches will lead the evolution of textile printing from conventional to digital. The key components in digital textile printing systems are the printer, especially the fabric handling and the garment mounting system, the software, including printing, color management and workflow management, and the ink. Among these three key components, the ink, especially the pigment ink, requires the most advancement to approach the same level as traditional printing in terms

Pigmented ink formulation

221

of color strength, permanence, speed and reliability. Some of the leading developers in pigment textile inks for digital printing are Rohm and Haas, DuPont, BASF, Ciba, Trident and Sensient.

13.3 Pigmented ink formulation for digital textile printing This section will focus on the ingredient selection for a typical pigment ink for digital textile inkjet printing and the function(s) of each ingredient will also be discussed.

13.3.1 Overall considerations A typical pigment ink formulation for ink jet digital textile printing includes: · · · ·

·

· · · · ·

A pigment dispersion for color A polymeric binder, a solution polymer or latex for image durability Water, for aqueous inkjet inks ± a medium to carry other components A co-solvent, helping water to carry other ingredients through solubility and compatibility and enhancing the performance of other ingredients in terms of wetting and adhesion to the substrates and jetting properties Surfactants, for nozzle and substrate wetting and jetting reliability and also for stabilizing the key ingredients such as binder and pigment particles from coagulation Humectants, to prevent drying when not printing An antifoam agent to reduce foaming A viscosity control agent for damping control and droplet formation A penetrant to speed drying on porous media such as paper and textile A biocide to prevent spoilage.

Aqueous pigmented ink jet inks have existed for some time. But the challenge for textile printing is to incorporate enough binder in the ink for washability and at the same time to maintain low viscosity, ink stability and jettability. To maintain low viscosity, ink stability and reliable jetting, solids level and formulation space are limited. It is important and sometimes critical for an ingredient to serve multiple functions. For example, a nonionic surfactant of long chain polyethylene glycol with a hydrophobe serves as a surfactant to aid stability and to control wetting and surface tension, as a humectant to slow water evaporation and therefore preventing the ink drying near the nozzle, and as a rheology modifier to control the viscosity profile and therefore jetting and drop formation. A cross-linkable and ink medium soluble but water insoluble polymer could serve as a dispersant to disperse the pigment, as a binder to bind the pigment particles to the fabric for wash and rub resistance, and as a rheology modifier and possibly as a humectant. The key ingredient selection, their

222

Digital printing of textiles

functions and interaction and synergy, will be discussed in detail in the following sub-sections.

13.3.2 Pigment dispersion selection Pigment dispersion is probably the most important ingredient, and the rest of the ingredients in the ink jet textile ink formulation either serve to position the pigment particles to the right place on the substrate through a given print head in order to generate beautiful images, or to bind the pigment particles to the substrate so that the image can last a long time with respect to different types of resistance such wash and rub. The pigment and the set of pigments selected determine the color gamut, the color density, the brightness and the UV resistance of the individual ink and the ink set. These properties are not unique to ink jet and are similar to conventional printing inks, so they will not be discussed further here. The unique properties which are important in pigment dispersion selection for ink jet inks are stability in terms of the formulated inks, particle size and size distribution, viscosity, surface tension and pigment solids. The stability of the pigment dispersion with respect to a variety of formulation ingredients, solvents, low surface tension surfactants and polymeric binders is very important because if the ink is not stable, the other properties become meaningless. The particle size and size distribution, on the one hand, affects the image quality, especially color density, in terms of ink holdout and the effective use of each pigment particle for light absorption. Bigger particles are probably good for holdout, leading to higher color density, while smaller particles are probably better in terms of the effective use of each pigment particle for light absorption. When there are competing factors, optimum particle size exists for a given substrate. Although it is not relevant for textile printing, smaller particles tend to yield better gloss for glossy substrates. On the other hand, the particle size and size distribution have a lot to do with settling of the pigment in the ink, colloidal stability, and clogging of the nozzle and therefore the jetting reliability. If the particle size is too big or the distribution is too broad, it will have an adverse effect in start-up and reliable jetting due to settling and clogging. Too many small particles (below 0.05 microns) can also have adverse effects in terms of stability and jetting reliability, because smaller particles have higher surface area to volume ratio, and therefore higher dispersant demand, and at the same time the surface area per particle is smaller, therefore if the zeta potential or charge density is the same, smaller particles have less total charge per particle and the repulsive barrier is lower than that of big particles and in turn may be less stable. Small particles may also provide high probability for clogging due to particle congestion. Typically pigment dispersions with low viscosity and high surface tension are more preferable. Low viscosity in pigment dispersion means leaving more room

Pigmented ink formulation

223

for other ingredients such as polymeric binders and greater ease in maintaining overall viscosity of the final inks. If the surface tension of pigment dispersions is high, there are always ways to lower the surface tension of water-based systems by adding surfactants and organic solvents. On the other hand, if it is low, it will put a limit on the surface tension of the final ink. Finally the solid level is a very important concern. Typically higher is better, because it leaves more room and flexibility for adding other ingredients to the ink formulation without compromising the pigment solids loading in the final ink: 10% is at the lower end of most pigment dispersion suppliers, 30 to 40% is at the high end and 20% is typical. There are two major types of pigment dispersion, polymeric dispersant stabilized dispersions and self-dispersed dispersions. The type is not important as long as they provide the right properties outlined above. The self-dispersed type tends to be more stable with respect to solvent selection, while polymeric dispersant stabilized dispersions tend to have better permanence, such smear resistance benefited from the polymeric dispersant. Some of the key pigment dispersion suppliers are Rohm and Haas, Lanxess, Clariant, and Cabot. This is by no means a complete list.

13.3.3 Binder selection While pigment is an important ingredient to provide the image, the binder is another key ingredient to maintain the permanence of the image with respect to washing and rubbing in the case of textile printing. Incorporating polymeric binders in ink jet inks is difficult in general. It is even more difficult for textiles because it requires not only high levels but also low Tg for good hand and feel. The following example will help to illuminate the difficulty. Most people have the experience of using Elmer'sTM glue, which is a soft and low Tg polymeric binder. How often we need to cut the tip in order to remedy the clogging to use it again, and it has only one big nozzle. As stated in Section 13.2, the specific challenges are · To keep the soft and low Tg binder from clogging the nozzles. · To maintain the low viscosity and formulation space for jetting reliability and at the same time to load enough binder for wash and crock resistance, especially for low viscosity print heads. · To achieve adequate (1 to 2 years) stability and shelf lifetime with regard to sedimentation, homo- and hetero-coagulation and phase separation. To overcome these difficulties, leading companies in this field such as DuPont, Rohm and Haas, and BASF have developed proprietary polymer and formulation technology, which may include ink medium soluble, but water insoluble random or block copolymers with or without cross-linking functionality, dispersants, binders and other property combined cross-linkable multi-functional

224

Digital printing of textiles

agents and protected cross-linkable latex polymers. For example, in the case of protected latex polymers, the soft and sticky binders are protected by a thin layer of shell, like an egg; it is not sticky in the ink before the protective shell is broken, and becomes sticky glue only when the protective shell is broken in the later curing stage.

13.3.4 Co-solvent and humectants Pigment is the important ingredient to provide the image, and binder is the key ingredient to maintain the permanence of the image. But both need a carrier to deliver them to the substrate. In the case of water-based pigment textile inks, the carrier is mainly water (50 to 80%) mixed with water-soluble organic compounds, which are called co-solvents and humectants based on their functions. Co-solvents are organic compounds such as 2-pyrrolidone and propandiols, which help water to incorporate other ingredients into the system better. For example, 2-pyrrolidone may help water to dissolve some surfactants better and to make some polymers more soluble. Humectants are hydroscopic organic compounds such as polyethylene or propylene glycols with or without alkyd ether capping groups on one or both ends, glycerol, sorbitols, etc. Hydroscopic means capable of `pulling' water vapor from the air back to the liquid phase, which slows down or completely stops the drying of the ink when humectants reach a certain concentration under a given humidity and temperature condition. This is very important to prevent the ink from drying on the nozzle and from clogging the nozzle both during the printing and in the idling state. A single ingredient or compound, for example, 2-pyrrolidone, often serves as both cosolvent and humectant.

13.3.5 Surfactant selection Surfactants are another key ingredient in terms of delivering the pigment and the binder from the ink to the substrate through the print head. High HLB (hydrophilic and lipophilic balance) surfactants are used typically for aiding the colloidal stability of the systems, and low HLB surfactants are used to lower the surface tension, so the ink can wet the nozzle capillary to establish and maintain the meniscus at the nozzle tip. The importance of maintaining the meniscus at the nozzle tip both in the steady state and in the dynamic state during jetting cannot be overemphasized because it is so critical for start-up, reducing latency (defined as number of firings needed before the ink establishes the first stable drop of jetting), increasing the elapsed time between jetting without refreshing and ultimately long-term reliable continuous printing. For some print heads, reliable jetting or printing can be achieved even when the nozzle plate is wetted. This low HLB surfactant is also a major factor which determines the interaction between the ink and the substrate and therefore controls or affects wetting,

Pigmented ink formulation

225

bleeding, dot-gain, dot-quality and ultimately the image quality. Surfactants affect these properties through a physical parameter, namely surface tension (both static and dynamic). The most popular surfactants used for this purpose are relatively short-chain ethylene glycol nonionic surfactants such as the Air Products SurfynolTM line of products like SurfynolTM 465. Anionic surfactants such as AerosolTM OT are also used.

13.3.6 Rheology modifier selection While a surfactant is the key ingredient used to control surface tension, a rheology modifier is the ingredient used to control the viscosity of the ink, or more precisely the rheology profile of the ink, which includes the yield stress and the viscosity at different shear modes and rates. The yield stress along with the meniscus has a major effect on the latency. The viscosity at low shear rate along with the surface tension determines the fill-up and the priming of the nozzles to establish the initial meniscus and the ready-to-jet condition. The viscosity at high shear rate, up to 1 million reciprocal seconds, is probably more relevant to fluid dynamics and the drop formation at the nozzle tip and thereafter. The viscosity and the mass density of the ink affect the oscillation and the damping of the ink chamber and therefore the jetting. Water-soluble polymers such as polyethylene glycols (PEG) with molecular weight ranging from several hundreds up to hundreds of thousands could be rheology modifiers. The rheology modifiers along with the co-solvents, humectants and the total solids in the case of pigmented inks with or without binders together determine the viscosity. Ideally the rheology modifier is selected with the following considerations in mind. 1.

2.

3.

It should not strongly associate with multiple pigment and/or latex particles, causing coagulation and precipitation, unless the coagulation is well controlled. Associative thickening should be avoided since even loose association through the rheology modifier molecule may increase yield stress and cause phase separation. Depletion flocculation should also be avoided because it may cause phase separation and non-uniform color density in the print.

In case 2, the rheology modifier molecule will be more likely in the particle rich phase, while in case 3 the opposite applies in that the rheology modifier molecule will be more likely in the particle deficient phase. In general, the viscosity of ink jet ink is low, below 20 cps. Even at this low viscosity, the rheology profile cannot be overlooked and it may be important that it is controlled intentionally with relatively high molecular weight water-soluble polymers.

226

Digital printing of textiles

13.3.7 Other miscellaneous ingredients selection Pigments, along with dispersants, binders, co-solvents, humectants, surfactants and rheology modifiers (viscosity adjusters), are the main functional ingredients of pigment ink jet ink formulation for textile printing. Other ingredients such as de-foamers, penetrants and biocides may be added as needed. The defoamer is for defoaming as its name suggests. The penetrant helps the ink vehicle to be absorbed into the substrate faster and therefore may make the print touchable sooner. Biocide is for preventing bacteria growth and maintaining ink shelf lifetime. It is always a good rule to keep the formulation simple; if not needed, do not put it in or take it out. When possible, use one ingredient to serve multiple functions.

13.3.8 Putting them together ± synergies among all the ingredients In the previous sections, the function and selection of each ingredient were discussed. In this section, I will briefly talk about how to put them together and the synergy among all the key ingredients. As stated in the previous sections, pigment is to provide the image, binder is to keep the image permanent, water along with co-solvents is the vehicle to carry the pigment and binder, and surfactants and rheology modifiers are to provide the right surface tension and rheology for reliable jetting through the nozzle and for the proper interaction with the substrate to create a high quality image. Each ingredient needs to do its own job and at the same time to work together in harmony. If the binder is polymeric latex particles, the binder particles and pigment particles need to have the right energy balance so that they will not be so attractive to each other to create stability problems and at the same time they should not repulse each other in a way to create pigment and latex particle phase separation. Co-solvents such as 2-pyrrolidone along with water are not only a carrier for pigments and binders, but also help to dissolve the low HLB surfactant to make it more effective and to transport it quickly through the medium to ensure low dynamic surface tension. Co-solvents and low HLB surfactants work together to create better wetting condition for both the nozzle and the substrate. Co-solvents often soften the binder and substrate to enhance adhesion. Long EO chain high HLB surfactants can serve as stabilizers, humectants and also rheology modifiers. They can also emulsify the low HLB surfactant to prevent it from forming an oily layer on top. The low and high HLB surfactants together with the co-solvent may form a good cleaning solution for metal and semiconductor surfaces, so the ink is self-cleaning. Polyethylene glycols (PEG) serve as both humectants and rheology modifiers.

Pigmented ink formulation

227

13.4 Tests and test methods for pigmented textile inks As in any field, in aqueous pigment ink jet textile ink formulation, knowing the ingredients and their functions is not enough; we also need to know the important properties and how to test these properties.

13.4.1 Overview of tests and test methods The following 22 properties/test methods are important in developing pigmented textile ink jet inks. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Viscosity Surface tension pH Particle size (PS) and PS distribution Total solids Ink mass density Ink filterability Foaming Air content in the ink and degassing. Drying rate at controlled temperature and humidity and re-dispersability

11. Heat-aging stability (3 days, 10 days and 28 days by PS, ST, viscosity, pH, total solid at 60ëC) 12. Settling rate at regular or accelerated gravity 13. Phase separation 14. Ink fill-up 15. Continuous jetting reliability (x number of square meters of continuous printing without defect) 16. OD or color density 17. L.a.b. 18. Printability (resolution, inter-color bleeding, and color uniformity) 19. 3A wash fastness 20. Dry and wet crocks 21. Dry cleaning fastness 22. Regulatory and safety (wet ink and dry printed sample). These 22 properties and test methods are separated into seven groups, which consist of a basic property group (1 to 6), a jetting reliability property group (7 to 10), a shelf lifetime property group (11 to 13), actual jetting reliability tests (14 and 15), an image quality related property group (16 to 18), an image permanence property group (19 to 21) and a regulatory and safety group (22).

228

Digital printing of textiles

13.4.2 Basic physical properties The first six properties (1 to 6) are the basic physical properties. The required value may be unique, but the properties themselves are not. These properties, especially the first three, are specified for a given design of print head. The viscosity (or more precisely the rheology) determines the fluid dynamics for the specific design of the print head geometry, e.g. the sufficient supply of the ink for starting the jetting and during jetting and the drop formation. Otherwise, start-up, latency and ink-starvation massive nozzle dropout can occur. The simple low shear rate BrookfieldTM viscosity is far from enough to know the rheology profile of the ink. Even if the rheology profile is known, it is still difficult to predict the fluid dynamics and drop formation due to the complicated geometry of the ink pathway and the driving waveform. One thing print head manufacturers have done is to do computer-based simulations using finite element analysis methods. Nonetheless, the simple Brookfield viscosity provides a starting point and gives the formulators something to think about. The optimum surface tension for jetting is determined by the surface energy of the channel of the nozzles vs. that of the front face of the print head in such a way that the ink does not ooze out and wet or dirty the front face, but has the maximum force to wet the channel to maintain the proper meniscus at the orifice of the nozzle. The proper meniscus at the orifice of the nozzle must be maintained both when it starts to jet and during the jetting, which means both the static and dynamic meniscus needs to be right. This requires both static and dynamic surface tension to be right. Again the simple static surface tension only provides a starting point and gives the formulators something to think about. In some situations, stable jetting can be achieved even when the front face is maintained wet during jetting. In this case, the ink has to be slow-drying, otherwise nozzle clogging or partial clogging could be a major issue. It is also important to understand that even when the static surface tension of the ink is lower than the surface energy for the front face of the nozzle, the front face may not be wet during the jetting because the dynamic surface tension may be higher than the surface energy of the front face. The optimum surface tension for jetting is not likely to be the optimum surface tension for wetting or controlling the bleeding or coalescence on the substrate of interest. Compromise is unavoidable. The pH and other chemical characteristics of the ink are important in terms of its compatibility with the construction materials of the print head. The compatibility means that the ink should damage neither the print head materials nor the adhesive bond between different head parts and also that the print head materials should not leach out any material to change the ink composition or properties in any way affecting the stability, jetting or ink performance on the substrate. The particle size and distribution, total solids and mass density of the ink are properties affecting jetting in very subtle ways.

Pigmented ink formulation

229

13.4.3 Ink and print head interaction properties Beside the chemical compatibility between ink and print head mentioned in the previous section, the next four properties (7 to 10) are particularly important in terms of jetting reliability, especially for piezo-based print heads. The ink flow from the ink supply to the substrate during the ink jet jetting process is indeed a series of filtering processes through very small (micron size) orifices such as filters in the damper and the nozzles themselves at very high shear rate (~1 million sÿ1). The filterability is related to both the content of large particles and shear stability to some degree. It is highly important in ink jet inks because it could shorten the lifetime of the print head. The air content can be measured using an oxygen meter. Typical water contains about 7±8 mg oxygen per gram or 7±8 ppm. For piezo print head, less than 4 mg/g or ppm may be desirable. This is because the high air content in the ink (foam or dissolved) makes the compressibility of the ink high, and when the compressibility is high, the periodically pulsed jetting force or energy is converted to thermal energy and is wasted, instead of being used for ejecting the ink drop. This is particularly important for piezo heads because the jetting push for piezo is relatively high force, short duration and low volume capacity. When the bubbles are big (foam), they could clog the nozzles for both piezo and thermal print heads. Both chemical and mechanical methods of defoaming and degassing have been used. The drying rate at a given temperature and humidity determines the idle time and the continuous jetting reliability at that temperature and humidity due to crusting, and these problems magnify when the front face gets dirty due to oozing, misting and filament springback due to poor drop formation. The redispersability of the ink also determines how easily the `spitting' and `wiping' of the maintenance cycles can recover the problem.

13.4.4 Shelf lifetime properties and actual jetting reliability tests Properties 11 to 13 determine the shelf lifetime of the ink. The heat-aging stability test is an accelerated test, which measures basic ink properties at different time intervals at 60ëC heat aging. One week of heat aging is approximately equivalent to three months at room temperature (20±25ëC), and four weeks is approximately equivalent to one year. The test can and should be extended to all other properties at the desired final heat-aging time interval before commercialization. Settling rate at regular or accelerated gravity and phase separation are two other properties which are very important in determining shelf life time, especially for polymeric binder containing pigment ink jet inks. Properties 14 and 15 put properties 1 through 10 into real jetting tests. Test 14 is about how easily one can get the print going perfectly when a new set of ink cartridges is installed, while test 15 is about how long the reliable jetting can last

230

Digital printing of textiles

once the printing gets going. Tests 14 and 15 can be performed on cheap paper substrates. They do not need the right textile substrate because they only test how well the ink interacts with the printer in terms of jetting reliability. In the next section, we will discuss how well the ink interacts with the intended textile substrate in terms of printing quality.

13.4.5 Ink and media interaction properties and image quality Properties 16 to 18 are about how the ink interacts with the substrate and the image quality for the textile substrate of interest. Optical density (OD) for black, and color density for cyan, magenta, and yellow, are used to assess the intensity that each process color ink can offer, while the L.a.b. measurements provide the full color space (gamut) which a given set of process color inks can achieve for a given printer, printing mode, software and substrate combination. The color density and gamut information is typically stored in a file called the color profile for the given combination. The color intensity and color gamut along with resolution (how fine a line can be printed), inter-color bleeding (how sharp the interface can be between two colors) and color uniformity determine the image quality and the printability. These properties are obtained through spectra colorimeters and image analysis software. To obtain a high quality image is the first step while making the image durable and permanent is the next key step in digital textile printing.

13.4.6 Durability and permanence of printed images Durability and permanence of the printed images are the key for a successful set of pigment inks for inkjet textile printing. Unlike the reactive, acid and dispersed dyes, the pigmented inks rely on external polymeric binders to achieve the durability and permanence of the printed images. The polymeric binders provide the permanence properties through mechanical binding of film formation and chemical interaction and cross-linking with the substrate. The mechanical binding works for all fiber types of textile substrates; cotton, polyester, cotton and poly blend, and silk to name a few. But chemical interaction and crosslinking only occur when the polymer and the fiber of the substrate have the right pair of chemical reaction functional groups. Therefore, people claim the bindercontaining pigment inks work for all fiber types due to the mechanical binding and work best when there is chemical cross-linking between the polymer and the fiber. The hydroxyl group on the cotton fiber is one of the best examples. Both mechanical binding and chemical cross-linking require the printed samples to be heated. The heating temperature and time depend on the glass transition temperature (Tg ) for the mechanical binding and the chemical reaction rate for the cross-linking, which typically ranges from 100ëC to 200ëC and from 30 seconds to 5 minutes. Two types of heating devices are used for the image

Pigmented ink formulation

231

fixation: heat presses and hot air ovens. The heat transfer of hot air ovens is through air flow while the heat press is by contact, so the hot air oven takes longer and/or needs a higher temperature than the heat press to reach the same curing point. In the case of a heat press, not only the time and temperature but also the pressure is very important to determine the cure and the fixation for wash and rub, because higher pressure means better heat transfer. The cure conditions need to be optimized for each ink set and fabric combination to achieve the best possible permanence properties without damaging the fabrics. After proper curing and fixation, water and detergent wash fastness, wet and dry crocks or rubs and dry cleaning fastness tests (19 to 21) can be performed, based on standard test methods from the Technical Manual of the American Association of Textile Chemists and Colorists, in the same way as for conventional screen pigment printing.

13.4.7 Regulatory and safety considerations Although regulatory and safety considerations are beyond the scope of this chapter, their importance should not be overlooked. Both the wet ink and the dried print samples need to be safe for handling and wearing or use. Before commercialization, the safety and regulation of both the ink and printed samples should be assessed by the right professionals.

13.5 Optional pre- and post-treatments for pigmented digital textile printing The pre-treatment for pigment printing is optional and the post-treatment is simple: dry heat. For pigment digital textile printing, due to limited formulation space and low viscosity, the pigment, binder and other ingredients loading are limited, so pre-treatment with polymers, cationic materials, catalysts, reactive cross-linking agents, etc., may be needed in some applications to enhance color density, image quality, durability and permanence of the image. Spray or screen print polymer binders over the printed images may be a necessary post-treatment for difficult substrates or applications which require extra durability.

13.6 White ink White ink is critical in digital textile printing. Without white ink, it is not possible to print pre-assembled dark garments such as T-shirts; in the case of roll-to-roll fabric printing, a large percentage of inks and printing time may be wasted to print the background color with a small percentage of image features. Despite its necessity, the development of white ink for digital textile printing has been slow and difficult. Two general approaches are discharge and hiding. The discharge approach is to destroy the pi conjugation system of the visible light-

232

Digital printing of textiles

absorbing dye molecules so that it will not absorb visible light any more. The hiding approach is to put a light-scattering layer on top of the substrate so that the light will not reach the dark substrate to be absorbed. In the case of discharge, the discharging agent has to match the chemistry of the dye molecules so the selection of dyed fabric or T-shirts is limited and it is rare for the discharging to be complete. In the case of hiding, inorganic or organic pigments can be used to scatter light. The inorganic pigment can be any metal oxides or salts such as titanium dioxide with high refractive index. Organic pigments can be solid bead with relatively high refractive index or with a hollow core. The metal oxides or salts, especially titanium dioxide, may be more effective, but tend to be high density, so settling is a big issue in low viscosity ink jet inks. Organic pigments may not have settling issue or abrasion to wear the print head, but could be less effective. To obtain enough whiteness, pre-treatment may be required or discharge and hiding may be combined. No ideal solution has yet been found.

13.7 Sources of further information and advice For general background on inkjet, Inkjet Technology and Product Development Strategies, written by Stephen F. Pond and published by Torrey Pines Research, is recommended. It covers print head design, ink formulation, media, interactions among these three key components of inkjet and the overall system integration. It also offers some information about the market and the existing and new applications of inkjet technology. For basic knowledge on textile testing, the Technical Manual of the American Association of Textile Chemists and Colorists is a standard book. For the most current events and developments, the following websites should be surfed frequently: · · · · · · ·

IT Strategies, www.it-strategies.com Lyra Research, www.lyra.com Web Consulting, www.web-eu.com www.tc2.com FESPA, Federation of European Screenprinters Associations, www.fespa.com ISS Imprinted Sportsware Show, www.issshows.com ITME, International Textile Machinery Exhibition

14 Formulation of aqueous inkjet ink H N O G U C H I and K S H I R O T A , Canon Inc., Japan

14.1 Dye±fiber interaction There are several types of inkjet inks for textile printing, depending on kinds of fabric and dyes. In each combination various dyes and textiles have been developed and improved to satisfy advancing practical requirements. Table 14.1 shows a classification established in the textile industry on the fabric±dye set, from which a primary choice of dye is made. The principles of dyeing of polymer are exemplified in Fig. 14.1, in which the inter-molecular interaction and chemical bonding between dye molecule and polymer in fiber are shown. Acid and direct dyes are bound by ionic bond, reactive dyes on the cotton are bound by covalent bond, disperse dyes are bound by Van der Waals forces and hydrogen bonding. The printing processes give favourable chemical and physical conditions for these bindings. Process conditions promote adsorption

14.1 Inter-molecular interactions and chemical bonds.1

Table 14.1 Types of fabric and dye2 Cotton

Rayon

Direct dye Bat dye Sulfur dye Naphthol dye

l l n n

l s n s

Reactive dye for cotton Acid dye Metal chelated acid dye

l

l

Chrome dye Reactive dye for wool Disperse dye Cationic dye l good; n fair; s applicable; © special use.

Wool

Silk

Nylon

n s

s

s l l

n l l

l l

l l

s s s

s ©

Acetate

Triacetate

Polyester

Vinylon

Acryl

n n n

l

l s

l

l ©

© ©

s l

Formulation of aqueous inkjet ink

235

Table 14.2 Formulation of dye-based aqueous ink Dye Water-soluble organic solvent 1 Water-soluble organic solvent 2 Surfactant Acid or base to adjust pH Water Additives (see Table 14.5)

0.3±10 wt.% 5±15 5±15 0.1±2.0

on the fiber surface, diffusion into the fiber, and dissolving in the fiber polymer or formation of chemical bond. In the case of inkjet printing on textiles these post-processes, including instruments for fixing, can be applied. The formulation of the ink composed of materials for water-based systems are shown in Table 14.2. The main differences between desktop inkjet printers and textile printers are kinds of dye, their concentrations, and physical properties. The reliability and printing characteristics of the formulation are not inherent properties but depend on print head and process architecture. Ink formulation and properties must be tested and tuned on each printing system.

14.2 Organic solvents and surface energy of ink The surface energy of ink has very important effects on various working properties. Absorption speed influences printing speed, i.e., the scanning speed of the print head and travelling speed on production. The water-soluble organic solvent and surface-active agent determine the absorption speed of inks. Table 14.3 shows the kinds of solvent used in aqueous inks and how organic solvent affects the surface energy of ink. By choosing water-soluble organic solvent and its concentration it is possible to reduce vaporization of water and to tune properties of ink such as viscosity and surface tension. Table 14.4 shows how surface energy of ink affects various aspects of inkjet printing and has major effects on reliability and print performance.

14.3 Time-dependent phenomena and surface-active components The most important factor in ink formulation is to optimize the ink properties to the print head. Dynamic behavior of ink is important in creating the objective size of droplets, in stabilizing the speed, and in reducing satellites. Theoretically, drop formation from a liquid stream in air is controlled by two equations:3

236

Digital printing of textiles

Table 14.3 Surface energy of solution of soluble organic solvent and water Chemical name

(mN/m) Chemical name

Ethylene glycol Diethyleneglycol Triethyleneglycol Polyethyleneglycol 300 Thiodiglycol Hexyleneglycol N-methyl-2-pyrrolidone 2-Pyrrolidone

-Buthyrolactone 1,3-Dimethyl-2-imidazolidinone Sulforan Dimethylsulfoxide Trimethylol propane Trimethylol ethane Neopentyl glycol Ethylene glycol monomethyl ether Ethylene glycol monoethyl ether Ethylene glycol monobutyl ether Ethylene glycol monophenyl ether Diethyleneglycol monoethyl ether Diethyleneglycol monobutyl ether Diethyleneglycol dimethyl ether Diethyleneglycol diethyl ether Triethyleneglycol monomethyl ether Triethyleneglycol monoethyl ether Tripropyleneglycol monomethyl ether

54.7 53.8 52.6 52.4 47.2 40.3 53.1 48.5 51.8 51.3 43.1 46.8 49.9 53.7 43.6 57.1 49.0 30.2 41.8 51.9 34.7 51.0 42.9 52.4 44.0 44.0

Diacetonealcohol 49.6 N,N-bis-hydroxyethyl urea 56.6 Urea 53.4 Acetonyl acetone 36.9 1,2-Cyclohexane diol 35.7 1,4-Cyclohexane diol 52.2 3-Methyl-1,5-pentane diol 41.3 3-Hexene-2,5-diol 37.9 Glycerine 57.3 Glycerin monoallyl ether 46.3 Glycerin triacetate (triacetine) 38.1 Water 73.7 Ethanol 47.8 2-Propanol 39.6 1-Methoxy-2-propanol 41.4 1-Butanol 26.5 Furfuryl alcohol 48.8 Tetrahydrofurfurylalcohol 53.5 1,2-Butane diol 49.9 1,3-Butane diol 48.0 1,4-Butane diol 42.8 2,3-Butane diol 38.4 1,5-Pentane diol 38.7 Triethyleneglycol monobutyl ether 33.9 Dipropyleneglycol monomethyl ether 45.4

Ratio of mixture is water : solvent ˆ 90 :10 by weight.

Table 14.4 Effects of surface energy of ink on printing process System

Ink properties

Printing and image

(mN/m)

Wetting to ink container Linear flow in capillary Frequency response Drop and satellite formation Foaming and defoaming Wetting to nozzle surface Start-up property Surface tension Solubility of colorants Stability of dispersion Weber number and Reynolds number Dissolving of impurities from container Wetting width on printing substrate Penetration speed and depth in substrate Mottle (tone homogeneity) Dot gain Color value

Formulation of aqueous inkjet ink

237

Weber Number: We ˆ U 2 D= Reynolds Number: Re ˆ UD= where

 U D  

ˆ ˆ ˆ ˆ ˆ

specific gravity (kg/m3), flight speed of liquid (m/s), diameter of flight liquid (m), surface tension (N/m), dynamic viscosity number (m2/s).

The inkjet nozzle requires inking properties so that this number is within a certain range to create an objective drop. In this regard it is important to measure dynamic surface tension and viscosity for a few microseconds. Figure 14.2(a), (b) shows dynamic surface tension of solvent and water mixtures. We can tune the dynamic and static behavior of formulated ink by selecting an organic solvent. The rheological behavior of ink at the microsecond scale relates to drop formation from the ink stream ejected from nozzles. Dot placement, spreading, and penetration are phenomena of a few milliseconds. Surface energy in this time range has important effects on image formation. Inks placed on the fabrics spread and are absorbed into pores in fibers, and then go inside the fibers. The dynamic surface energy of ink affects both drop formation and image formation. By measuring the static and dynamic surface tension of ink we can tune absorption speed to printing speed. Fabric surfaces vary in nature. Natural fabrics show a hydrophilic nature and synthetic fibers have a hydrophobic nature. Inks for hydrophilic fabrics have a surface-active solvent or surfactant added. The reason is increased wetting speed.

14.4 Additives Surface-active components in ink influence capillary flow and drop formation, homogeneous coloration, wiping properties, and reduction of foaming by decreasing the interfacial tension of dissolved air. Other factors affecting the choice of the kind and concentration of solvent and surfactants are solubility and stability of ink. In Table 14.5 additives commonly used are listed. They are for finishing a textile product as ingredients in pre-treatment, inkjet inks, and/or post-treatment. The choice of additive also relates to pre-treatment and posttreatment of each production process.

14.5 Reliability Reliability has various aspects. Start-up is the principal issue in on-demand water-based inkjet systems. By evaporation of water, the concentration of solid components increases near the nozzle surface. This results in an instantaneous barrier for uniform ejection of the ink stream and for drop formation. The barrier reduces transformation efficiency of energy from ejection pressure to ejection

238

Digital printing of textiles

14.2 Dynamic surface tension of (a) organic solvent in water; (b) surfinol AE100 in water (acetylene glycol derivative type non-ionic surfactant).

speed and size of droplet. This is called start-up failure. Figure 14.3 shows the evaporation speed of solvents from a mixture. As shown in Fig. 14.3, evaporation of water is slower by incorporating of water-soluble organic solvents having two or more hydroxyl group or a hydrophilic nature. Start-up failure causes the image edges and ends of the substrate to lose color and sharpness. Cleanliness around the surface of the ejection nozzle also affects steady ejection, so the ink is rendered self-cleaning by redissolving the residual portion of the ink on the nozzle surface. Stability of

Formulation of aqueous inkjet ink

239

Table 14.5 Ink additives Surfactant Anti-oxidizing agent IR absorber Anti-flammable agent Perfumes Antiseptics

UV absorber Hygroscopic agent Fluorescent whitening agent Fastness enhancer Fixing accelerator Disinfectant

dye from oxidative, reductive, and hydrolytic breakdown generates unfavorable species on shifting physical properties of ink. Physical shift of ink causes print failure such as clogging, precipitation, pH shift and interaction between printhead materials. Increase of low volatile organic solvents usually give better reliability, although there are some deficiencies in this method. Image bleeding on HT fixing, color stain on washout, mottling after drying, and diffusion transfer on the process after printing may occur, so optimum choice and concentration becomes important.

Abbreviations: ETOH: ethanol; IPA: isopropyl alcohol; 2-BuOH: 2-butanol; H2O: water, EG: ethyleneglycol; DEG: diethylene glycol; TEG: triethylene glycol; TDG: thiodiglycol; GLY: glycerine; NPG: neopentyl glycol; TMP: trimethylol propane; TME: trimethylol ethane; PENTA: pentaerythritol; 126HT: 1,2,6-hexane triol; 125PT: 1,2,5-pentane triol. Condition for measurements: weighting method 25ëC, 60% RH, no air flow.

14.3 Evaporation speed of solvents from 1:1 mixture of water and organic solvents.

240

Digital printing of textiles

14.6 Production process of inkjet-printed textiles Figure 14.4 shows a typical production process of inkjet textile prints. In the process, pre-treatment and post-printing relate strongly to ink properties. Usually pre-treatment material and loading weight by padding or coating on fabrics must be determined carefully. The pre-treatment materials are hydrophilic, and thus sensitive to moisture content in storage environments. Therefore in storage of treated fabrics, humidity control becomes important. The conditioning of pre-treated fabrics before printing is also important in attaining target image quality. Ink formulation is tuned to adopt these conditioned fabrics on spreading and absorption. If moisture content becomes lower, the saturation and color density curve show reductions. Post-treatment after printing is also related to formulation of ink. If the water-soluble organic solvent is excessively concentrated image bleed will occur during the fixing stage.

14.4 Production process of printed textiles by inkjet.

14.7 Reactive dye ink 14.7.1 Printing with reactive dye ink Printing process conditions with reactive dye ink for cotton are exemplified in Table 14.6. The pre-treatment solution commonly contains water-soluble polymer, base (sodium carbonate), and salts (sodium alginate), and urea. All the

Formulation of aqueous inkjet ink

241

Table 14.6 Reactive dye ink process on cotton fabrics Process

Material

Instrument

Pre-treatment Sodium alginate Urea Sodium carbonate Padding Drying Tenter Printing Dye  0.3 mg/cm2 Inkjet printer Steaming HT steamer Washing Wash cool water Soaping 1 g/L nonionic surfactant Washing Hot water Washing Cool water

Condition 100±200 g/kg 100 g/kg 30 g/kg Pick up ratio 70±80% 100±120ëC At saturation density* 102ëC 8min saturated steam 100ëC 3min 60±80ëC

* Dye load weight depends on kind of fabric and condition of treatment.

compounds in the pre-treatment solution should be selected from compounds that do not react with the reactive dye. Amine and hydroxyl group moiety can react with reactive dye. From this point water-soluble organic solvents may possibly react with reactive dyes. This is one of the causes of low yield of printed ink and coloration of wash water along with the printing process. The role of sodium carbonate is as a catalyst to promote hydrolysis by increasing pH. The role of sodium alginate and urea is homogeneous dyeing and keeping moisture in the fabric. There are many alternatives in each component. Ink droplets are laid down evenly by the film-formed hydrophilic nature of pretreatment materials and the fiber surface. For promoting rapid wetting of ink to the pre-treated surface, inks often contain surface-active compounds.

14.7.2 Reactive dye and reaction mechanism In Table 14.7 commercial reactive dyes are listed. They are the latest dyes for textile printing, including inkjet printing. Historically, various reactive dyes Table 14.7 Reactive dye for textile printing Products by Nippon Kayaku Co.

Products by Ciba Specialty Chemicals, Inc.

KAYACION Yellow P-5G liquid 33 KAYACION Orange P-G liquid 20 KAYACION Red P-BN liquid 33 KAYACION Red P-4BN liquid 25 KAYACION Blue P-3R liquid 40 KAYACION Turquoise P-3GF liquid 33 KAYACION Black P-NBR liquid 40

CIBACRON Yellow MI-100 CIBACRON Golden Yellow MI-200 CIBACRON Orange MI-300 CIBACRON Red MI-400 CIBACRON Red MI-500 CIBACRON Blue MI-600 CIBACRON Turquoise MI-700 CIBACRON Gray MI-800 CIBACRON Black MI-900

242

Digital printing of textiles

Table 14.8 Reactive moieties in reactive dyes Type of reactive moiety DCT: VS: MFT: DCQ: CFT: MCT: TCP: MNT: MCT/MCT: VS/VS: MCT/VS: MFT/VS:

Reaction temperature

Dip Vat Printing dyeing dyeing

dichloro triazine Low vinylsulfone Medium monofluoro triazone Medium dichloroquinoxaline Medium monochlorofloro triazine Medium monochloro triazine High trichloro pyrimizine High nicotinic acid triazine High polyfunctional (uni moiety) High polyfunctional (uni moiety) Medium polyfunctional (same moiety) Medium to high polyfunctional (same moiety) Medium

l l l l l l l l l l l l

s l s

n

l

l

l n l

l good; n fair; s applicable.

have been created by attaching reactive moieties to acid and direct dyes. They are summarized in Table 14.8. Of the 12 groups in Table 14.8 the vinylsulfone (VS) type and the monochloro triazine (MCT) type are the majority of the market. Furthermore MCT is the most popular in inkjet inks. Reactive dyes undergo hydrolysis at the initial stage of the reaction, therefore to get long-term stability of formulated ink it is very important to control the pH range and to choose compounds having acidic and basic nature. In the course of development, the kind of acidic or basic compounds or buffer must be determined by careful experiments. The reactive dye generates species such as Na‡Clÿ, SO42‡ by hydrolysis. The washout process is important in removing inorganic and free organic compounds to give the right quality.

14.7.3 Stability of reactive dye ink Figure 14.5 shows possible reactions during storage and in printing and fixing on cellulose. Reactive dyes can react not only with the fiber nucleophile (cellulosate anion) but also with nucleophiles, such as amino moiety and hydroxyl, present in ink or pre-treatment ingredients. From this mechanism there are several unfavorable courses of reaction by reactive dyes. This reaction reduces the efficiency of coloration of a polymer by a dye±fiber reaction (fixation) and results in dye wastage. Therefore some part of the dye is inevitably lost in the post-treatment stages. If the wash-off process is incomplete, the product suffers from reduced wash fastness. A more efficient material and condition for reactive dye has long been sought. The inkjet process has not changed the situation in this regard.

Formulation of aqueous inkjet ink

243

14.5 (a) Reaction of vinylsulfone type reactive dye.

14.7.4 Dye load and coloration The yield of dye on dyeing is limited by the unreacted part of the dye as explained in Section 14.7.3. For this reason the concentration of reactive dye decline is relatively high compared to desktop printer ink. There is also another reason in the high surface area of fabrics. Figure 14.6 shows the dye load on printing versus K/Smax. K/Smax represents the K/S value at the reflection maximum of the final fabric, where K/S is the ratio of absorption coefficient to scattering coefficient defined in Kubelka±Munk theory. In Fig. 14.6 experimental results by disperse dye are also plotted and show the lower dye load weight and higher efficiency on the K/S tone curve. The slower saturation in the tone curve mainly comes from dye loss by hydrolysis. The total amount of ink

244

Digital printing of textiles

14.5 (continued) (b) Reaction of monochlorotriazine type reactive dye.

Formulation of aqueous inkjet ink

245

14.6 Dye load on printing versus K=Smax of printed fabrics (K=S values are at reflection maximum of measured textile samples).

on the fabric can be increased by a multi-pass method, but in such a case speed and resolution are sacrificed. For many years, the lower yield in the reactive dye has remained unsolved in textile printing.4

14.8 Disperse dye ink 14.8.1 Printing with disperse dye ink Printing process conditions with disperse dye ink for polyester are exemplified in Table 14.9. The pre-treatment materials give homogeneous wetting to the polyester and ink-receiving layer, giving good image quality and smooth transport to the fiber. Pre-treatment ingredients are selected and formulated from these objectives. Polyester is hydrophobic in nature, so the pre-treatment with hydrophilic polymer has important effects on print quality.

14.8.2 Kind of disperse dye for inkjet Commercial dyes for coloration of polyester fabrics are exemplified in Table 14.10. For inkjet printing by disperse dye, there are two methods: (i) direct

246

Digital printing of textiles

Table 14.9 Disperse dye ink process on polyester fabrics Process

Material

Instrument

Pre-treatment Sodium alginate Citric acid Padding Pick up ratio Drying On tenter Printing InkJet printer Steaming HT steamer Wash off Cool water Soaping 1 g/L nonionic surfactant Soaping 1±2 g/L hydrosulfite 2 g/L causticsoda 1 g/L surfactant Wash Hot water Wash Cool water

Condition 100±200 g/kg 0±2 g/kg 70±80% 100±120ëC Dye 0.1 mg/cm2 170±180ëC 8min saturated steam 40ëC 60±90ëC

40±60ëC

printing, and (ii) dye sublimation transfer printing. For sublimation transfer, relatively lower molecular weight and more hydrophobic disperse dyes have been adapted. The structural group of the molecule is the same as that of the D2T2 dye for photo print. The sublimation transfer method is composed of inkjet printing onto a transfer sheet and transfer to fabrics by a heated press. It needs no other post-treatments such as steaming and chemical fixing. Because of its simplicity, the system has been adapted to small production runs in the consumer market. The dispersion technology is the same as for direct printing with disperse dye.

14.8.3 Ink formulation for direct printing Disperse dyes are prepared as water-borne dispersions by a similar process of pigment dispersion as for inkjet. Inks with disperse dye differ greatly from screen-printing inks in terms of range of particle size, kind of dispersing agent and final viscosity.5 The role of polymers in the textile paste for the screen ink is moved to the pre-treatment polymer. The dispersed colorants are usually prepared as concentrations (10±20 wt%) of dispersion in water and are highly purified. The range of mean particle size is roughly from 100 to 250 nm (in a distribution from 20 to 400 nm) by eliminating coarse particles including polymer emulsion. The viscosity range of the concentrated dispersion is from 10 to 50 mPa.s. By using low-viscosity, highly particulated dispersions, inks for polyesters are formulated. Table 14.11 shows the preparation processes of disperse dye ink. The purification processes to get objective levels on storage and ejection stability are important in inkjet disperse inks. For performing the dispersion process the choice of dispersing agent is significant. Table 14.12 lists the system

Table 14.10 Commercial disperse dyes for polyester fabrics For transfer printing

For direct printing Products by Nippon Kayaku

Products by Ciba Specialty Chemicals Inc.

Products by Ciba Specialty Chemicals Inc.

KAYALON Polyester Yellow 4GN KAYALON Polyester Orange R-SF 200 KAYALON Polyester Light Red B-S 200 KAYALON Polyester Pink RCL(N) KAYALON Polyester Blue 2R-SF KAYALON Polyester Turquoise Blue GL-S C200 KAYALON Polyester Black BRN-SF paste 100

TERASIL Yellow DI-100 TERASIL Red DI-200/01 TERASIL Red DI-300/01 TERASIL Blue DI-400 TERASIL Blue DI-500 TERASIL Gray DI-600 TERASIL Black DI-700 TERASIL Violet DI-800

TERASIL Yellow TI-1000 TERASIL Orange TI-2000 TERASIL Red TI-3100 TERASIL Red TI-3200 TERASIL Turquoise TI-4000 TERASIL Blue TI-5500 TERASIL Black TI-7000

248

Digital printing of textiles

Table 14.11 Preparation process of disperse dye ink Purification of disperse dye Dispersion Purification

Addition of additives Ink preparation Aging Purification Bottling

Elimination of salts and organic impurities Roll-milling, bead-milling, high-pressure milling 1 Coarse particle by ultra-centrifugation, ultra-filtration 2 Inorganic impurities by reverse osmosis 3 Solublilized part of disperse dye

By ultra-centrifugation

requirements and physico-chemical properties of the dispersion. The requirements from the system and the physico-chemical properties of the ink denote the design criteria for dispersion. The technical criteria for designing a dispersed dye are (i) dispersability on ink preparation and (ii) reduction cleaning on washoff. These two are frequently incompatible because the former is stabilization and the latter is easy removal. Dye concentration in ink is determined by making a standard curve to correlate K/S against the weight of printed ink under system conditions. The general tendency is exemplified in Fig. 14.6. Part of the dye is not absorbed in the fiber but is washed out at a reduction cleaning stage under high pH Table 14.12 Dispersion design and dispersing agent Dispersion design System requirements Physico-chemical properties Dispersing agent 1 Fine particulation

Wetting to active surface of dye particle

1 Poly-sulfonic acid and its derivative

2 Storage stability

Adhesion to dye particle

2 Naphthalene sulfonic acid and its derivative

3 Ejection stability

Low and linear viscosity

3 Lignine sulfonic acid and its derivative

4 Smooth diffusion of particle to fabrics

Low viscosity and good wetting to ink channel

4 Low molecular weight carboxylic acrylate polymer

5 Wash-off property

Good solubility to water

5 Low molecular weight aqueous polyester

6 Stain in white

Good solubility to wash-off cleaning liquid

6 Aqueous polyurethane

7 Keep touch in fabric

Easy removal from fabric

8 Keep air permeability Easy removal from fabric

Formulation of aqueous inkjet ink

249

conditions. The relation depends on all the printing and washing conditions. The concentration of dye in each ink is determined by preparing such a relation as in Fig. 14.6. The yield of dye is better than that of reactive dye ink, because of physical diffusion of disperse dye into synthetic fibers.

14.8.4 Flow rheological property The preparation of the aqueous dispersion has significant effects on reliability and print performance. How to reduce impurities and how to tune the dye towards the targeted particle size and its distribution are important technical issues, especially in terms of cost. These properties have effects on static storage stability by precipitation, and long-term ejection stability by aggregation, and frequency response through high-speed refilling flow by non-linear viscosity. All these aspects associated with disperse dye, in most cases, present larger difficulties than pigment dispersion in inkjet. The work of Konica-Minolta should be referred to on these points.6 The series of dispersed dye inks were prepared and examined in terms of storage and ejection performance and nonlinear behaviour of viscosity. Figure 14.7 shows three types of shear rate versus viscosity curve by disperse dye inks. Ink 1 shows a lowering of viscosity by increasing shear rate. The tendency in Ink 1 is analyzed by imperfect adsorption of dispersant. Unadsorbed dispersant is a cause of agglomeration and larger particles. When ink contains

14.7 Viscoelasticity of three disperse dye inks. Instrument: MCR300 Modular Rheometer (Physica Messtechnik GmbH).

250

Digital printing of textiles

Ink 1 Ink 2 Ink 3

Dispersant/carrier medium

Auxiliary agent

Optimized for each dye Same for all dyes Same for all dyes

Not added Not added Added

14.8 Sedimentation properties of three disperse dye inks.

these particles, viscosity behavior tends to curve 1. The curve for Ink 2 shows a large decrease in viscosity by increasing the shear rate. The cause of the large non-linearity in Ink 2 has been explained as structural viscosity by flocculation. Figure 14.8 shows the results of an experiment relating absorption by sedimentation to centrifugal force applied. From Fig. 14.8 Ink 1 contained larger particles, but Inks 1 and 3 did not contain components of larger particles. This means that the causes of non-linear viscosity of Inks 1 and 2 in Fig. 14.7 are different. The color components in Ink 2 for the black shade were three disperse dyes. Further analysis of the surface energy reveals that if the wetting property of one dye component is weak, the particles tend to aggregate or agglomerate. Ink 3 shows linear viscosity and hence enough jettability and long-term stability. The inks having non-linear viscosity showed a short-term threat of head clogging. These research results in Figs 14.7 and 14.8 suggest that (i) linear viscosity behavior is favorable to inkjet ink, and (ii) wetting of dye and combination of hydrophilic±hydrophobic balance of dispersing agent is important in solving these failures.

14.9 Acid and direct dye ink formulation Dyeing of nylon/silk/wool using acid or direct dye-based ink is the most popular way in inkjet textile printing. Table 14.13 shows the latest commercial acid and

Formulation of aqueous inkjet ink

251

Table 14.13 Acid and direct dyes for textile printing Products by Nippon Kayaku

Products by Ciba Specialty Chemicals Inc.

KAYASET Yellow N5G KAYASET Red B KAYASET Red 130 KAYASET Blue F-R KAYASET Blue A-2R KAYASET Black 922(D)

LANASET Yellow SI-100 LANASET Orange SI-150 LANASET Red SI-200 LANASET Red SI-330 LANASET Blue SI-400 LANASET Turquoise SI-500 LANASET Gray SI-600 LANASET Black SI-710

direct dyes for textile printing. On choosing a dye, compared to desktop inkjet printer inks, dyes having lower solubility in water have been adopted for textile printing inks, especially for apparel use. Fiber polymers with repeating moieties, such as amide, peptide and urethane, have a hydrophilic nature and work an ionic attractive force upon acid/direct dyes. Because of these properties, it is not always necessary to apply pre-treatment and fixing agent. Although the washfastness of dyed products must satisfy the criteria for quality standards in apparel products, these interactions are sometimes not enough to give high washfastness of products. From this tendency the dyes for printing have relatively lower solubility in water. In formulating ink, solubility and long-term stability are key for reliability issues. Commonly, water-soluble organic solvents having high solubilizing power also exhibit effects on compounds of the ink container and nozzle channel materials in the print head. Careful tests on solubility of dye in certain vehicle solvents under both elevated and lowered temperatures are planned in the manufacturing stage. In the finishing stage fixing agents are applied to increase wash and humidity fastness. The post-treatments give improvements but, in some cases, have disadvantages in other aspects of usability. To overcome these disadvantages, pre-treatment is also another option.

14.10 References 1. Yoshiki Akatani technical brochure, Fixing of Dye on Various Textile and Inkjet Printing, Nippon Kayaku Co. (2005). 2. Nagase Sangyo Co., in the text Textile Printing Seminar (1999). 3. Akira Asai et al., `Impact of an ink drop on paper', IS&T NIP7 Proceedings (1991). 4. Burkinshaw, D.K., Dyes and Pigments, 33, 11, 1887. 5. Piriya Putthimai et al., `Comparison of textile print quality between inkjet and screen printing', Surface Coating International Part B, B1, 88, 1-82, March 2005. 6. Yasuhiko Kawashima et al., `The development of new disperse dye inks for inkjet textile printing', IS&T NIP19 Proceedings (2003).

15 Effect of pretreatment on print quality and its measurement Y K K I M , University of Massachusetts-Dartmouth, USA

15.1 Introduction A printed textile fabric can be produced by various methods. In the US, rotary screen (69%), flat screen (20%), transfer (10%) and roller (1%) printing are traditionally employed for the manufacture of printed fabrics. Current textile printing speeds are 27 to 55 m (30 to 60 yards) per minute at a 1.5 m (60 inch) printing width. The number of colors in a design can be up to 24. However, typical textile print designs have an eight-color pattern. Also, average lot sizes are 2700 linear meters or longer.1 US and EU textile markets are facing challenges from developing economies elsewhere around the globe. The US Textile/Apparel Complex has adopted a demand-activated manufacturing architecture (DAMA) to survive into the twenty-first century. This manufacturing strategy requires `quick response', time-based competition, small lot sizes (typically 450 m or less), large variety, and linkage into the supply chain.2 It has been recognized that the textile printer can achieve these goals by employing digital printing technology. This provides minimized printing lead time, quick sample printing and small lot size. Current digital textile printing, based on inkjet technology, is mainly limited to proofing and sample production. Further developments in ink formulation, jet design, pre/post wet processing requirements and higher throughput will provide the solution for extending digital printing technology beyond the mere preparation of proofing samples. Hopefully, the extension of digital printing into a full production mode will enable the developing textile printing market to thrive in this century. From the standpoint of traditional textile printing, emerging digital printing is too slow and expensive compared to the mass production characteristics of rotary screen printing. Since the late 1980s customers have demanded short runs, high quality and customization. Digital printing technology supports this customer demand together with digital linkage to the supply chain.3 Rotary screen printing is the most popular method (close to 70%) of traditional textile printing. There are many advantages of rotary screen printing:

Effect of pretreatment on print quality and its measurement

253

high-speed production, economical long runs, and a large color gamut from the spot color. Typically rotary screen printers are integrated with dryers and other finishing equipment to provide a continuous printed fabric production ± print and dry, or print, steam, wash and dry. The resulting printed cloth has excellent fastness to light, crock (dry and wet), and wash. Although rotary screen printing offers many benefits, there are also many drawbacks such as low machine efficiency due to pattern changes and correcting printing problems (40% downtime). Also short print runs are not easily accommodated by rotary screen printing. Traditional analog printing involves a lengthy and expensive sampling process in addition to the machine efficiency problem. The design is converted into screen files and screens are engraved. Once the screens are ready, colors are matched and patterns are `struck-off' on the print machine. With an average strike-off time of 5±6 hours and screen engraving turnaround of two to three days, the total time from design origination to finished product can be several weeks.4 Digital textile printing can offer solutions for these run-size and time problems. Digital printing reduces the turnaround time between design origination and finished product. Digital printers do not require a lengthy setup time between patterns and can print continuously. In addition, digital printing also provides the elimination of screen cost in sampling and offers profitable shortrun production. Printing without screens eliminates the registration problems and more importantly provides mass customization.5 Nevertheless, there are disadvantages associated with digital printing of textiles. It is an emerging new technology and the available inkjet printer models are equipped with inkjet heads typically targeted toward consumer home/office applications. This means that the print heads were designed to be disposable and would not meet the durability demanded in a production environment. Here, there was a lack of proper ink chemistry designed for reliability. Special coatings or lamination were required to achieve proper print quality. Furthermore, original digital printers provide only CMYK process color, which severely limits the volume of color gamut. However, the most advanced wideformat textile digital printers in mid-2004 come equipped with six to eight process color inkjet heads. Digital textile printing arguably broke the initial technical entry barrier due to recent advances in high speed inkjet printers, specialized inks, and color management software. It can displace screen printing in short, medium and even some high volume production runs on the basis of quality, cost and speed.6 Due to the low viscosity of inkjet printing inks, textile print media require a pretreatment for improved print quality and post-treatment of the fabric to allow for ink fixation. Print quality, in general, is affected by printing hardware/ software (see Parts I and II), substrate (Chapter 12) and ink (Chapters 13 and 14). The present chapter addresses print quality issues related only to substrate and pretreatment: available pretreatments of textile substrates for inkjet printing,

254

Digital printing of textiles

print quality characterization, and pretreatment effects on the print quality of digital textile prints.

15.2 Textile pretreatments for inkjet printing Traditional screen printing delivers printed fabrics with brilliant colors, sharp and clear patterns, and soft fabric handle. However, it is very difficult to achieve these desirable qualities with inkjet technology. This is due to the different ink delivery methods employed in the traditional method compared to inkjet printing. The viscosity of inkjet inks is very low (up to 5 mPa
s) in order to meet available print head technologies, while traditional printing pastes (inks) have viscosity close to 5000 mPa
s. The printing paste prevents the dissolved dyes from running on the fabrics in the analog printing.7 There are a wide range of inkjet ink types for substrates constructed from different fiber types. The most common type is water-based inks, which are as low in viscosity as water. Thus, textile fabrics and nonwoven webs require preparatory treatment to achieve acceptable print quality. The type of textile pretreatment depends on fiber contents of fabrics, accordingly on inks based on dyes or pigments, which have substantivity to the fabric's fiber contents.

15.2.1 Matching fiber material and ink chemistry Woven printing substrates are the most important in textile printing. The volumes of knits and nonwovens used for printing are quickly increasing at the expense of wovens. Printing clothes usually are constructed by weaving or knitting. Woven printing fabrics are produced on a shuttle-less weaving machine by interlacing two sets of yarns, i.e. warp and weft threads, while knitted printing fabrics are formed by intermeshing yarn loops on a circular knitting frame or a warp knitting machine. Typical width of the fabrics is 1.5 m (60 inches). The cotton substrates are the most widely printed (48%), and others consist of cotton/polyester blends (19%), polyester (15%), and viscose (13%). Fabrics made from nylon, acrylics, wool and silk play a minor role in textile printing.8 Inks for inkjet printing are formulated from the same colorants used in traditional dyeing and printing, but dyes and pigments used in inkjet inks require high purity, submicron particle size, and high tinctorial strength for reliable print head operation and print quality. In addition, the ink's physical and chemical characteristics must be compatible with the fiber chemistry. Cellulosic fibers are the most frequently used fiber type for print fabrics. Cotton, linen, viscose, and polynosic fibers all consist of cellulosic polymers. Dye classes for inkjet ink formulation suitable for cellulosic fiber printing fabrics are direct dyes, vat dyes and reactive dyes. While direct and vat dyes are relatively large in molecular size and fixed by physical forces, the formation of covalent bonds between reactive dyes and cellulosic fiber makes reactive dyes

Effect of pretreatment on print quality and its measurement

255

the dye of choice. They are of small molecular size and good solubility. Reactive dyes provide a full gamut of colors, are brighter and faster diffusing, and the excess is easily removed in the washing-off process.9 Polyester fibers are the most commonly used fibers among synthetic fibers, because of their desirable properties and low cost. Azo, anthraquinone, coumarin, and quinoline disperse dyes are used after reducing the particle size under 1 m by milling in the presence of a dispersing agent. Proper fixation and post-treatment provide excellent wet-fastness properties. Polyester/cotton blends have become increasingly popular in recent years. Regenerated cellulose fibers, e.g. viscose, are also mixed with polyester fibers. The popularity of the polyester/cellulose fiber blends stems from the optimum balance of physical properties and wearer comfort. In traditional screen printing, several techniques and dyes and dye combinations were tried with good results, but with difficulties. Ink formulation can be divided into two classes: a single class of colorant and two classes of dye. The former includes pigment, insoluble azo colors, selected vat dyes and selected disperse dyes. Inks formulated from pigments with binder are simple to apply and suitable for fiber blends. Other single dyes show good fastness, but available hues are limited. The latter case uses disperse dye for polyester and reactive dye for cotton in analog printing. However, a pigment and disperse dye combination in polymeric binder will work for inkjet inks.9 Polyamide fibers, which include nylon 66, nylon 6 and nylon 11, were the first synthetic fibers produced in commercially significant quantities. Warp knitted nylon fabrics are used for printing. Dyes are selected from the ranges of acid, metal complex acid and direct dyes according to physical and chemical requirements of inks.9 Protein fibers are used in higher priced luxury goods due to their desirable appearance and properties far exceeding those of synthetic polymeric fibers. Thus the protein fiber substrates are prime candidates for high value added, short run print production. They are wool, cashmere, and silk. In theory, acid, basic, and direct dyes are suitable for use as protein fiber inks. Ink formulation is limited, however, to acid dyes. Acid dyes are selected because of high colorfulness, acceptable light and wet fastness. For added cost, reactive dyes can be used in inks in higher durability articles made of protein fibers. In general, all these ink systems require different dye classes for different fiber substrates, and typically require post-steaming, washing and drying to achieve acceptable properties in the printed fabric. There are diverse opinions on post-treatment required in inkjet printing. The different opinions arise primarily due to the varying concepts of where in the textile fabric manufacturing process inkjet printing will actually be performed. Those who believe that inkjet printing will replace current printing technologies in the textile industry feel that postprocessing will be readily acceptable. Those who feel that inkjet printing will be practiced by design studios, apparel manufacturers and retail establishments argue for little or no post-processing.10 Many researchers reported that pre-

256

Digital printing of textiles

treatment and/or ink formulation can eliminate post-treatment requirement in textile inkjet printing. Further information on ink formulation is found in Chapters 13 and 14. Pretreatment of textile substrates in inkjet printing is required to optimize the interaction between the low viscosity ink drops jetted from the print head and capillaries in fibers, yarns and fabric structures. In the next section, we will discuss ink±textile substrate interaction in the context of understanding the needs of pretreatment technology and materials used.

15.2.2 Ink±textile substrate interaction The choice of textile substrate influences image quality (i.e., inter-color bleed, dot quality, color, visual perception, etc.), ink drying time, and fastness (light, wet, gas, etc.). Fabric substrates are three-dimensional structures, and low viscosity inks can wick into macro-capillaries between yarns and fibers. Inks can also diffuse into the micro-capillaries in fibers. The wicking and diffusion rates are controlled by the surface tension of ink, ink viscosity, yarn and fabric structures, and the polymer morphology of the fiber. Ultimately, dye molecules in the ink droplets must be fixed on or near the surface of the textile fiber substrate for sharp and brilliant color images. The fixing mechanism depends on the dye/fiber combination as discussed in the previous section. An ink drop jetted onto a fabric substrate wets the surface, then spreads. The wetting and spread are controlled by the surface tension and viscosity of ink, and this initial ink±media interaction ultimately determines the dot gain. The subsequent interactions in the form of wicking (capillary flows of the dye fluid in fibers and yarns) and diffusion determine dot quality, line quality, inter-color bleeding, and mottle. Finally, the solvent (i.e. water) in water-based ink systems is lost by evaporation and absorption. Dot gain comes from physical dilation and the optical Yule-Nielson effect.11 The optical dot gain is a measure of the extent to which the area measured by optical density is larger than the real area covered by ink. This is the consequence of the light scattering in the substrate due to light not emerging from the point where it entered, as pointed out by Yule-Nielson. Physical dot gain measures the larger dot size by ink±substrate interaction including spreading. Inkjet printed dot models are shown in Fig. 15.1. Linear dot gains are defined with parameters in Fig. 15.1 as
Dx …%† ˆ 100  …Dx ÿ D0 †=D0 ;
Dy …%† ˆ 100  …Dy ÿ D0 †=D0

Dot area gain is calculated from linear dot gains as below: " 2 #  D ÿ1  100
A …%† ˆ D0

…15:1†

…15:2†

Effect of pretreatment on print quality and its measurement

257

15.1 Dot gain models: (a) ideal model, and (b) dot with a satellite.
x and
y are printer resolution, which is inverse of printer addressability.

 ˆ …Dx ‡ Dy †=2. where D The interaction between ink and textile media also influences the line quality. In addition to line width gain, the fibrous substrate with capillary and micropores generates edge raggedness, unsharpness and feathering. According to ISO 13360, edge raggedness is expressed as the standard deviation of differences, di , between actual contour and the best-fit straight line as shown in Fig. 15.2. Inter-color bleeding is most in the invasion of one color into an adjacent area. It stems from the slow rate of ink penetration into the substrate and precipitation of colorant in the second ink caused by additives in the first ink. This can be minimized by pretreatment and ink set selection. Image noises in the printed area are expressed as graininess for high spatial frequency and mottle. Graininess is a measure of non-uniformity in optical density of spatial scale smaller than 250 m, while mottle is that of larger than 250 m. Interaction between ink and media generates spatial image noises, i.e. graininess (fine scale) and mottle (coarse scale). Textile substrates have inherently high surface roughness and texture, which causes mottle noises. The uneven density is caused by ink drops coalescing at the surface and low porosity of the substrates. To reduce this tendency, the textile substrate must be pretreated with high ink absorbing porous particles and/or polymers, which serve as surface finishes on the fabric. These finishes condition the fabric surface to accept printing inks. We have established the fact that the image quality of digital print depends on substrate surface preparation in the form of chemical and physical pretreatment.

15.2 Edge raggedness, defined as the standard deviation of the differences di (ISO/IEC 13660-2002).

258

Digital printing of textiles

The optimal pretreatment process is selected to match the ink system and the fiber content of the substrate.

15.2.3 Pretreatments of textile substrates for inkjet printing In traditional screen printing, ink pastes contain thickening agent and additives required for fixing chosen colorants. For instance, spot color pastes based on reactive dyes contain alkali (e.g. sodium bicarbonate), urea, and sodium alginate as a thickener. Due to poor resistance to alkali and some additives of inkjet print heads, inkjet inks must be formulated with sub-micrometer size purified dyes and very few additives compatible with print head technologies. Thus all other auxiliaries including thickener are used in pretreatment formulations for target textile substrates. Last decade, we saw rapid development in textile substrate pretreatments for inkjet printing to meet the growing demands of diverse applications. Textile fabric materials are porous, soft and pliable. In order to achieve wellregistered, clear and sharp printing results using water-like printing inks without thickening agents in digital printing, the textile substrates need to be pretreated. The bleeding on untreated fabric in digital printing is another issue that should be taken very seriously. In design studio, retail or end user applications, postprinting washing is not preferred. Thus pretreatments intended for the applications must not only control the wicking and diffusion of ink for higher print quality, but also protect colorant integrity after drying. Patent literature shows that a large number of patents describing pretreatments for inkjet printing of cotton fabrics with reactive dye inks were issued in past decades.12,13 This is understandable from the fact that approximately 50% of worldwide prints are produced with cotton fabrics. Taniguchi et al.14 first pretreated polyamide fabrics with bisphenol derivatives, then inkjet printed the pretreated fabrics followed by post-printing washing and clearing off of unfixed dye. The printed fabric showed better color depth, no staining and good wash fastness. Hees et al.15 disclosed a pretreatment formula (LupraJet HD) for inkjet printing of polyester and polyester/cotton blends. The disclosed pretreatment works well with inks based on pigments or disperse dyes. Readers who need further information on pretreatment are referred to Chapter 12. We will discuss the effects of pretreatment on print quality in the next section.

15.3 Effect of pretreatments on print quality Print quality problems in digital printing of textiles can be categorized into four issues: (1) appearance-related issues including line definition, text quality, resolution, image noise, optical density, tone reproduction and (to a lesser

Effect of pretreatment on print quality and its measurement

259

extent) gloss; (2) color-related issues including color gamut, color matching and color registration; (3) permanence issues including light fastness and water fastness; and (4) usability issues including the presence of defects and `hand'. Many print quality issues, not surprisingly, are common to both conventional and digital printing techniques. However, inkjet printing introduces a number of peculiarities of its own, for example jaggies (digital artifacts in edges), banding (lines of missing color), and satellites (extra drops of ink). Clearly, for digital printing of textiles to advance, significant improvements in print quality must be achieved.16

15.3.1 Print quality measurement Print quality is defined as how closely the printed dot resembles that intended on an individual and/or collective basis, while image quality is the closeness of the final printed image to that intended.17 There are many attributes determining perceived image quality: artifactual (unsharpness, graininess, digital artifacts, redeye, etc.), preferential (color balance, contrast, saturation, memory color reproduction, etc.), aesthetic (lighting quality, composition), and personal (preserving a cherished memory, conveying a subject's essence). The last two types of attributes are not quantifiable and not directly related to print quality.18 Over the past few decades, there has been a virtual explosion in printing and copying technologies, particularly digital printing technologies. These developments have forced many people in the printing/printer business to ask some basic questions about how to specify and assess print quality (PQ) in a consistent way that is not dependent on the specific printing technology they are using. Aside from color measurement equipment (spectrophotometers) and densitometers, there have been relatively few objective means of evaluating PQ of lines, text, print uniformity, registration, etc. In response to this situation, the International Organization for Standardization (ISO) has developed international PQ standards, ISO/IEC 13660 2001(E): Information Technology ± Office Equipment ± Measurement of image attributes for hard copy output ± binary monochrome text and graphic images. Several PQ equipment manufacturers have implemented the ISO standards in their automated PQ equipment.19 The standards are applicable to digital textile printing, even though their intended area of application is paper media digital printing. Digital print quality attributes derived from ISO 13660 are summarized in Table 15.1.

15.3.2 Pretreatment effects on print quality Print quality improvements in digitally printed textiles after pretreatments have been reported elsewhere.7,12±15,20,21 Sapchookul and coworkers20 evaluated effects of pigment to binder (P/B) ratio on inkjet ink properties (surface tension, viscosity, shear stress and rate) as well as effects of P/B ratio and pigment

260

Digital printing of textiles Table 15.1 Print quality attributes Image element

Quality attribute

Dot

· Dot location · Dot gain · Dot shape · Edge raggedness · Satellites

Line

· Line width · Edge sharpness · Edge raggedness · Optical density · Resolution (modulation)

Solid area

· Optical density (tone reproduction) · Color (lightness, chroma, hue, gamut) · Noise (graininess, mottle, background, ghosting)

dispersion techniques on print quality (color gamut and color gamut volume, air permeability, stiffness, and crock fastness). They found that a P/B ratio of 1:2 for all inks gives an optimum color gamut, color gamut volume, and crock fastness. Fan and coworkers investigated the effect of pretreatments on digital textile prints. They treated woven and knitted cotton fabrics with various pretreatment formulations containing alginate, silicone-based textile softener and nano-silica powder. Digital image analysis and optical microscopy were used to compare various print quality parameters of pretreated fabrics.21 In the next sections, effects of fabric structures, preparation and pretreatment on print quality (i.e. line width and color gamut) are discussed from results of the author and coworkers' previous research work.16,21

15.3.3 A case study of cotton fabric prepared with a special pretreatment For the case study, we acquired and characterized test fabrics. The fabric structural details are shown in Table 15.2. Test Fabrics, Inc. (Pittston, PA) prepared the cotton fabrics by bleaching and/or mercerization (caustic soda treatment). Print quality of un-pretreated fabrics The main objective of the print quality study with un-pretreated fabrics was to explore the relationship between fabric properties and print quality. The results were used as a baseline for a print quality investigation with specially pretreated cotton fabrics. The un-pretreated fabrics listed in Table 15.2 were printed with an Epson Stylus Color 1520 desktop inkjet printer and the OEM ink set supplied

Effect of pretreatment on print quality and its measurement

261

Table 15.2 Description of cotton fabrics studied* Style

Treatment

Thread count

Yarn size type

Weight (g/m2)

423 twill

Mercerized

108  72

258

428 sateen

Bleached

96  56

437 knit 407 poplin

Bleached Mercerized

38  44 100  50

400M print

Mercerized

80  76

419 broad

Mercerized

132  72

14/1  14/1 carded 20/1  14/1 carded 30/1 combed 20/1  17/1 combed 40/1  32/1 carded 40/1  40/1 combed

235 124 189 107 120

* Reprinted with permission of IS&T: The Society for Imaging Science and Technology, sole copyright owners of The Journal of Imaging Science andTechnology.21

with the printer. This printer and its ink set are not specifically designed for textile printing. Line quality analysis Our results show that different fabric properties affect line quality quite differently. Of the properties studied, one of the most significant is fabric structure. The results of our structure comparisons are shown in Fig. 15.3. As the figure shows, the plain weave fabrics have the highest line width gain, followed by the twill and sateen woven fabrics. The knitted fabric has the lowest gain. However, in the case of the knitted fabric, another important factor may come into play, namely the hydrophobic character of the fabric as demonstrated by wicking tests. The results of these tests are shown in Fig. 15.4. Here, the average line width gain is plotted against the water/alcohol wicking ratio, which is a good indicator of the hydrophilic/hydrophobic nature of the material. From these data, it is clear that the knitted fabric is hydrophobic, whereas the other fabrics are hydrophilic. The correlation suggests that in addition to the effects of structure shown in Fig. 15.3, the hydrophilic/hydrophobic nature of the fabric (or the finish on the fabric) strongly influences the ink±fabric interaction. Graininess (image noise) analysis The effect of fabric structure on graininess (image noise) was noticeable, and the fabric variable with the greatest impact on image graininess was found to be yarn size. The results are shown in Fig. 15.5. Yarn type was also considered, but was found to have no significant impact on graininess. Generally, as gray level

262

Digital printing of textiles

15.3 Effect of fabric structures on line width gain. Samples were prepared with the fabrics as received. Reprinted with permission of IS&T: The Society for Imaging Science and Technology, sole copyright owners of IS&T NIP14 Conference Proceedings.16

15.4 Correlation between average line width gain and water/alcohol wicking ratio. Reprinted with permission of IS&T: The Society for Imaging Science and Technology, sole copyright owners of IS&T NIP14 Conference Proceedings.16

Effect of pretreatment on print quality and its measurement

263

15.5 Effect of fabric structures on graininess. Reprinted with permission of IS&T: The Society for Imaging Science and Technology, sole copyright owners of IS&T NIP14 Conference Proceedings.16

increases from 10% to 100%, graininess decreases. This coincides with the fact that image noise is sensitive to brightness variation. In other words, noise is most noticeable in the highlight and mid-tones regions; it is affected mostly by the size of the yarn and to a lesser degree by the fabric structure. Color gamut and color accuracy We were surprised to find that the color appearance of all samples tested was quite similar. Quantitatively, the color gamuts of all the samples were about equal. However, two observations, illustrated in Figs 15.6 and 15.7, are worth mentioning. Figure 15.6 compares the color gamuts of the two fabrics, style 419 and 407. Both of the fabrics consist of mercerized combed yarns, but their sizes are 40 and 20 cotton count, respectively. It appears that the color gamut of print on the fabric (style 407) made of the larger size yarns is larger than that of the finer size yarns. The difference in color coverage area of a*±b* plot (numerical color gamut) is 15.5%. Secondly, although the numerical color gamuts for the plain weave (style 400M) sample and the knitted sample are very close, there is an apparent downshift in the a*±b* plane for the knitted sample, indicating a color shift between the two types of fabric structures (see Fig, 15.7). This apparent shift toward bluer shade stems from the structural difference and/or bleaching treatment of the knit fabric, which gives bluer substrate.

264

Digital printing of textiles

15.6 Effect of yarn size on color gamut. Reprinted with permission of IS&T: The Society for Imaging Science and Technology, sole copyright owners of IS&T NIP14 Conference Proceedings.16

15.7 Effect of fabric structure on color gamut. Reprinted with permission of IS&T: The Society for Imaging Science and Technology, sole copyright owners of IS&T NIP14 Conference Proceedings.16

Effect of pretreatment on print quality and its measurement

265

Print quality of pretreated fabrics The cotton fabrics described in Table 15.2 were used in this study. All were woven fabrics except style 437 which is a cotton `T-shirt' knit. The absorbency of cotton fabrics was evaluated according to AATCC Test Method 79: Fabric Absorbency Measurement. As shown in Table 15.3, the absorbency characteristics of these fabrics indicate that surface preparation of the fabrics is different. Notable is the observation that the Knit (437) and the Sateen (428) have a greater affinity for the 2-octanol than they do for water. This shows that these fabrics have hydrophobic surfaces, especially the knit fabric. The fabric wicking properties were determined by INDA IST 10.0-70 Method 10.3: Fabric Wicking Behavior Measurement. The results are shown in Table 15.4. The data indicate that a slight wicking anisotropy can exist relative to the warp and weft direction in some fabric types. Here a P/F ratio of 1.00 denotes a uniform wicking behavior of the fabric in the warp and the weft directions. As shown, most of the woven cotton fabrics evaluated wicked fluid at a faster rate in the warp direction than in the weft. The W/O ratio (average value of water and 2-octanol warp and weft wicking rate) is an index of the relative ability of the fabrics to wick hydrophobic versus hydrophilic fluids. With the exception of the Sateen (428) and to a much greater degree the knitted (437) fabric, most of the fabrics studied had a hydrophilic character. This is indicative of the scouring, mercerizing and bleaching treatments given to these fabrics. It is suspected that the knit fabric has some lubricating finish on the yarns as a processing aid. Pretreatment of cotton fabrics for digital printing Some preliminary digital printing studies were conducted on all the fabric samples described in Table 15.2. The results indicate that the cotton knit (437) and the plain weave print (400M) fabrics were the fabrics deemed suitable for our further digital printing experiments. Overall, this was expected since twill, Table 15.3 Water and 2-octanol absorbency* Fabric

Absorption time (s) Water

423 twill 428 sateen 437 knit 407 poplin 400M print 419 broad

2.6 20.4 >1200 5.8 18.6 11.3

2-octanol 2.1 1.8 1.0 10.6 14.4 16.6

* Reprinted with permission of IS&T:The Society for Imaging Science andTechnology, sole copyright owners of The Journal of Imaging Science andTechnology.21

266

Digital printing of textiles

Table 15.4 Water (W) and 2-octanol (O) wicking rate Fabric

Fluid

Warp (P) cm

Weft (F) cm

P/F

W/O

423 twill

W O

5.87  0.3 4.70  0.1

5.38  0.2 3.97  0.1

1.09 1.18

1.30

428 sateen

W O

3.37  0.1 4.07  0.1

3.60  0.2 4.03  0.2

0.94 1.01

0.86

437 knit

W O

0.20  0.0 5.80  0.0

0.05  0.0 5.40  0.1

4.00 1.07

0.03

407 poplin

W O

5.58  0.1 4.17  0.1

4.97  0.1 3.80  0.1

1.12 1.10

1.32

400M print

W O

4.73  0.2 4.40  0.2

3.80  0.2 3.73  0.2

1.24 1.18

1.05

419 broad

W O

5.75  0.1 4.32  0.1

4.70  0.2 3.70  0.1

1.22 1.17

1.32

* Reprinted with permission of IS&T: The Society for Imaging Science and Technology, sole copyright owners of The Journal of Imaging Science andTechnology.21

sateen, poplin and broad fabric styles have inherently coarse surface textures. Furthermore, these fabric styles are less used in textile printing. Moreover, the selection of cotton fabrics, 400M print and 437 knit is based on the absorbency and wicking behavior shown in Tables 15.3 and 15.4. Therefore, fabric 437 cotton knit and the 400M cotton print were the only two fabrics selected for the detailed pretreatment studies herein reported. The selected fabrics were treated with the recipes listed in Table 15.5. The alginate (Prime Alginate T-400) from Table 15.5 Pretreatment recipes Treatment 1 2 3 4 5 6 7 8 9 10 11 12

Alginate (g)

Silicone (g)

Silica (g)

Water (ml)

4.00 2.00 0 0 0 0 4.00 4.00 2.00 4.00 2.00 2.00

0 0 4.00 2.00 0 0 4.00 2.00 3.00 3.00 2.00 2.00

0 0 0 0 2.00 1.00 1.00 2.00 2.00 2.00 1.00 2.00

196 198 196 198 198 199 191 192 193 191 195 194

Reprinted with permission of IS&T: The Society for Imaging Science and Technology, sole copyright owners of The Journal of Imaging Science andTechnology.21

Effect of pretreatment on print quality and its measurement

267

Multi-Kem Co. and the silicone (Ultratex CSP) from Ciba Specialty Chemicals Co. were used for this study. The nano-silica used was from Nissan Chemical Industries (MEK-ST). Inkjet printing of pretreated fabrics The digital textile printing was carried out with an Epson Stylus Color 980 printer. OEM-supplied CMYK inks were used as received. A special pattern was designed to facilitate the evaluation of basic printing parameters after printing. The pattern contains primary colors, red, blue, yellow, green and brown, and lines with different width from 0.5 point to 4.5 points. Epson Photo Paper was used to print the same pattern for print quality reference. The pretreated fabrics were paper-backed for easy transport through the media path of the inkjet printer. After printing, the print qualities were analyzed using an optical microscope and a digital image analysis to quantify the print qualities in terms of color-related metrics (L*a*b*) and line quality. Line quality analysis Line dilation data for all 13 pretreatment samples were reported.20 For the plain cotton weave and knit fabrics studied, a pretreatment containing 2% alginate, 1.5% silicone softener and 1% silica shows good balance in colorant retention and line width control. In addition to this pretreatment formula, line qualities of 1% alginate only (traditional printing paste thickener) and photopaper were included in Figs 15.8 and 15.9. Pretreatment containing only alginate has higher line width gains for both woven and knit structures. Nano-silica-containing pretreatment provides line quality close to photopaper quality. Results show that digital textile printing quality on plain weave and knitted cotton fabrics is influenced by the fabric pretreatments, the most noticeable being the line width. The line quality on these pretreated cotton fabrics was not significantly affected by the fabric structure and the hydrophilicity of the fabric surfaces as long as the pretreatment can give cotton fabrics a balanced hydrophilic/hydrophobic character. We will expand future studies to quantify line quality parameters such as blurriness, raggedness and optical density. Color depth and color gamut The visual quality of the textile prints on pretreated fabrics was quite good, considering that neither the printer nor the ink set was optimized for printing on fabrics. The fineness and sharpness of detail, the fineness of line, and the saturation and the quality of the color were all quite acceptable. Where color is concerned, the CIE L*a*b* color system is used in many applications. L* is the lightness, a* is the red±green aspect and b* is the yellow±

268

Digital printing of textiles

15.8 Effect of pretreatment on line quality (Woven 400M fabric).

15.9 Effect of pretreatment on line width (Knit 437).

Effect of pretreatment on print quality and its measurement

269

blue aspect of a specific color. The higher the L* value the lighter the color, the higher the a* value the redder the color, and the higher the b* value the more yellow the color. It is observed in Fig. 15.10 that the pretreated woven fabric with 2% alginate, 1.5% silicone and 1% silica (shown as Nano silica) pretreatment shows good results in color depth for all five color shades used. The depths of five colors printed on nano silica treated woven 400M were all very close to that of photopaper. On knit fabric (Fig. 15.11), the combination of 2% alginate, 2% (or 1.5%) silicone and 1% silica gave the better color depth in yellow and red shades, when compared to the photopaper. From Figs 15.10 and 15.11, we can conclude that the pretreatment recipe containing 2% alginate, 1.5 to 2% silicone and 1% silica was good in achieving dark shades for all five colors (red, yellow, blue, brown and green) on both woven and knit fabrics. That means this combination would render color strength on the pretreated cotton fabrics. Color gamut plots for selected pretreatments are shown in Figs 15.12 and 15.13. In general, the color appearance of all prepared samples was quite similar. Quantitatively, the color gamuts of all the samples were about equal except for photopaper. This is understandable due to fact that the printer manufacturer optimized ink for its photopaper supplied. However, two observations illustrated in Figs 15.12 and 15.13 are worth mentioning. It appears that the color gamut for the knit fabric (style 437) is as large as that for the photopaper as shown in Table 15.6. Secondly, the numerical color gamuts for the knit samples treated with 1%

15.10 Effects of pretreatments on the depth of shades (Woven 400M).

270

Digital printing of textiles

15.11 Effects of pretreatments on the depth of shades (Knit 437).

15.12 Effect of pretreatment on color gamut (Woven 400M, mercerized 100% cotton).

Effect of pretreatment on print quality and its measurement

271

15.13 Effect of pretreatment on color gamut (Knit 437, bleached only).

silicon and nano silica are 15.7% larger than those of the woven samples. However, there is an apparent expansion leftward (greener) and downward (bluer) in the a*±b* plane for the knitted sample, indicating a colorfulness differential between the two types of fabric structures. Figure 15.12 also shows that all pretreated woven fabrics, 400M, were slightly greener (lower a* values) compared to the photopaper except for the green color itself which showed no significant difference between the pretreated fabrics as well as between the pretreated fabrics and the photopaper. This green effect could be from the fabric itself or from the pretreatments resulting from the unnoticeable color. In Fig. 15.13, none of the pretreated knit fabrics showed excessive a* value differences among them, which is similar to the results for their woven counterparts. However, it is believed that the pretreatments may be the cause of the color variation, because the colors on the pretreated knit fabric, 437 knit, also showed a greener tone. Overall, Figs 15.12 and 15.13 show that Table 15.6 Numerical gamut comparison (Woven 400M versus Knit 437) Pretreatment

Woven 400M Numerical gamut (arbitrary unit)

Knit 437 Numerical gamut (arbitrary unit)

Photopaper Nano silica Alginate

4537.59 3917.83 3512.62

4544.88 4534.11 4155.88

272

Digital printing of textiles

the color gamut was affected only slightly by the pretreatments. The b* values had similar trends among pretreatments except that the photopaper had a little bluer tone. Based on the experimental results, a pretreatment recipe of 2% alginate, 1.5% silicone softener and 1% silica is recommended for the digital printing of woven and knitted cotton fabrics.

15.4 Concluding remarks and future trends In this chapter, we reviewed the effects of the pretreatment on print quality of digitally printed textiles. Several digital print quality characteristics were measured on a variety of pretreated cotton fabrics. The main observations in this study can be summarized as follows: 1.

2.

3.

4.

5.

The subjective visual quality of inkjet printed cotton fabrics, either as received or pretreated, was as good as that printed on plain paper or photopaper. Several important print quality attributes including line quality, image noise, optical density and color quality were measured. The results clearly show that these objective print quality metrics are needed to evaluate the efficacy of pretreatments, and to investigate the substrate/pretreatment interaction for advancing digital textile printing technology. The effects on print quality of several key fabric properties were studied. These include fabric structure, yarn size, yarn type and pretreatment. The test results suggest that the most significant fabric variables are fabric structure, yarn size and the hydrophilic/hydrophobic nature of the fabric (i.e. absorbency). Fabric absorbency and fabric wicking behavior were measured. The studied knit and sateen fabrics were found to have the most hydrophobic surfaces. Woven fabrics showed some anisotropic wicking behavior. Effects of fabric pretreatments on the quality of digital printing on plain cotton weave and a knit fabric were studied in detail. A special print pattern for quality analysis was designed for use in this study. In the experimental work, a number of recipes derived from the different combinations of alginate, silicone softener and fine silica particles were used as pretreatments for the cotton fabrics. Overall this study showed that digital textile print quality is influenced by fabric pretreatments, the most noticeable being the appearance-related quality, i.e., the line width. The print quality was not significantly affected by fabric structure, and the hydrophilicity of the unpretreated (as received) fabrics. Pretreatments can give cotton fabrics the required characteristics of digital printing substrates, i.e., the balanced hydrophilic/hydrophobic characteristics. It was found that digitally printed cotton fabric with the optimum pretreatment can have as good a quality as that of the digitally printed

Effect of pretreatment on print quality and its measurement

273

photopaper substrate under the conditions pertaining to this case study. This indicates that quality digital printing onto textile fabrics is achievable with proper fabric pretreatment and inkjet technology. Once an optimum solution for fabric pretreatment is developed, digital textile printing can be used not only for preparing prototype samples but also for production quantity and quality printed textile fabrics. While digital printing technology is a fairly mature technology for the paper printing industry, such developed technology cannot be directly applicable to the textile printing industry. This is because there are significant differences in the physical and chemical nature of various fabric substrates as well as the process variations depending on target markets. Printers currently being used for textiles were originally graphics printers and did not address all of the needs of the industry in terms of width, speed and substrate handling. Several companies have begun addressing these problems, and the future of digital printing of textiles is beginning to take shape. International Dyer magazine reported that Reggiani, an Italian textile inkjet printer manufacturer, sells a machine with a printing speed of 120 m2/hr. Other wide format textile inkjet printer manufacturers (DuPont, Mimaki, Mutoh, Cannon, etc.) also market high speed six to seven head machines.22 In spite of this advancement, there remain many future challenges and barriers for this exciting new technology to enter the traditional textile screen printing markets. We will briefly discuss these issues by which print quality is determined, and conclude the chapter. First, printer system design addresses textile-specific problems such as variations from paper printing in colorant volume loading, spot colors versus process colors, substrate texture, dimensional instability and chemistry of substrate. Other crucial system parameters to be optimized for textile printing include ink viscosity, gray levels, drop size, drop frequency, number of print heads, and data communication that will impact on the resulting print speed and print resolution. Color reproduction and management are the next major challenges for both the paper printing and the textile printing processes to achieve higher colorrelated print quality. In particular, this challenge is more severe for textile printing if digital printing technology is employed to substitute the table-strikeoff process. The use of process color and inks with physical±chemical properties which differ from the analog printing process will add to color matching difficulties. In a similar fashion to color reproduction, digital prints must emulate the engraving appearance of the analog production prints, if it is adopted for the table-strike-off process. Print quality attributes in terms of dots, lines, large areas including tonal gradations and print growth must have reasonable resemblance between the two processes. In this respect, the printers and the engravers must

274

Digital printing of textiles

work closely together to provide the digital printing system with the necessary knowledge of analog print processes and techniques. The physical and chemical properties of the ink must be designed to fit the substrate characteristic in order to achieve an acceptable level of various color fastness properties such as crocking, washing, and light fastness. In addition, the proper treatments of the fabric before and after the printing process are essential to ensure good fastness properties. Commercial-scale digital printing requires sophisticated fabric pretreatment and post-treatment. Proper pretreatment promotes satisfactory appearance and color-related print quality, while proper post-treatment allows fashionable effects and provides additional values such as improved handle, luster, color fastness or soil resist properties to the fabric. On the contrary, inconsistent or poor treatment will result in color and print mark variations. These two valueadding steps also present a problem to users who do not have such facilities. In addition, any inconsistent fabric speed, fabric wrinkling, loose fibers and the touching of the inkjet nozzle due to improper pretreatment will present a great challenge to the print quality result. Textile printers prefer inks that are compatible with colorants used in the analog printing process to facilitate coordination and to maintain similar color effects. As shown in Section 15.2.1 of this chapter, inks for digital printing are available in reactive dye containing inks for cellulose, acid dye inks for protein fibers, disperse dye inks for polyester, and pigment inks for all types of fibers. However, due to the print head operating restrictions, some inks may not perform satisfactorily in terms of fastness and/or color gamut requirements. Other challenges for the growth of digital textile printing sector are printing speed, print width, ink and other costs. Indeed, we can end this chapter by quoting Brooks G. Tippett: `The future of textile printing will be digital, though no one can predict when we will see it unfold.'4

15.5 References 1. Eastwood, B. and Malachowski, R., Cranston Print Works Co., Cranston, RI, Private Communication, June 1998. 2. Tincher, W., Cook, F., Carr, W. and Failor, B., `Keynote paper: Printing on textile substrate', pp. 368±369, IS&T 46th Annual Conference (1993). 3. Randal, D.L., `Digital imaging for textiles ± next generation', http:// www.techexchange.com/thelibrary/digitalimagingNG.html. 4. Tippett, B.G., `Recent developments in digital textile printing inkjet production becomes a reality', pp. 372±378, Proc. AATCC Ann. Int. Conf. and Exhibition, 21± 24 October 2001. 5. Poetz, T., `Inkjet printing: Present situation and prospects', International Textile Bulletin, Vol. 48, Issue 5, pp. 80±83, October 2002.

Effect of pretreatment on print quality and its measurement

275

6. Pearlstine, K., `Understanding digital textile inks', International Dyer, pp. 27±29, June 2004. 7. Hees, U., Freche, M., Kluge, M., Provost, J. and Weiser, J., `Inkjet printing: New product for pretreatment of textiles', International Textile Bulletin, Vol. 49, Issue 2, pp. 64±66, 2003. 8. Ross, T., A Primer in Digital Textile Printing, www.techexchange.com/thelibrary/ DTP.html, May 2001. 9. Gutjahr, H. and Koch, R.R., `Direct print coloration', Chapter 5 in L.W.C. Miles (ed.), Textile Printing, 2nd edn, Society of Dyers and Colorists (1994). 10. Tincher, W.C., Qiang Hu, Xiaofei Li, Yingnan Tian and Jianming Zeng, `Coloration systems for ink jet printing of textiles', Proc. 14th Int. Congr. on Advances in Non Impact Printing Technologies, p. 243 (1998). 11. Emmel, P., `Physical models for color prediction', Chapter 3 in G. Sharma (ed.), Digital Color Imaging Handbook, p. 223, CRC Press, Boca Raton, FL (2003). 12. Taniguchi, M., JP 2004-143621 (2004) and JP 2004-162247 (2004). 13. Takikawa, S., Kishi, H. and Takaishi, N., JP 2004115953 (2004). 14. Taniguchi, T., Nakamura, T. and Tanaka, K., JP 2004-131919 (2004). 15. Hees, U., Kluge, M., Freche, M., Freyberg, D., Siemensmeyer, K., Heissler, H. and Raulfs, F.-W., DE 10244998 (2004). 16. Tse, M.-K., Briggs, J.C., Kim, Y.K. and Lewis, A.F., `Measuring Print Quality of Digitally Printed Textiles', Proc. IS&T NIP14 Int. Conf. on Digital Printing, pp. 250±256, Springfield, VA (1998). 17. Hurd, A.L., Inkjet Academy: Theory of Ink Jet Technology, an IMI Digital Printing Summer Course, Cambridge, MA, 19±20 July 2004. 18. Keelan, B.W., Handbook of Image Quality, Marcel Dekker, New York (2002). 19. Quality Engineering Associates, Inc., Personal Image Analysis System User's Guide, Burlington, MA (2002). 20. Sapchookul, L., Shirota, K., Noguchi, H. and Kiatkamjornwong, S, `Preparation of pigmented inkjet inks and their characterisation regarding print quality of pretreated cotton fabric', Surface Coatings International, Part A: Coatings Journal, Vol. 86 (A10), pp. 403±410 (2003). 21. Fan, Q., Kim, Y.K., Perruzzi, M.K. and Lewis, A.F., `Fabric pretreatment and digital textile print quality', Journal of Imaging Science and Technology, Vol, 47, Issue 5, pp. 400±407 (2003). 22. Scrimshaw, J., `Advances of digital dream', International Dyer, pp. 14±16, June 2004.

16 Ink jet printing of cationized cotton with reactive inks P J H A U S E R , North Carolina State University, USA and M K A N I K , University of Uludag, Turkey

16.1 Introduction In recent years, ink jet printing has found an increasing application in the printing of textiles. It has demonstrated considerable benefits in terms of strikeoff, sampling and more recently in the production of short lots of printed textiles.1,2 It is also expected that with further advancements in software, printer and ink technologies being made every day, ink jet printing of textiles will become ever more important in the near future.3 Reactive dye based inks are used commonly to print cotton and other cellulosic fibers. In ink jet printing, unlike conventional reactive printing, thickener, alkali and urea are applied onto the fabric by a padding process prior to application of color, and this process has a crucial importance on the print quality.4±8 After ink jet printing, steam fixation, washing and drying processes follow. In general, the reactive inks used in ink jet printing often have a degree of fixation to cotton of only 70%.9 In order to achieve the necessary high level of wetfastness, the unfixed dye must be removed effectively. For that reason, timeconsuming, energy intensive and expensive washing-off procedures are required similar to the conventional washing-off processes used with fiber reactive dyeing. This washing-off process has a major negative environmental impact owing to the large amount of dye and chemicals removed and the large amounts of water required. Furthermore, unfixed reactive dyes in the wastewater may pose an environmental hazard.12 This problem can be minimized by increasing the fixation rate of reactive inks as high as possible, and by reducing the amount of chemicals used in the pretreatment process. For this purpose, either reactive dyes9 or fabric10,11 can be modified, or fixation-enhancing chemicals can be used.4 The chemical modification of cotton prior to dyeing and printing in order to improve its dyeability with anionic dyes such as reactive, direct, acid, sulphur and vat dyes has received considerable attention in recent years. All of these modifications introduced cationic groups in the form of quaternary, tertiary or secondary amino residues. In this way, anionic reactive dyes are attracted by the

Ink jet printing of cationized cotton with reactive inks

277

cationic charges on the fiber, and as a result a high degree of dye±fiber fixation, a reduced washing off procedure, reduced or no electrolyte use in dyeing, and wetfastness properties equivalent to the untreated cotton can be obtained.12 Some recent literature has reported that near 100% dye fixation rates could be obtained in conventional reactive dye printing, as well as with acid and direct dye printing by cationization of cotton with 2,3-epoxypropyltrimethylammonium chloride12,13 prior to printing. The actual cationizing agent used in this work was also 2,3-epoxypropyltrimethylammonium chloride (I).3 This reactive material is conveniently prepared in situ by the reaction of 3-chloro-2-hydroxypropyltrimethyl-ammonium chloride (II) with alkali according to Fig. 16.1. 3-Chloro-2-hydroxypropyltrimethylammonium chloride is commercially available as a 69% solution in water and was used as received. 2,3-Epoxypropyltrimethylammonium chloride will react with alcohols under alkaline conditions to form ethers (Fig. 16.2). The reaction product of this epoxy with cotton is a modified fiber with the structure shown in Fig. 16.3. A by-product of reactions of 2,3-epoxypropyltrimethylammonium chloride in aqueous alkaline solutions is 2,3-dihydroxypropyltrimethylammonium chloride (III), a species that cannot react with cotton at the conditions used here (Fig. 16.4). Depending on the specific reaction conditions, 20±50% of the epoxy groups will hydrolyze to form this inactive material. As a result of reaction with 2,3-epoxypropyltrimethylammonium chloride, cotton will have covalently bound cationic dye sites that are present at any pH value. When dyeing or printing, these cationic dye sites will strongly attract negatively charged anionic reactive dyes to form an ionic bond. Furthermore, due to the enhanced electrostatic attraction, the reactive dye molecules reach the hydroxyl groups in the fiber faster, and the covalent bonding rate is accelerated.12

16.1

16.2

278

Digital printing of textiles

16.3

16.4

This chapter will review two recent studies14,15 that investigated potential improvements realized by cationizing cotton fabrics prior to ink jet printing with reactive inks.

16.2 Experimental 16.2.1 Materials A scoured, bleached and mercerized, optical brightener-free woven (twill 3:1) 100% cotton fabric, with a weight of 235 g/m2, a density of 46 threads/cm in the warp and 20 threads/cm in the weft directions was used throughout both studies. As the cationizing agent, a 69% solution in water of 3-chloro-2-hydroxypropyltrimethylammonium chloride was used. The inks were Cibacron Yellow MI-100, Cibacron Red MI-500 (magenta), Cibacron Turquoise MI-700 (cyan) and Cibacron Black MI-900. A medium viscosity alginate thickener was used in the pretreatment pad bath. A naphthalene sulphonate based anionic scouring agent was used for washing of the prints. Other chemicals in this study were commercial grade sodium hydroxide (50% w/w), soda ash, urea and acetic acid (97%) used as received.

16.2.2 Cationization The cationization was carried out according to the pad-batch method,14 on a Mathis HVF padder. The cationic reagent was used at concentrations of 50, 75, 100 and

Ink jet printing of cationized cotton with reactive inks

279

125 g/l with corresponding sodium hydroxide concentrations of 31.0, 46.5, 62.0 and 77.5 g/l, respectively. Sodium hydroxide concentrations were calculated according to the method optimized by Tabba.16 The fabrics were padded through the cationization baths at 100% wet pickup, wrapped in plastic, and stored at room temperature for 24 hours. After removal from the plastic, the batched fabrics were rinsed with warm water at 40ëC, neutralized with 2 g/l acetic acid at 40ëC, cold rinsed, and then dried on a conveyor-type drying machine at 100ëC.

16.2.3 Pretreatment process for ink jet printing Prior to printing, the untreated and the cationized fabric samples were padded on a Mathis HVF padder with soda ash (10, 20, 30, or 40 g/l), urea (100 g/l) and thickener (0, 4, 8, or 12 g/l). The wet pickups were adjusted to 70%, and all fabrics were dried at 100ëC.

16.2.4 Printing processes The printing processes were performed on a Stork Amber ink jet printer with a pass number of 4 and resolutions of 360 and 720 dpi. All prints were air dried, and steamed on an Arioli sample steamer at 100ëC.

16.2.5 Washing procedures Throughout the study, the untreated printed samples were washed according to the following five-step washing procedure: (1) 5 min cold rinsing at 20ëC; (2) 5 min warm washing at 40ëC; (3) 10 min hot washing at 95ëC by the addition of 3 g/l scouring agent; (4) 5 min warm washing and neutralization at 40ëC by the addition of 0.5 g/l acetic acid; and (5) 5 min cold rinsing at 20ëC. Cationic cotton samples were washed either according to the same procedure or with the following three-step washing procedure: (1) 5 min warm washing at 30ëC by the addition of 3 g/l scouring agent; (2) 10 min hot washing at 70ëC; and (3) 5 min cold rinsing and neutralization at 20ëC by the addition of 0.5 g/l acetic acid. The liquor ratio was 30:1 for all washing processes. The washed samples were dried on the conveyor dryer at 100ëC.

16.2.6 Testing methods The color properties of the printed samples were determined using a Datacolor SF300 spectrophotometer with SLI-Form/NG software (SheLyn, Inc.). Relative color strength, staining on white ground and dye penetration were determined using the Kubelka±Munk equations; (K/S)f, (K/S)g and (K/S)r respectively.17 For dye penetration, (K/S)r values were measured on the reverse side of the fabrics. All K/S values were taken at the max of each dyestuff.

280

Digital printing of textiles

Color fastness to washing was tested according to AATCC Test Method 61 (2A) and color fastness to crocking was tested according to AATCC Crockmeter Method 8. Color fastness to light was tested according to AATCC Test Method 16E, 20 AATCC fading units (AFU), and the color change evaluated with the AATCC color change gray scale.18 To compare the outline sharpness of the prints, a special print pattern containing lines with different widths was used. The widths of the lines on the fabrics were measured by using an Image Analysis System/Version 3.00 (B.A.R.N. Engineering) to evaluate print quality.8

16.3 Results and discussion 16.3.1 Effect of cationic reagent concentration In order to investigate the effect of the cationic reagent concentration on print properties of cationic cotton, each fabric was printed with the cyan reactive ink at a resolution of 720 dpi. The pretreatment was 40 g/l soda ash, 100 g/l urea, and 8 g/l thickener. Printed samples were steamed for 10 min, and then washed according to the five-step washing procedure. The results are presented in Table 16.1. Values of (K/S)f from Table 16.1 clearly demonstrate that cationization enhances the color strength of reactive ink jet prints. The average increase of the color strength was about 34%, suggesting a potential for significant reduction in ink usage. This increase could arise from improved dye substantivity due to the introduction of cationic groups into the cotton, and lower print paste penetration with cationic cotton. As shown from (K/S)r values, cationization decreases penetration of the prints dramatically, leading to higher surface coloration. The Table 16.1 Effect of cationic reagent concentration on the K/S and color fastness values of fabric ink jet printed with cyan reactive ink K/S valuesa

Washfastness

Crockfastness

Cationic reagent conc. (g/l)

…K=S†f b

…K=S†g c

…K=S†r d

Color change

Staining of cotton

Dry

Wet

Untreated 50 75 100 125

18.90 25.22 25.37 25.41 25.75

0.34 0.54 0.29 0.24 0.21

0.94 0.37 0.35 0.35 0.36

5 5 5 5 5

5 4±5 5 5 5

5 4±5 4±5 4±5 4±5

3±4 2±3 2±3 2±3 2±3

a

max ˆ 670 nm. …K=S†f ˆ color strength on front side of fabric. c …K=S†g ˆ color strength of staining on white ground. d …K=S†r ˆ color strength on reverse side of fabric. b

Ink jet printing of cationized cotton with reactive inks

281

lower print penetration may be due to the lower open hole area of the cationic fabric by swelling in sodium hydroxide.12 Furthermore, the penetration was also restricted by the strong ionic attraction between cationic fiber and anionic reactive ink. Values of (K/S)g show that the staining on the white grounds is lower for cationic cotton than for untreated cotton, except with the cationic reagent concentration of 50 g/l. It was also recognized that the washing baths with cationic cotton were nearly clear, especially for the higher cationic reagent concentrations. This result can be attributed to the fact that the higher levels of cationization in the printed areas were enough to prevent any dye loss, and thus reduce the staining on the white ground of the fabric. Generally, cationization had no significant effect on the washfastness ratings which are very good for both fabrics (Table 16.1). On the contrary, the crockfastness of cationic cotton was reduced by 0.5±1.0 rating units. This decrease in crockfastness can be attributed to the higher overall dye concentration and higher surface coloration on the cationic cotton.

16.3.2 Effect of cationization on outline sharpness To determine the effect of the cationization on outline sharpness, each fabric was printed with the cyan ink at a resolution of 720 dpi. The fabric treated with 100 g/l cationic reagent was used as the cationic cotton. The pre-treatment was 40 g/l soda ash, 100 g/l urea, and 8 g/l thickener. Printed samples were steamed for 10 min, and then washed according to the five-step washing procedure. The line widths of the prints were measured by image analysis methods, and the results are shown in Fig. 16.5. This figure shows that the warp direction lines are generally thicker than the weft direction lines on both fabrics. This might be due to the different wicking power of weft and warp yarns. Furthermore, the lines for cationic cotton, especially in the weft direction, are narrower than for untreated cotton. This can be attributed to prevention of spreading of the printed inks by the strong ionic attraction that arises between cationic cotton and the anionic reactive ink. As a result, it is possible to say that cationization enhances the outline sharpness of the prints.

16.3.3 Effect of cationization on steaming time The possibility of reducing steaming time for ink jet printing via cationization was investigated with a series of experiments. Printings were carried out using the cyan ink on both untreated and cationic cotton treated with 100 g/l cationic reagent. The pretreatment was 40 g/l soda ash, 100 g/l urea, and 8 g/l thickener. Printed samples were steamed at different times, and then washed according to the five-step washing procedure. The results are summarized in Table 16.2. As can be seen from the (K/S)f values, color strength of the printed untreated cotton increases by increasing the steaming time up to 8 minutes. On the

282

Digital printing of textiles

16.5 Effect of cationization on line width.

contrary, there is no significant increase with the cationic fabrics, and the (K/S)f values are more or less similar for all cationic samples from 2 minutes to 10 minutes. However, the (K/S)g value and washfastness rating for 2 minutes steaming time on the cationic cotton are slightly worse than for higher steaming times. Taking into account the staining on the white ground and washfastness ratings, it is possible to say that the minimum steaming time must be 4 minutes for cationic cotton and 8 minutes for untreated cotton. Table 16.2 K/S values and color fastness ratings at different steaming times K/S valuesa Fabric type

Washfastness

Crockfastness

Steaming …K=S†f b …K=S†g c …K=S†r d Color Staining Dry time change of cotton (min)

Wet

Untreated

2 4 6 8 10

17.32 18.83 20.44 20.92 20.70

0.38 0.36 0.40 0.38 0.38

0.38 0.49 0.79 0.81 0.84

4 4±5 4±5 5 5

5 5 5 5 5

5 4±5 5 5 5

4 3±4 3±4 3±4 3±4

Cationic (100 g/l)

2 4 6 8 10

27.21 27.68 27.20 27.52 27.53

0.33 0.29 0.28 0.28 0.25

0.27 0.31 0.32 0.34 0.37

4±5 5 5 5 5

4±5 5 5 5 5

4±5 4±5 4±5 4±5 4±5

2±3 2±3 2±3 2±3 2±3

a±d

SeeTable 16.1.

Ink jet printing of cationized cotton with reactive inks

283

The (K/S)r values of the reverse side of untreated and cationic fabrics demonstrate that penetration has been increased by increasing the steaming time. Furthermore, the rate of increase on the cationic cotton due to steaming time is lower than for the untreated cotton because of the strong ionic attraction. The crockfastness results confirm the results obtained in the previous experiment, except for lower steaming times with untreated cotton. The higher crockfastness ratings for two minutes steaming time might be due to lower dye concentration on the untreated fabric.

16.3.4 Effect of cationization on alkali concentration In order to determine the possibility of reducing the alkali concentration when printing reactive inks on cationic cotton, a series of experiments were carried out at various alkali levels with untreated cotton and cationic cotton treated with 100 g/l cationic reagent. The pretreatments were with 100 g/l urea, 8 g/l thickener and either 0, 10, 20, 30, or 40 g/l soda ash. Printing was done with the cyan ink at 720 dpi. The steaming time was 10 minutes and the printed samples were washed with the five-step washing procedure. Table 16.3 gives the (K/S)f values which demonstrate that, as expected, the color strength of the untreated cotton increased with increasing soda ash concentration. In contrast, the (K/S)f values with cationic cotton showed no significant trend. The color yield with no alkali was quite low with untreated cotton since some alkali is necessary for the reactive ink to covalently bond with the fiber, whereas with cationic cotton, no alkali is necessary for high color yields since the anionic dye molecules can form strong ionic bonds with the cationic fiber. The highest washfastness achieved with untreated cotton was with Table 16.3 Effect of cationization on alkali concentration for pretreatment process K/S valuesa Fabric type

Washfastness

Crockfastness

Soda ash …K=S†f b …K=S†g c …K=S†r d Color Staining Dry conc. (g/l) change of cotton

Wet

Untreated

0 10 20 30 40

1.41 19.76 20.29 20.52 20.57

0.52 0.46 0.39 0.43 0.44

0.99 1.11 1.02 1.06 0.98

2 4 4±5 5 5

4±5 4±5 5 5 5

5 5 5 5 5

4±5 3±5 3±5 3±5 3±5

Cationic

0 10 20 30 40

27.29 26.80 27.26 26.87 26.28

0.66 0.34 0.33 0.35 0.35

0.26 0.31 0.37 0.41 0.44

4±5 5 5 5 5

4±5 5 5 5 5

5 4±5 4±5 4±5 4±5

4 3 3 3 3

a±d

SeeTable 16.1.

284

Digital printing of textiles

30 g/l soda ash. Cationic cotton, on the other hand, had maximum washfastness with only 10 g/l soda ash. White ground staining, (K/S)g, was near optimal also at 30 g/l soda ash for untreated cotton and 10 g/l soda ash for cationic cotton. Crockfastness results, both wet and dry, generally indicated slightly worse performance for printed cationic cotton, probably due to the higher level of surface dye obtained with that system (higher (K/S)f and lower (K/S)r).

16.3.5 Effect of cationization on thickener concentration The thickener was applied at various concentrations (0, 4, 8, or 12 g/l) in the pretreatment bath to determine the effect of cationization on the optimal level of thickener. The concentrations of soda ash and urea in the pre-treatment were 40 and 100 g/l respectively. The fabrics, untreated cotton and cotton cationized with 100 g/l reactant, were printed with the cyan ink at 360 and 720 dpi resolutions. The printed samples were steamed for 10 minutes and then washed with the five-step washing procedure. The results are shown in Table 16.4. As seen in the (K/S)f values, cationization significantly enhanced the color strength of the printed fabrics regardless of thickener concentration or print Table 16.4 Effect of thickener concentration on K/S values and color fastness rates of cyan ink jet printed fabrics K/S valuesa

Washfastness

Crockfastness

Fabric Thickener …K=S†f b …K=S†g c …K=S†r d Color Staining Dry type conc. (g/l) change of cotton (print resolution)

Wet

Untreated (360 dpi)

0 4 8 12

14.56 16.26 17.01 18.67

0.26 0.24 0.26 0.24

0.70 0.61 0.56 0.47

5 5 5 5

5 5 5 5

5 5 5 5

3±5 3±5 3±5 3±5

Cationic (360 dpi)

0 4 8 12

20.15 22.34 23.13 22.95

0.10 0.14 0.09 0.14

0.26 0.23 0.23 0.19

5 5 5 5

5 5 5 5

5 5 5 5

3 3 3 3

Untreated (720 dpi)

0 4 8 12

18.26 18.83 19.21 21.19

0.37 0.37 0.35 0.38

1.36 1.08 0.97 0.88

5 5 5 5

5 5 5 5

5 5 5 5

3±5 3±5 3±5 3±5

Cationic (720 dpi)

0 4 8 12

25.50 25.24 25.93 26.23

0.37 0.42 0.29 0.26

0.47 0.38 0.37 0.36

5 5 5 5

5 5 5 5

4±5 4±5 4±5 4±5

2±5 2±5 2±5 2±5

a±d

SeeTable 16.1.

Ink jet printing of cationized cotton with reactive inks

285

resolution. The cationic fabric forms strong ionic bonds with the anionic ink, leading to high surface coloration without the need for thickener. With the untreated fabric, color strength increased with thickener concentration as expected since the more viscous ink penetrated the fabric less, leaving more color at the surface. Values of (K/S)f for cationic cotton printed at 360 dpi were comparable to those for untreated cotton printed at 720 dpi. Since four times more ink is theoretically used in 720 dpi prints than in 360 dpi prints, the savings in ink from using cationic cotton can be quite significant. Values of (K/S)g, a measure of background staining, are extremely low for the cationic fabric printed at 360 dpi resolution. The wash baths were also quite clear for these fabrics. This indicates that the level of cationization was high enough to strongly bond all the ink molecules, preventing any color migration to the white background. Color values on the reverse face of the fabrics, (K/S)r, are considerably lower for the printed cationic fabrics than for the printed untreated fabrics, providing another indication that ink penetration is significantly decreased due to strong ink±fiber interactions.

16.3.6 Effect of thickener concentration and cationization on outline sharpness In order to evaluate the effect of the thickener concentration on outline sharpness, the line widths of fabrics printed with cyan ink at 360 dpi were measured by an image analysis method, and the results are shown in Table 16.5. This

Table 16.5 Effect of thickener concentration and cationization on outline sharpness Designed line Thickener width (mm) conc. (g/l)

Actual line width in weft direction (mm) Untreated

Cationic

0.5

0 4 8 12

0.744 0.689 0.675 0.679

0.644 0.553 0.573 0.593

1.0

0 4 8 12

1.168 1.150 1.127 1.150

1.127 1.056 1.047 1.047

1.5

0 4 8 12

1.673 1.660 1.631 1.671

1.664 1.531 1.595 1.631

286

Digital printing of textiles

shows that, in general, the lines for cationic cotton are finer than for untreated cotton as seen in Section 16.3.2. The results also show that thickener concentration has a significant effect on the outline sharpness. The line widths on the untreated fabrics reduce by increasing the thickener concentration from 0 to 12 g/l. This result demonstrates that a suitable amount of thickener is required for preventing the entry of ink to the capillaries, since blockage of the capillary spaces reduces the wicking effects.5 On the cationic fabrics, generally, the best outline sharpness (the finest lines) has been obtained with a thickener concentration of 4 g/l, and the line widths increase above 4 g/l. Higher thickener concentrations than 4 g/l with cationic cotton may cause more surface coloration, and thickness of the lines increases again. Thus, it is possible to say that cationic cotton requires less thickener than untreated cotton because cationization reduces pore size (capillary spaces) of the fabric and a lower thickener concentration may be enough for the blockage of the capillary spaces. In addition, ionic attraction between cationic cotton and ink may have an important effect on preventing the spreading of printed ink.

16.3.7 Printing results with CMYK colors and reduced chemical levels The previous experiments indicated that ink jet printing with cyan reactive ink on cationic cotton can be carried out with reduced alkali and thickener concentrations and a less lengthy after-wash compared to reactive printing on untreated cotton. To confirm these results with a variety of inks, the following experiment was carried out. Untreated cotton and cotton cationized with 100 g/l cationic reactant were printed with CMYK (cyan, magenta, yellow, black) reactive inks at 320 dpi resolution. Different pretreatments were used for untreated and cationic cotton fabrics. Untreated fabrics were padded with 8 g/l thickener, 40 g/l soda ash, and 100 g/l urea, while cationic fabrics were padded with 4 g/l thickener, 20 g/l soda ash, and 100 g/l urea. Both fabric types were steamed for 10 minutes after printing. The printed untreated fabrics were washed with the five-step washing procedure, while the printed cationic cotton fabrics were washed with the threestep procedure (see Section 16.2.5). K/S and fastness properties of the fabrics are given in Table 16.6. The color strength values, (K/S)f, of all inks with cationic cotton are significantly higher than the color values of the inks with untreated cotton. The improvements range from 25% higher for cyan to a remarkable 89% higher for black. The color values of the reverse side of the fabrics, (K/S)r, were also significantly less for all inks on cationic cotton, indicating that the inks penetrated less with cationic cotton. The higher (K/S)f values could arise from less penetration as well as from increased color utilization, since it was noted that the wash baths for all cationic fabrics were colorless, implying near 100%

Table 16.6 K/S values and fastness properties for fabrics ink jet printed with CMYK colors with reduced chemicals for cationic cotton K/S valuesa

Light fastness (20 AFU)

…K=S†g c

…K=S†r d

Color change

Staining of cotton

Dry

Wet

Yellow Magenta Cyan Black

13.58 17.97 18.51 11.16

0.03 0.02 0.27 0.02

1.08 0.46 0.37 0.76

5 5 5 5

5 5 5 5

5 5 5 5

4 3±5 3±5 3±5

4 4 4±5 4

Yellow Magenta Cyan Black

17.93 25.90 23.11 21.12

0.17 0.18 0.10 0.08

0.24 0.24 0.17 0.21

5 5 5 5

5 5 5 5

5 4±5 5 5

3±5 2±5 3 2±5

3±5 4±5 4 4±5

Color

Untreated

Cationic

Yellow max ˆ 430 nm; magenta max ˆ 530 nm; cyan max ˆ 670 nm; black max ˆ 610 nm. SeeTable 16.1.

b±d

Crockfastness

…K=S†f b

Fabric type

a

Washfastness

288

Digital printing of textiles

color fixation. However, the staining of the unprinted areas, (K/S)g, was higher on cationic cotton for all inks except cyan. A higher cationization level or use of inks with a greater interaction to cationic cotton could overcome this difficulty. Washfastness, dry crockfastness, and lightfastness results were essentially comparable for the untreated and cationic fabrics. Wet crockfastness ratings were generally 0.5±1.0 units lower for cationic cotton than for untreated cotton. Since the cationic cotton fabric clearly had more surface color than the untreated cotton, a more reliable evaluation would be to compare the wet crockfastness of untreated and cationic cotton fabrics printed to the same color depth.

16.4 Conclusions The results demonstrate that cationization of cotton with 2,3-epoxypropyltrimethylammonium chloride can be used to improve reactive ink jet printing properties. As a result of cationization, the color yield of the prints was increased significantly, while the ink penetration was reduced. Hence, cationization permits faster printing, especially with dark-colored designs, since darker shades can also be printed with lower resolution on cationic cotton. Print outline sharpness was also increased. Due to the greater interaction of reactive inks with cationic cotton, ink consumption, steaming times, alkali concentrations, thickener concentrations, and washing procedures could all be reduced from the levels necessary with untreated cotton, leading to a shorter and less costly printing process. However, cationization had some adverse effects on the wet crockfastness and white ground staining values with some reactive inks. To obtain lower white ground staining and higher wet crockfastness rates, further research investigating different reactive inks which have more substantivity, other anionic dyes such as direct and acid dyes, different thickeners and different print parameters is needed.

16.5 References 1. Hees U, Freche M, Kluge M, Provost J and Weiser J (2002), Int. Conf. on Digital Printing Technol., San Diego, CA, 242. 2. Dawson T L (2001), Color. Technol., 117 185. 3. Clark D (2001), Proc. AATCC Int. Conf., Greenville, SC 379. 4. Aston S O, Provost J R and Masselink H (1993), JSDC, 109 150. 5. Kulube H M and Hawkyard C J (1996), ITB Dyeing/Printing/Finishing, 3 5. 6. Kulube H M and Hawkyard C J (1998), South African J. Sci., 94 469. 7. Lavasani M R B and Hawkyard C J (2000), Melliand Int., 6 152. 8. Fan Q, Kim Y K, Lewis A F and Perruzi M K (2002), Int. Conf. on Digital Printing Technol., San Diego, CA, 236. 9. Li X and Tincher W C (1999), Text. Chem. Colorist & Amer. Dyes. Rep., 1 307. 10. von der Eltz A, Schrell A and Russ W H (1994), USP5348557.

Ink jet printing of cationized cotton with reactive inks 11. 12. 13. 14. 15.

289

Hutter G F and Matzinger M D (2000), USP6156384. Kanik M and Hauser P J (2002), Color. Technol., 118 300. Hauser P J and Kanik M (2003), AATCC Review, 3(3) 25. Kanik M and Hauser P J (2003), Color. Technol., 119 230. Kanik M, Hauser P J, Parrillo-Chapman L and Donaldson A (2004), AATCC Review, 4(6) 22. 16. Tabba A H (2000), Cationization of cotton with 2,3-epoxypropyltrimethylammonium chloride, MSc Thesis, North Carolina State University, USA. 17. Broadbent A D (2001), Basic Principles of Textile Coloration, Bradford, Society of Dyers and Colourists. 18. AATCC Technical Manual (2002), Research Triangle Park, North Carolina, American Association of Textile Chemists and Colorists.

Part IV Design and business

17 Digital printing and mass customization M F R A L I X , [TC]2, USA

17.1 Introduction Since it was coined in the book Future Perfect in 1987 (Davis and Meyer, 1998), `mass customization' has become an industrial household name and can now be found in numerous articles written on the subjects of innovation, technology management, product development, and supply chain management. Shifts in consumer behavior have caused many industries, including the soft goods industry, to examine the trend toward mass customization and the technologies that support it. It is a well-known fact that consumers want more personalized products. The companies that will remain competitive in today's marketplace are those that can correctly anticipate consumer wants and incorporate effective business strategies with emerging technologies to respond to those wants. Digital ink jet printing is an emerging technology that has the potential to satisfy consumer expectations as well as impact company strategy. Ink jet has taken over the office printing market and is now the technology of choice for home printers, even though the print resolution is not quite as good as that of laser printers. Equipment suppliers, textile manufacturers and apparel producers continue to develop applications for the use of ink jet and other digital printing technologies for printing fabrics. The supply chain implications of this technology in the soft goods industry have been apparent for some time and are significant. How do companies in the sewn products industry position themselves to implement such new concepts? How are the technologies identified? What are the systems and processes that must be implemented, and how are they structured? What role should technology management play in the short term versus long term decisions that are made? Can accurate projections be made regarding the future of a breakthrough technology within an industry? Will digital printing provide a mechanism for the soft goods industry to meet some of the requirements of its customers? The answers to these questions require that industry leaders think about how their organizations are preparing for the future.

294

Digital printing of textiles

This chapter focuses specifically on digital ink jet printing as an enabling technology for mass customization and its potential impact on the soft goods supply chain. In addition, a comparison will be made between a traditional supply chain for woven fabric and a supply chain that will probably exist for digitally printed garments. The intent is to better understand the impact of implementing digital technologies and to explore the relationship to systems that are more capable of adapting to changes in consumer wants and expectations. These concepts have been tested using industry representatives that have a thorough knowledge of advanced technology and are familiar with the latest developments in the soft goods industry. Several conclusions can be drawn from the analysis. First, business and manufacturing processes can be automated through the use of digital systems. Automation is not restricted to digital technology; however, digital processes can easily be reconfigured and are not as sensitive to changes in production requirements or product attributes. Section 17.2 consists of a review of the trends in supply chain strategies. The most important take away is that mass customization strategies are infiltrating all types of companies and are rewriting the rules about how products and services are becoming inseparable. Second, technology forecasting and technology management strategies serve as a basis for exploring the impact of a breakthrough technology such as digital printing. Understanding these methodologies provided a foundation to approach expert thinkers with the vision of a totally digital supply chain. Section 17.3 speaks to the fact that there are limitations to mass customization. Such limitations may be process oriented or technology oriented. For example, not everything about a product must be customizable and a production batch size of one may be the ultimate example of mass customization; however, it is not a requirement. Section 17.4 identifies three requirements for mass customization that are met with digital printing. The need for speed, the need for seamlessly flexible automation, and the need for integration throughout the order-to-delivery process are explored. Section 17.5 reinforces the point that product life cycles in the soft goods industry are becoming shorter and shorter. Coupling that with the fact that consumers want more individualized products means that such technologies as digital printing are positioned to capitalize on the changing market expectations. In Section 17.6 reference is made to such demands as time compression, speed to market and mass customization that are currently placed on the soft goods industry. The importance of technology forecasting is reviewed as it relates to these demands and how mass customization using digital printing may cause a reorientation of the supply chain. Section 17.7 provides an overview of how supply chains are ordered. An analysis of traditional printing supply chains is used to set the stage for the section that follows on digital printing supply chains. The key point in Section 17.8 is that if digital printing is to be used for the customization of products, decisions must be made closer to the point of consumption and the responsibility for the coloration of the fabric must shift.

Digital printing and mass customization

295

An overriding theme in this chapter on the use of digital printing for mass customization is that one must look beyond the current state of the technology and/or the supply chain strategies that might employ it. There are gaps between the digital process technologies that have been developed and the `digital islands' that have been created. There are opportunities to integrate the digital islands and this serves as the basis to think in terms of a totally digital supply chain. Since digital systems are not as sensitive to changes in product configuration, a digital supply chain should not have the same constraints that exist in a traditional physical product supply chain. The last section is dedicated to the future. Current developments promise to improve the efficiency and lower the cost of digital printing. Not only will digital printing allow for the customization of individual print patterns, it will eventually drive the development of the mass customization of solid colors using digital technology.

17.2 From craft to mass production to mass customization Prior to the industrial revolution, manufacturing was considered a craft. Products were typically custom made to meet the needs of a particular individual. Even though many products were similar, parts from one product could not necessarily be interchanged with the same parts on another product. Since products tended to be relatively expensive, access was limited primarily to the upper class or aristocracy. With the advent of the industrial revolution and the concept of interchangeable parts, like products began to be produced in large quantities and were made available to the middle class. Because of the large production quantities of like products, the costs were low enough that they became affordable for most people. The concept of interchangeable parts relates more directly to hard goods even though the soft goods industry also adopted the principles of mass production. In the soft goods industry finished garment parts are not yet totally interchangeable; however, large quantities of the same style will be cut and assembled as a group. While it is not practical to remove a component from a finished garment and reattach it to another garment, any component in a batch of garments ready for assembly can often be used on any of the garments in that batch. In addition, many of the components, which are also referred to as `cut parts', can be used on multiple sizes of the same product. Mass customization has emerged as a practice that combines the best of the craft era with the best of the mass production era. Not to be confused with custom-made, mass customized products may still be manufactured in relatively large quantities; however, each item might be slightly different based on the needs and desires of the individual end customer. Joe Pine (1993) refers to the goal of mass customization to be to provide enough variety so that the wants of

296

Digital printing of textiles

the consumer are satisfied, whereas the goal of mass production was to produce at sufficiently low enough cost so that everyone could have one. Prior to the 1980s, the mass production system dominated the soft goods industry. It was characterized by large batches of product that moved slowly through the manufacturing process. Huge inventories were carried in raw materials, work-in-process and finished goods. As a result, lead times were extremely long and commitments for delivery anywhere along the supply chain could be months. Even the sewn product manufacturers operated on a 12±16 week rolling forecast. Relationships between players in the supply chain were often adversarial and companies operated independently of their suppliers as well as of their customers. At the end of a selling season it was common for manufacturers and retailers to markdown leftover inventory. In 1986 the concept of Quick Response was introduced to the soft goods industry. It was the industry's first collaborative attempt to better manage the supply chain linkages between manufacturers, their customers, their suppliers and their suppliers' suppliers. Such technologies as EDI (Electronic Data Interchange) and bar coding were intended to improve the accuracy of data and information flows through the pipeline. Instead of adversarial relationships, Quick Response called for partnerships between the retailers and their suppliers. Through better information flows, retailers could count on reduced lead times from their suppliers and better assure that the right product would be available at the right time. Even today, many companies continue to focus their efforts on SCM (Supply Chain Management). During the early 1990s agile manufacturing was introduced as a strategy to help the United States regain its position as a world leader in manufacturing. Japan, for example, had taken considerable market share from US companies, particularly in the automotive and electronics industries. Agility further refined the concepts taught by Quick Response, JIT (Just-in-Time), and lean manufacturing and, according to Goldman (1997), was the collection of best practices and trends that were already underway in some of the leading US companies. Through the use of information technology, flexible automation, a knowledgeable work force and team-based short cycle manufacturing, manufacturers would be better able to respond to the needs of individual customers. In 1991, mass customization was introduced in the literature as an example of agile manufacturing (Nagel et al., 1991). It was not, however, presented as a requirement for a company to be considered `agile'. Since Pine's (1993) book on the subject, mass customization has become an established business practice and is now a topic that is discussed in almost all work that pertains to product development or supply chain management. A study of the shift in supply chain strategies from the batch and queue system to Quick Response to agile manufacturing to mass customization reveals several trends. Information is becoming instantly accessible. Production batch sizes and minimum order quantities are getting smaller and in many cases it is

Digital printing and mass customization

297

possible to produce a single item cost-effectively. Product differentiation is increasing exponentially with the end consumer having more input into the configuration of the finished product. The `order to delivery' cycle time is often expressed in terms of hours instead of weeks or months.

17.3 Limitations of mass customization The ability for manufacturers to offer mass customization is limited by their ability to get consumer information to the `workplace' doing the customization. Mass customization is also limited by the extent to which production workers have been cross-trained and empowered to accept responsibility for the manufacturing and `customization' process, so that they can accurately respond to those needs. In addition, manufacturers are constrained by the lack of available technology that can be reconfigured quickly, easily, and cost effectively to meet consumer needs. On the other hand, mass customization does not mean that everything about a product is customizable. This may have been true in the craft era, and may still be true for some products, but it is not true for mass customization. Pine says that `variety in and of itself is not customization ± and it can be dangerously expensive'. Customizable features must include only those things the customer determines are important and the customized products should not necessarily cost any more, even though research shows that many people are willing to pay more for customized products that are delivered quickly. Information technology and automation play a key role in mass customization in that they create the linkage between a customer's preferences and the ability of a manufacturing team to construct products based on those preferences. In the case of apparel, ink jet printing technology offers the potential for color preference to become a customizable feature. There are other product features, such as garment fit, that are also desirable as consumer options. Because technology development is an ongoing process, product features that are not presently customizable can become so when affordable technology is developed that makes it possible. The selection of fabric color at the individual garment level (which includes fabric print specification) is an area that currently offers little opportunity for mass customization. The primary reason is that to do so requires the production of 1 to 1Ý yards of individualized fabric. This requirement is dramatically different from the way in which traditional textile production technologies have been developed. Sewn products manufacturers typically commit to fabric purchases months in advance of their receipt at the production plant and they are required to make yardage commitments and issue purchase orders for fabric by color and/or print design. It is also a common practice for textile producers and fabric finishers to

298

Digital printing of textiles

require minimum purchase quantities of 1000 yards or more because until now it has not been economical to dye or print small lots of fabric. Digital technology provides an opportunity to shorten the lead-time on precolored fabric purchases and minimize the amount of raw materials inventory that must be carried by sewn product manufacturers. In fact, digital technology is already being applied to some of the pre-production processes associated with color application. Mass customization of printed garments will require the use of digital printing technology even though the reverse is not true; i.e., the implementation of digital systems can be accomplished as a replacement for current fabric printing technology without the need to implement a mass customization strategy. The primary digital printing focus is on ink jet technologies because ink jet allows the direct application of dyes and/or inks to the textile substrate without the need for an intermediate step in the process. It is also anticipated that this technology will shift some of the responsibility for fabric coloration from textile manufacturers and converters to apparel and other sewn products manufacturers.

17.4 Time, technology, and connectivity The primary focus of this section is on the technology management issues surrounding the use of digital printing for the soft goods industry. Therefore, it is essential to explore the extent to which direct digital printing fits the description of enabling technologies for mass customization and personalized products. Companies that can accept input from customers in the design of their products and can manufacture and deliver to customer requirements in a very short period of time at a cost close to mass production methods will create tremendous new opportunities to capture market share. This customer input might be in the form of color preference, personal body measurements, print design, fabric type, garment features, or price point. Stan Davis and Christopher Meyer (1998) state that `Connectivity, Speed, and Intangibles ± the derivatives of time, space, and mass ± are blurring the rules and redefining our businesses and our lives. They are destroying solutions, such as mass production, segmented pricing, and standardized jobs, that worked for the relatively slow, unconnected industrial world.' They state further that `Speed means a shift from relying on prediction, foresight, and planning to building in flexibility, courage, and faster reflexes.' Souder and Sherman (1994) put mass customization in the context of technology-based challenges by asking two questions. `How can organizations make their products more responsive to changing customer needs?' and `How can organizations adapt their operations (manufacturing or service and distribution) so that process technology pays no quality penalty for the speedier deliveries, small order quantities, and reliable delivery schedules?' The

Digital printing and mass customization

299

implications of these two questions are tremendous. In the context of digital printing, organizations could use the technology to potentially respond to whatever the current color combinations or fashionable print designs happen to be. Because the printing process uses a digital data stream from input to output, the system is indifferent to production quantities. After the designs have been created, a production order for 1000 yards of one print is not considered to be different from one yard each of 1000 different prints. Suzanne Berger refers to the `convergence of future consumer preferences, market forces and technological opportunities' as the drivers that will cause some industries to implement `totally flexible' production systems (Berger et al., 1991). She refers directly to: `custom-tailoring of products to the needs and tastes of individual customers'. Again, direct digital printing can enable the production of individualized products. Tom Peters (1991) states that among other things, successful firms will be `oriented toward differentiation, producing high value-added goods and services, creating niche markets'. John Seely Brown (1997) says that `technology will become so flexible that users will be able to customize it ever more precisely to meet their particular needs ± a process that might be termed ``mass customization''.' A potential digital printing market that has been discussed with representatives of several retailers involves the introduction of a line of designer blouses. Each blouse would be digitally colored with a `limited edition' print that is signed and numbered by the artist/designer. Assuming that high-end fabrics such as 100% silk are used and that the garment construction is currently fashionable, additional value is created because each blouse would be uniquely identified and the supply would be controlled. To put it yet another way, Handfield and Nichols (1999) contend that `the second major trend facing organizations today is the demand for ever-greater levels of responsiveness and shorter defined cycle times for deliveries of highquality goods and services.' They also say that the only companies that will be successful are those that have the ability to `mass customize'. And finally, the authors of Blur, who state that the rate of change today is so rapid that it is only a blur, make the following claim: It used to be one size fits all. Now, Porsche says it never makes the same car twice. Whatever your offer, you must tailor it each and every time, with the needs of the individual buyer or user in mind. Cheeseburgers, hotel rooms, pants, software programs, office chairs, retirement plans, skis, kitchen appliances . . . . Today, every offer, no matter its nature, can be customized, along many dimensions (Davis and Meyer, 1998).

17.5 Product life cycles The result of the convergence of time, technology, and connectivity is that companies are required to introduce products more frequently with shorter

300

Digital printing of textiles

expected product life cycles. Also, the days of mass production characterized by huge volumes of identical products have given way to strategies that call for individualization of products made in arbitrary quantities. Traditional manufacturing techniques are being challenged at the same time that technological innovations are providing alternative mechanisms for meeting customer expectations. As product life cycles become shorter and the volume demanded of each product becomes smaller, the need for technologies that are designed for shorter runs and smaller quantities will become greater. Direct digital printing is one response to the need for short runs of printed fabric and it is already being used on a limited basis for the production of new print designs. The demands of consumers and the roles played by manufacturers are not unique to the soft goods industry; however, the soft goods industry is more affected by changes in consumer tastes and whims than most other industries. Many of the end products also have an extremely short product demand life span. To achieve a leadership role and to respond effectively, the textile and apparel industry must innovate. Manufacturers and suppliers, including machinery and equipment vendors, must recognize the trends of their customers and the expectations they have for quality, personalization, service, and delivery. They must identify the strategic directions, including technologies, required to meet these expectations.

17.6 Forecasting the opportunities Time compression, speed to market, and mass customization are terms that are commonplace in today's globally competitive environment. What is not commonplace are the integrated solutions to these challenges. In the case of direct digital printing, the first applications have been substitutions for other printed fabrics. The potential to replace dyeing for solid colors or Jacquard fabrics is not as feasible at this time. The questions that were posed in the opening section can possibly be better answered with a more complete understanding of technology management and the various methods of technology forecasting that have evolved. Technology forecasting can be useful to steer development programs toward innovations that could either reduce costs to increase profits, or provide new services for customers. Technology forecasting can also be used to help identify new business opportunities, and may lead to a modification of corporate goals. The desire to be different and the desire to be unique have a major influence on the demand that individuals in the US have for clothing. If it were possible to predict consumers' future wants and expectations, companies could align their technology development and implementation strategies with the expectations of their customers. Until a viable business practice is established, it is appropriate to question the potential impact of digital printing on the soft goods supply chain. Kurt Salmon

Digital printing and mass customization

301

Associates (KSA) completed a study of mass customization in 1997 (KSA, 1997) and concluded that the early adopters of mass customization technologies will have an advantage over their competitors. The study also showed that the manufacturer's profit on a seasonal garment sold through department stores can increase from 8.6% to 18.1% without affecting the retailer's gross margin. When comparing fashion garments sold through department stores, the manufacturer's profit increases from 6.9% to 17.9%, again without impacting the retailer's gross margin. In addition to reviewing the results of the KSA study, an analysis of technology forecasting is appropriate. Technology forecasting can be used to determine not only when or if digital printing will have an impact, but how and to what extent it will impact the industry. It can also be used to predict what may happen to the competitive environment of the organizations that conduct business in the soft goods supply chain. The most common form of technology forecasting is trend analysis. It is easy to understand and generally assumes that the future will be an extension of the past. As with most processes, there are simple as well as elaborate approaches to conducting trend analysis (Cetron, 1969; Martino, 1993). The easiest trend analysis method to apply is trend extrapolation. It is also popular because it is relatively inexpensive to conduct. By mapping the past graphically, the future can be predicted through an extension or `extrapolation' of the past. It is simple and inexpensive but should not be considered reliable beyond a relatively short time period. It would be impractical to rely on this method to predict technology events a decade in the future. In fact, decisions that look beyond a one-year time frame should be made using an alternative forecasting method (Martino, 1993). Trends in the soft goods industry include: · · · · · · ·

Shorter product life cycles Lower desired inventories throughout the supply chain More differentiated products New forms of retailing such as television and the Internet Increased use of information technology Reduction in time to market Marketing to a customer of one.

Time-series estimation goes one step further in that it attempts to account for the variations in the slope of a trend line over time. For example, it can be used to identify specific trends such as seasonal variations. It is more complicated than trend extrapolation and often requires the use of a computerized statistical analysis package. A common projection using time series analysis is the forecast of sales by quarter. Regression analysis can also be used for time series estimation; however, it is especially useful when more than one variable is required to explain the present state or predict the future state. Regression can be used for non time-series forecasts as well. It has become very popular because of the availability of such

302

Digital printing of textiles

computer programs as SAS and Statgraphics and because it simplifies understanding of more complicated forms of trend analysis such as econometrics. S-curves have been used to graphically illustrate the growth pattern of new technologies. Fisher and Pry (1972) used them to show that the rate of acceptance of new product introductions is generally slow at first. This slow start is followed by a period of fairly rapid growth or technology advancement as the new products penetrate the marketplace. Once market saturation occurs and the product has reached the last stages of its life cycle, growth and advancements taper off (Fig. 17.1). Anderson and Tushman (1997) refer to the introduction of a breakthrough product or technology as a technological discontinuity. A technological discontinuity triggers a technology cycle that begins with an era of ferment during which the new technology `displaces its predecessor during an era of substitution'. Following the discontinuity but overlapping the era of substitution is an era of design competition where more refined versions of the technology are introduced. `The emergence of a dominant design marks the end of the era of ferment and the beginning of a period of incremental change' (Anderson and Tushman, 1997). During this time `the rate of design experimentation drops sharply and the focus of competition shifts to market segmentation and lowering costs (via design simplification and process improvement)' (Anderson and Tushman, 1997). This change corresponds with the last stages of the life cycle of a technology, which would in turn be followed by another technological discontinuity. Figure 17.2 provides a graphical view of the advancements in

17.1 Technology S-curve, general form (Betz, 1993).

Digital printing and mass customization

303

17.2 Technology S-curve for progress in lamp technology (Betz, 1993).

lamp technology following two breakthroughs, one in incandescent lamps and the other in fluorescent lamps. While digital printing is currently in the very early stages of its life cycle in the soft goods industry, ink jet printing has entered the era of ferment for the office printer market. The market `is expected to mature in the next few years' (I.T. Strategies, 1999). Ink jet printers have replaced laser printers for many applications and the era of design competition has also begun. Ink jet offers lower cost, a smaller footprint, and greater flexibility than laser printers. Since its capability is still somewhat limited for textile substrates, it is not yet an accepted substitution in the soft goods industry for analog printing methods (Clark, 1999). Believers in trend analysis recognize that history can and does repeat itself, that business is cyclical, and that problems encountered in one industry are eventually experienced in another. Historical analogies is the formal name given to the method of trend analysis that projects the future based on a study of past practices in other industries and draws inferences from the lessons learned (Cetron, 1969). Digital printing in the soft goods industry should track closely what has happened in the paper industry. Analog technologies have been replaced by digital technologies and ink jet is now the dominant technology. A

304

Digital printing of textiles

similar substitution should occur as the soft goods industry transfers learning from the paper printing industry.

17.7 Traditional supply chains Supply chains are characterized by the movement of goods from the initial raw materials provider, through one or more manufacturing processes, distribution, retailing, and delivery to the end consumer. Many of the terms used to identify a supply chain are interchangeable; however, some distinctions should be made. `Supply chain', `industrial value chain', `value-added chain' and `integrated supply chain' are used interchangeably. Each of these terms may be used to refer to a set of organizations that collectively and sequentially bring a product from raw materials to the end consumer, or may refer to the set of activities required to produce a product and deliver it to the end consumer. An individual organization may be responsible for multiple activities that get captured as a single step in a supply chain. At other times each activity will be identified as a separate step. Figure 17.3 illustrates these two scenarios for a soft goods supply chain: one that has been segmented into organizational functions; and one that has been segmented into the activities that are performed. Depending on the company, `supply chain' can mean either the set of organizations of which it is a part or the activities that it conducts.

17.3 Soft goods supply chains.

Digital printing and mass customization

305

A tremendous amount of work has been done to track the flow of materials and information through a supply chain. The most notable in the soft goods industry was accomplished through a project known as DAMA (DemandActivated Manufacturing Architecture). This project resulted in the creation of process maps for such products as men's cotton slacks (DAMA, 1995), bed sheets, and nylon supplex parkas. `These process maps included the movement of the goods and the conversion steps through which the goods passed, the information behind each step, quality checkpoints, and time elements for each step' (Kuglin, 1998). The soft goods industry supply chain consists of fiber producers, yarn manufacturers, fabric manufacturers, fabric finishers, apparel and other sewn products manufacturers, wholesalers, retailers, and consumers. The type of organization that will control a particular step in the chain depends on the finished product that is being made. For example, for a woven product such as men's slacks, some of the significant supply chain steps and major players could be as follows: · · · · · ·

Fiber producer s: Invista, KoSa (formerly Hoechst-Celanese) Yarn manufacturers: Dixie Yarns, Parkdale, Unifi Textile manufacturers: Burlington, Milliken, Greenwood Mills Fabric finishers: Cherokee, Cranston Print Works, Lortex Apparel manufacturers: Russell Corp., Levi Strauss, VF Corp. Retailers: Dillards, J.C. Penney

While the processes in the chain are somewhat distinct, clear boundaries do not always exist to differentiate companies that are traditionally called `textile manufacturers' from those that are referred to as `apparel manufacturers'. For example, Russell Corporation, a major supplier of knit products, will purchase fiber but will spin its own yarn, knit its own fabric, dye the fabric, cut the fabric into components, and sew the components into finished garments for distribution to its customers. Other knitwear manufacturers will purchase yarn, knit and dye the fabric, and cut and sew garments. Others will buy the fabric from which they cut and sew finished garments. Manufacturers of woven garments typically outsource the manufacturing and finishing of the fabric and focus on cutting and sewing the fabric into finished garments. When considering the investment in digital printing technology, fabric finishing becomes an even more important step in the soft goods supply chain. This step includes fabric coloration, which can involve dyeing or printing the fabric prior to cutting it into component parts. Figure 17.4 represents traditional supply chains for making and coloring fabric and sewing the fabric into finished garments. The black boxes represent activities that are typically performed by textile manufacturers; shaded boxes represent activities that are typically performed by either converters or textile manufacturers; and white boxes represent activities that are typically performed by apparel and sewn product

306

Digital printing of textiles

17.4 Alternatives for coloring and sewing fabric.

manufacturers. At Russell Corporation, all of the activities (knitting replaces weaving) shown in the second column are performed as a part of the internal supply chain. Companies today should be aware of the challenges affecting the supply chains of which they are a part and the changes that can impact their internal supply chains. For example, digital printing has the potential to significantly alter some segments of the soft goods industry supply chain. While some organizations may consider digital printing as merely an alternative technology for rotary screen printing, others consider it a `breakthrough' technology that has the potential to be applied to new families of mass customized products. Wheelwright and Clark refer to breakthrough products as those that `depart significantly and fundamentally from existing practice'. Furthermore, breakthrough products `may introduce highly innovative product or process technology, open up a new market segment, or take the business into a totally new arena' (Wheelwright and Clark, 1995). A digitally printed line of limited edition blouses would definitely enter into a niche market and a company like Brooks Brothers could market digitally printed ties as an extension to its custom-made dress shirts. Naturally, the market for these products would be limited to customers who are less sensitive to price, although it is common for new product innovations to be expensive at first and then to decrease in price as demand rises. It is also common for breakthrough or disruptive technologies `to be used and valued only in new markets or new

Digital printing and mass customization

307

applications; in fact, they generally make possible the emergence of new markets' (Bower and Christensen, 1997). Betz's research (Betz, 1993) shows that `major changes occur in an industrial value chain when basic inventions are discovered and introduced into a sector of the industrial value chain.' In the case of digital printing, such products as the mass customized, limited edition blouses can be manufactured more cost-effectively by shifting the responsibility for printing to the sewn product manufacturer. This transition is a major departure from traditional supply chain practices since fabric is usually printed before it is stocked at the apparel plant.

17.8 Direct digital printing supply chains Figure 17.5 diagrams two direct digital printing supply chains. One involves printing on cut parts and the other involves printing prior to cutting. Printing on cut parts can also be accomplished using belt screen printing technology. In both cases the responsibility for printing shifts from the textile manufacturer and/or converter (refer to Fig. 17.4) to the apparel manufacturer. The primary reason for shifting the printing activity is that printing fabric after the marker is made provides additional benefits.

17.5 Printing on cut parts versus uncut cloth.

308

Digital printing of textiles

17.6 Textile mill responsible for printing.

Markers are traditionally made by the sewn product manufacturer and typically are not generated until a request for finished product has been made. Even so, responsibility for printing could stay with the textile manufacturer, and responsibility for cutting or for cutting and marking could shift from the sewn product manufacturer to the textile manufacturer. Figure 17.6 illustrates these supply chain configurations. The decision regarding where to mark should be made based on who has responsibility for fabric utilization and the point of control for order information. The objective of this type of supply chain reorientation is to respond quickly and efficiently to the wants of the customer. The supply chains detailed in Figs 17.5 and 17.6 provide a preliminary look at where direct digital printing might fit in. As more and more processes make use of digital technology, `islands of digitalization' are being created. The extent to which these `digital islands' are integrated will also drive the type of information systems needed to support a `totally digital supply chain'. It is becoming more obvious that batch size becomes less important as supply chains become more digital. In other words, the point at which a product in the supply chain moves from an analog process to a digital process is the point at which batch size or production quantity is no longer a constraint. Research has demonstrated how digital printing, along with other digital process technologies,

Digital printing and mass customization

309

supports this claim. This is also an indicator of the direction that digital printing is likely to take in the soft goods industry.

17.9 Future trends in the digital supply chain By now, it should be apparent that supply chain processes are incorporating an increasing amount of digital technology. Just as `islands of automation' were being created in US factories during the 1970s and 1980s, `digital islands' are being created in supply chains today. These digital islands are changing the rules of supply chain management and, in some cases, are causing a reorientation of supply chain processes. The movement towards mass customization that has been underway during the nineties will eventually give way to another supply chain strategy. This strategy will be called the `digital supply chain'. A digital supply chain is created when the process steps within the chain have been converted from analog functions to digital functions and the digital functions have been integrated into a continuous stream of product data. Since digital processes are readily reconfigured, they are not constrained by batch size. The result is a supply chain that is not constrained by batch size or production quantity. Mass customization technologies will play a significant role in the digital supply chain because they too are indifferent to batch size or production quantities. For example, body scanning is changing the way brands and retailers think about fit and the way they think about standard sizing. It is an enabler for digital product development for mass customization, but also has applications to the ready wear market. In the future, fit models for every size will be scanned and custom patterns will be generated for each of the fit models. The result will be set of patterns that more accurately map the variations in size from small to large. This will also allow the segmentation of a population, a brand, or a group of retail customers or into a set of standard sizes that more accurately represent the target market. In fact, the entire product development process will become digital. From yarn formation to fabric formation, to designing silhouettes to draping digital fabrics onto the silhouettes and applying color or print designs, the process will not require the conversion to physical samples and will be accomplished in 3D. Product Development Management (PDM) packages will contain patterns, cost sheets, bills of materials, manufacturing specifications, fabric specifications, quality standards, and in some cases, color specifications and printing instructions that can drive digital printing machines. The entire data set will be distributed and accessed worldwide. Simulations of human movement are beginning to have the appearance of near reality. It is not a big stretch to recognize that we will be able to scan a physical body (consumer) and from the scan data create a virtual, morphable body model and drape digital clothes on it. The digital clothes will be converted

310

Digital printing of textiles

to the physical clothes only when the consumer is satisfied with the fit, design and performance of the product. Garments will be designed in 3D using body scanning technology and automatic pattern generation. Color and print will be applied and converted using digital technology; however, the colorants will be produced in such a way that pre- and post-treatment of fabrics will not be required other than a heat set. The same set of colorants will work on silk, cotton or polyester and will have the same appearance and the same performance results. In other words, light fastness, color fastness and wash fastness will be equivalent to existing standards. This will augment existing digital printing technology such that solids can be produced in very small quantities, very efficiently and in environmentally acceptable ways. The same type of colorants that will revolutionize the coloring of solid fabrics will also allow sewing thread to be colored on demand. One can imagine a device that would unwind, color, and rewind a cone of thread without the need to vat dye in large quantities. It will also allow color decisions to be made much closer to the point of consumption. The technology of ink jet printing and the strategy of digital product supply are already in practice to some extent in the soft goods industry. As the chemistry evolves and the performance of the application technology improves, machine instructions will be generated at the point of design and distributed digitally during the order fulfillment process. The result will be a much more responsive supply chain that allows consumers to order customized products that will be distributed digitally and converted locally within very short time frames.

17.10 References and bibliography Anderson, P.A., Tushman, M.L. (1997). Managing through cycles of technological change. In P.A. Anderson, M.L. Tushman, Managing Strategic Innovation and Change ± A Collection of Readings (pp. 45±52). New York: Oxford University Press. Berger, S., Dertouzos, M.L., Lester, R.K., Solow, R.M., Thurow, L.C. (1991). Toward a new industrial America. In J. Henry, D. Walker (eds), Managing Innovation (pp. 288±305). London: Sage Publications. Betz, F. (1993). Strategic Technology Management (pp. 219±271). New York: McGrawHill. Bower, J.L., Christensen, C.M. (1997). Disruptive technologies: Catching the wave. In J.S. Brown (ed.), Seeing Differently: Insights on Innovation (pp. 123±140). Boston: The Harvard Business Review Book. Brown, J.S. (1997). Research that reinvents the corporation. In J.S. Brown (ed.), Seeing Differently: Insights on Innovation (pp. 203±219). Boston: The Harvard Business Review Book. Cetron, M.J. (1969). Technological Forecasting ± A Practical Approach. New York: Gordon and Breach. Clark, D. (1999, September). Applications of digital ink-jet printing on textiles. Printing

Digital printing and mass customization

311

2000: Entering the Jet Age, pp. 4±10. Charlotte, NC: AATCC Symposium. DAMA (1995). Process steps for men's cotton slacks. Demand Activated Manufacturing Architecture Project. Davis, S., Meyer, C. (1998). Blur ± the Speed of Change in the Connected Economy (pp. 6, 104, 249, 219). Reading, MA: Addison-Wesley. Fisher, J.C., Pry, R.H. (1972). A simple substitution model of technological change. Technological Forecasting and Social Change, Vol. 3, pp. 75±88. New York: American Elsevier. Goldman, S. (1997, November). Lecture about agility. Graduate Seminar, College of Textiles, NCSU. Goldman, S.L., Nagel, R.N., Preiss, K. (1995). Agile Competitors and Virtual Organizations, Strategies for Enriching the Customer. New York: Van Nostrand Reinhold. Handfield, R.B., Nichols, E.L., Jr (1999). Introduction to Supply Chain Management (pp. 7, 41, 42, 48, 159). Upper Saddle River, NJ: Prentice Hall. I.T. Strategies (1999, April). Worldwide Printer and Supplies Market Report. KSA (1997, August). Mass Customization: A Key Initiative of Quick Response. Report by Kurt Salmon Associates. Kuglin, F.A. (1998). Customer-Centered Supply Chain Management (pp. 11, 16, 69, 105, 111, 143, 150). New York: ANACOM. Martino, J.P. (1993). Technological Forecasting for Decision Making, 3rd edn. New York: McGraw-Hill. Nagel, R., Dove, R., Goldman, S., Preiss, K. (1991). 21st Century Manufacturing Enterprise Strategy, An Industry-Led View. Bethlehem, PA: Iacocca Institute. Peters, T. (1991). Thriving on chaos: Facing up to the need for revolution. In J. Henry, D. Walker (eds), Managing Innovation (pp. 306±312). London: Sage Publications. Pine, B.J. II (1993). Mass Customization: The New Frontier in Business Competition. Boston, MA: Harvard Business School Press. Pine, B.J. II, Victor, B., Boynton, A.C. (1993). Making mass customization work. Harvard Business Review, September±October, pp. 108±119. Souder, W.E., Sherman, J.D. (1994). Managing New Technology Development (pp. 141± 142). New York: McGraw-Hill. Tincher, W. (1999, 30 January). The latest on digital printing. Presentation at the 7th Annual NTC Forum. Tippett, B.G. (1999, September). Ink jet printing of textiles ± A practical perspective. Printing 2000: Entering the Jet Age (pp. 21±29). Charlotte, NC: AATCC Symposium. Wheelwright, S.C., Clark, K.B. (1995). Leading Product Development. New York: Free Press.

18 Just-in-time printing K M A G U I R E K I N G , [TC]2, USA

18.1 Introduction The printed textile market is a competitive arena that requires considerable risk and investment in time and inventory via the conventional screen print process. As the soft goods industry has become increasingly reliant on the production of smaller quantities and shorter manufacturing cycles, industry leaders have looked to alternative methods for creating product. Within this environment a digital approach to textile printing has been recognized as a viable option for producing printed textiles. As an emerging manufacturing method, digital textile printing has also been identified as a key process that supports the goal of `JustIn-Time' production within the sewn products industry. This chapter will examine the adoption of digital inkjet technology for just-in-time printing and will discuss significant technological developments and issues that support successful implementation within the manufacturing environment. Before launching into this discussion, it is important to have a general understanding of the `Just-In-Time' concept and the role of digital textile printing within this manufacturing strategy.

18.1.1 The just-in-time concept Just-in-time or `JIT' is familiar terminology within the sewn products industry and is generally applied to a supply chain scenario in which product is manufactured and delivered in a timely fashion and in direct response to market demands. While the term is often used in reference to manufacturing specifically, the successful implementation of JIT also relies on strategies that support supply chain visibility and management. These strategies involve data collection and analysis at the retail level and various points along the supply chain; communication of demand for the product and its specifications to manufacturing; communication of demand for raw materials and parts among supply chain partners; logistical control from raw material through product delivery; and product monitoring and quality control throughout the manufacturing process.

Just-in-time printing

313

In practical terms, in order to deliver the right product to the consumer `justin-time', the retailer must have knowledge of both current inventory and consumer demand. The retailer must be able to communicate their inventory needs to the manufacturer, and the manufacturer must have the ability to rapidly respond, providing the specified product at the agreed upon quality level. Throughout the supply chain, there must be a shared awareness of manufacturing capabilities and limitations and manufacturers must maximize their flexibility so that they can rapidly adjust to changing needs within the just-intime environment. As you ponder the concept of `manufacturing flexibility' it is important to remember that the sewn products supply chain is relatively complex. Fabric embellishment, including printing, is merely one step in a long list of processes that turn raw materials into finished fabric, garment, or fabric structure. In the past, this complex supply chain has largely been driven by manufacturing and within this scenario, the manufacturer analyzed trends and attempted to predict consumer desire for product. The manufacturer designed the product according to their trend information and offered it for sale to the retailer. The retailer selected from available styles and filled their stores accordingly, hoping that they purchased the right product, in the right color or print, in an appropriate quantity. With the introduction of `just-in-time', there has been a shift from a manufacturing driven economy to a demand driven economy. Within this environment, flexibility is paramount as the manufacturer must respond to, rather than dictate, product need at the retail level. With this in mind, we can begin to look at the significance of the digital print process and the role of digital technology for `printing on demand'.

18.1.2 Printing on demand in the digital environment The development of digital processes has been key to the implementation of `on demand' manufacturing methods, and as we begin to examine the adoption of the digital print process it's helpful to review Wantuck's (1989) explanation of JIT. He describes JIT as a production strategy and specifically notes the link between quality and productivity, indicating that `The JUST-IN-TIME Strategy includes seven principles which can guide us toward world-class productivity. It requires that we: 1. 2. 3. 4. 5. 6. 7.

Produce to exact customer demand Eliminate waste Produce one-at-a-time Achieve continuous improvement Respect people Allow for no contingencies Provide long-term emphasis.' (Wantuck, 1989, p. 11)

314

Digital printing of textiles

Although each of Wantuck's seven principles is significant, we can highlight the first three as we look at the digital textile printing process specifically. As compared to analog or `conventional' printing, the digital print process provides the opportunity to print very short lengths of cloth according to individual customer specifications and changing market demands. It reduces waste by minimizing design setup and eliminating costly and time consuming changeover for new designs or colorways. It allows for multiple product printing and very quick change of print, colorway, and/or design element. If we look to the remaining four principles, digital textile printing is still an emerging process that offers the opportunity for early adoption, growth and continuous improvement through experience and technological development. It respects people by harnessing the creativity of individuals and extending the potential for unique designs and printed effects. By eliminating waste and increasing production flexibility digital printing reduces the need for contingency plans including carrying excess inventory. Finally, the adoption of digital printing is a long-term strategy that can be implemented in stages with the ultimate goal of providing prints on demand. Many manufacturers have started their exploration of digital printing by adopting the technology for sampling and product development. As they gain experience and confidence in emerging solutions, they can begin to look at digital printing as a manufacturing tool. As we examine the manufacturing benefits I've noted, they are largely the result of the digital approach. In the digital world it is possible to create and hold design information in a form that is easily retrieved and altered. As order information is communicated to the manufacturer, prepared for print greige goods can be converted to printed cloth. Design information can be customized using CAD technology and precise lengths of fabric can be specified at the print station and produced according to individual order specifications. Orders can be batched at the printer for efficiency and ease of processing. With this background in mind, we can begin to examine how available technology and ongoing development will enable successful adoption by the soft goods industry.

18.2 Enabling the process In its infancy, digital textile printing was predominantly a sampling or prototyping technology. While implementation of digital printing for this purpose offers considerable benefits, there are great rewards to be reaped by organizations and individuals that can successfully implement a process for printed fabric production. Our industry has been aware of these benefits for some time. However, we have been slow to adopt the approach for production. What is the reason for this? The answer to this question can be found if we examine the nature of supporting technology and processes.

Just-in-time printing

315

18.2.1 Supporting technology, processes, and materials On the surface, supporting technology would appear to refer specifically to printer hardware and software. However, the development of printed textiles involves a range of technologies, processes, and materials that support a series of production stages including design and product development, printing, fabric preparation and finishing, cutting and product assembly. Product development The textile printing process begins with concept design and imagery development. During this stage of product development, designers research the marketplace and trend resources for design inspiration. They develop prints and color palettes based on their research, keeping in mind their target customer and end product requirements. This step in the process has been digitally driven for some time through the use of CAD technology. Hand-painted designs are photographed or scanned and edited by designers utilizing off-the-shelf and industry-specific graphics programs. As designers have become more fluent in the use of CAD and the technology has become more intuitive in nature, designers are increasingly developing artwork directly in the CAD system. The advent of digital photography has even enabled a truly digital approach as artists develop concepts directly from digital resources. While conventional print methods have also been supported through the adoption of CAD for product development, the digital print method is reliant on digital designs to feed the process. The color reduction and separation process that is required for the engraving of screens has become almost exclusively a digital process. Although color reduction and separation, as well as the preparation of pattern repeats, are often undertaken for digital prints as part of a sampling and/or color management strategy, they are not a requirement of the digital print method. Even so, the design information must be delivered to the inkjet printer in digital form. The design file is often prepared and supplied in such universal formats as .tif or .jpg files that hold the pattern and color information the printer software requires for image processing and color management. These files are then interpreted for output at the printer by specialized software programs for color management and raster image processing or `RIP'. The digital design environment provides a wealth of flexibility that is key to the JIT scenario. As long as the design remains digital, it is possible to make quick changes to pattern or coloration. Digital designs are easily stored or archived and retrieved on demand for production purposes. It is possible to harness such design archives, so that customers can review designs from previous lines or seasons and update or reorder as the market dictates. In the digital print environment it is also possible to create highly unique designs that illustrate special tonal (see Fig. 18.1) or photographic detail (see Figs 18.2 and

316

Digital printing of textiles

18.1 Dahlia scarf, provided courtesy of [TC]2.

18.3) and are almost unlimited in terms of color number and design length. There is even the potential for print engineering according to the shape of cut parts and to enhance overall product design. Printing There are currently a range of hardware, software, and ink chemistry solutions in the marketplace for digital textile printing. With respect to printer hardware, machines have been engineered by vendors for a variety of applications including both sampling and short-run production. Early machinery introductions were largely modified and/or re-engineered paper printing machines with specialized fabric handling mechanisms and textile specific software solutions. These machines operated at very modest speeds, typically printing well under 10 meters per hour in quality modes. This equipment was criticized as being too slow for production purposes and has predominantly been used for samples/ prototypes and for one-of-a-kind or very small-scale and fine art production. More recently introduced production-oriented equipment often features more robust print head technologies and belt fabric handling mechanisms. While these technological introductions are also engineered for vast improvements in print rate, currently available production printers continue to operate at relatively modest speeds as compared to rotary screen technology.

Just-in-time printing

18.2 Rhododendron scarf, provided courtesy of [TC]2.

18.3 Iris scarf, provided courtesy of [TC]2.

317

318

Digital printing of textiles

Even so, industry leaders continue to analyze the merits of the digital method for production and question the `real efficiency' of the rotary environment, taking into account practical rates for the operation of rotary machines and the amount of printer downtime required to change colors and align screens. This is noted in contrast to the high degree of production flexibility offered by the slower digital printer. While digital print rates have initially been a barrier to the implementation of JIT, there appears to be growing interest by the soft goods industry as technological advancement has occurred. This interest is highest in higher-end market areas and those that benefit from short runs, product customization and/or special effect printing such as ties, scarves, and swimwear. What print speed is required for successful implementation of digital printing for JIT production? There is more than one vision for production printing. Some hardware developers have focused their efforts on the refinement of smallerscale machines that can be used in multiples to meet production demands, much like the `weaving mill approach'. Other developers have focused their efforts toward the design of more robust, higher-speed machines that are accompanied by larger price tags. While potential adopters questioned the price and reliability of early introductions of production-scale equipment, more recent technological introductions have attracted greater interest. There is likely a place for both approaches as production requirements depend on a variety of variables including product type, market, and business model. Printing for high-end and high-fashion markets may take advantage of the substrate and ink chemistry flexibility offered by the use of many small-scale printers for production, while production for more mainstream markets and products may be suited to the efficiencies offered by larger-scale machinery. It appears that both approaches are developing simultaneously within the printing industry. Print rate has not been the only barrier to technology adoption for JIT. The inkjet environment requires highly purified and specially formulated colorants for reliable jetting. These colorants are typically more costly than colorants for conventional print methods and early introductions to the marketplace presented issues related to nozzle clogging and recovery. Print head engineering and compensation along with refinement of inkjet chemistry for textiles have gone a long way to address this issue. Drop-on-Demand (DOD) piezo print head technology dominates the current marketplace and has also been a focus for research and development. In contrast to thermal inkjet technology, the piezo approach allows for greater flexibility in terms of ink chemistry and does not involve heating the ink chemistry, which can be at the root of print nozzle failure. Despite advances, machines will continue to require monitoring for problems and fabric inspected for quality. As previously noted, colorants for digital textile printing are specially formulated for the inkjet process. Research and development has resulted in the development of reactive, acid, and disperse dyes for inkjet, as well as pigments. Dye-based coloring systems including reactive and acid have been the focus for

Just-in-time printing

319

early development. These colorants are highly water soluble and are more easily formulated for inkjet to obtain a range of brilliant master hues for process printing. However, the fabric pretreatment and wet finishing processes required to achieve optimal print results and enable color fixation of these dyes have been a deterrent for adoption. Even so, auxiliary equipment and service providers are now available to address these issues. Disperse dyes and pigments have been more challenging to formulate, as color chemists address issues including jettability, color fastness, and color brilliance for these chemistry types. As a result of research and development, there are now a number of vendors for these colorants in the marketplace today. The continued development of textile pigments for inkjet may provide some manufacturers with greater substrate flexibility and will ease fabric preparation and finishing requirements. Despite the advances in color chemistry for inkjet, there is still room for development of the `perfect ink chemistry'. This ideal solution would eliminate the need for fabric pretreatment and would enable integrated printing, color fixation, and cutting for in-line manufacturing. Although this `ideal solution' is not commercially available, digital printers are using available colorant chemistry to print a wide range of substrates for an endless variety of product types. Software development has also been essential to the successful implementation of digital textile printing at all levels. Unlike paper, which can be made uniform in terms of surface character and absorption, textile characteristics can vary widely. As a result, specialized software has been designed for this application to optimize color output for specified fabric, ink, and printer combinations. Software solutions can also be utilized to manage color separations and print quality for the sampling strategy. Software features may include the ability to create and manage color profiles for printing, specify a range of repeat arrangements, indicate print length and width, select designs and/or colorways for printing, specify color by screen through the use of a spectrophotometer or by entering a numerical color value, and control print quality to replicate the screen printed effect (e.g. color order, spread, fall-ons, etc.). This kind of software has become an integral part of the digital print system for textiles and cannot be undervalued. The importance of software and the idea of an integrated systems approach have been recognized by hardware and ink vendors and addressed through development partnerships and joint marketing relationships. It is essential to note that color is the key feature for any textile product. With this in mind it is essential to be able to obtain a wide range of hues within any printing or dying system. This is possible within the spot color screen printing process as each hue is pre-mixed from a master set of colors and applied in stencil form. However, the process color printing approach utilized in the inkjet environment is a completely different procedure. In the inkjet world, colors are `mixed on the fly' in the form of ink droplets that are selected from a master set. While a wide range of printed effects are possible as a result of this approach,

320

Digital printing of textiles

early systems were criticized for limitations in color gamut and the appearance of dither, particularly in the areas of pastel and neutral tones. Technology solution providers have addressed these concerns through print head development, color number, ink chemistry formulation, and software. As a result, currently available systems have considerable capabilities and are able to produce large color gamuts and print quality that rivals the spot color approach. Fabric preparation and finishing For optimal print results, fabrics for digital printing must be specially prepared for the process. As with any prepared for print (PFP) fabric, the greige cloth may be scoured and bleached depending on the fiber type. This step removes oils and impurities in the fibers and ensures a clean white surface for the application of color. Fabrics for inkjet are then pretreated. Fabric pretreatment serves two purposes. Pretreatment typically involves the application of a thickener to maintain print precision and prevent color wicking. It also involves the application of chemical components such as an alkali or acid that enable fixation for dye-based printing. In conventional printing, the thickening agent and auxiliaries can be found in the print paste. However, thickening agents, binders, and alkali or acid components may cause clogging or degradation of the print head and are therefore applied as pretreatments to the PFP cloth. The fabric preparation procedure is relatively involved and requires specialized equipment for even application and quality print results. Outside the mill environment, this step is often outsourced to service providers and can add considerable cost to the base fabric. In the JIT production printing environment, longer runs of cloth will be consumed as compared to sampling and pretreatment may take place as an in-house activity utilizing existing openwidth pad or screen systems. Alternatively, smaller-scale equipment has also been developed for pretreatment application in the digital plant. Quantity and inhouse processing will certainly have an impact on the overall cost structure for digital printing. However, in the fashion area product designers will look to print a wide range of fabrics and fiber types. This will continue to be a challenge to product developers and printers who will develop strategies for preparing, obtaining or stocking PFP fabrics for the inkjet process with the goal of providing quick turnaround, while minimizing inventory. With respect to fabric finishing, a variety of equipment has been developed for the digital print process. Batch steamers have been adopted for sampling and very small-scale production purposes. However, JIT production printing will be enabled by larger-scale fixation equipment. Where digital printing exists alongside conventional rotary setups, printers may utilize existing production resources for fixation and washing. Digital mills will rely on a new generation of machinery that has been developed specifically for the digital process. This equipment includes in-line and continuous steamers, open-width washing

Just-in-time printing

321

devices, and drying technology. These machines have smaller footprints and price tags than full-scale production equipment and may be more suited to the shorter runs and more modest print rates experienced in the digital environment. Cutting and product assembly A discussion of enabling technology and processes for JIT is not complete without addressing fabric cutting and assembly. Cut and sew is part of the manufacturing procedure for most printed textile products. With this in mind, the ideal manufacturing solution would minimize the processing and skill requirements for assembly and maximize flexibility. Flexible manufacturing is a key component within the JIT supply chain. JIT printing is of limited advantage, if the fabric moves slowly through the cut and sew procedure. Single-ply cutting technology and short cycle manufacturing methods are process enablers and are well developed for sewn products. They provide an alternative to more traditional multi-ply cutting and the progressive bundle method of product assembly that result in larger quantities of work in process and reduced manufacturing flexibility. The cut and sew process may be further enhanced for specific product types through engineered printing. In the digital print environment, the imagery may be designed to print only within the shape of the product piece and engineered for pattern matching (see Fig. 18.4). This strategy can allow for optimal fabric utilization and simplify any hand cutting requirements by providing an outline for the cutter to follow. Engineered printing may be of particular value in areas such as the custom upholstered furniture business. Hand cutting of upholstery fabrics is relatively common for custom orders and requires a considerable amount of personal training to ensure pattern matching of printed goods. Engineered printing may effectively de-skill this work and allow greater flexibility and cross-training within the cut and sew area for upholstered furniture (see Fig. 18.5). Despite technological advances in cutting technology, mechanical cutting of engineered prints continues to be a challenge. While cutting machines with cameras for piece recognition and positioning are in existence today, the wet post-processing required for color fixation and wash-off of currently available colorants results in dimensional instability and makes the development of in-line print/cut systems challenging.

18.2.2 Technology development and integration It is apparent that currently available technology and systems are the result of ongoing research and development efforts. When our industry first began its exploration of digital printing for textiles, development was required in diverse areas of specialty including hardware, software, and ink chemistry. Developers

322

Digital printing of textiles

18.4 Printing marker for apparel, provided courtesy of [TC]2.

quickly learned that their research could not be undertaken in isolation and that inkjet printing for textiles required an integrated systems approach. The result of this thinking was the formation of development partnerships and joint marketing efforts. The introduction of the DuPont Artistri and the Reggiani Dream systems are two examples of joint efforts that have brought together expertise in areas including ink chemistry, print head technology, fabric handling, and color

18.5 A portion of a marker for upholstery, provided courtesy of [TC]2.

Just-in-time printing

323

management. The marketing partnership between Stork Digital and CAD technology provider Lectra Systems is another example of a systems approach for digital textile printing. As technology systems have developed for JIT printing, there is increasing recognition of the need to develop more integrated manufacturing solutions. As noted in the previous sections, printing is only one step in the production of a printed textile. Fabric preparation and finishing are also required in order to obtain prints with appropriate washfastness and serviceability characteristics. Development of a new generation of equipment that addresses the shorter runs and flexibility requirements of the digital method is key to the successful implementation of the JIT approach. There is growing acknowledgement of the need for in-line and modular preparation and finishing systems that can be arranged and adapted according to specific ink chemistries and the specialized needs of a given manufacturing environment. The development of innovative ink chemistries that eliminate the need for wet post-processing may even allow for the development of in-line cutting, the benefits of which were noted in the previous section. In addition to technology integration in the area of printers and auxiliary equipment, there has also been the development of integrated color management systems and strategies that assist the printer in reproducing color palettes specified during the product development stage. The term `color management' does not refer to a single activity, but rather a process that involves identification of color targets or palettes and the accurate communication and reproduction of this specification through the manufacturing supply chain. Color management strategies are supported by software solutions and color measurement hardware that enable the profiling of systems and transmission of numerical color data to aid the printer with accurate color reproduction as specified by the product developer or customer. The ability to predict and control color output is essential to the just-in-time model as manufacturers are required to `get it right the first time'. It is apparent that technology development and integration is occurring simultaneously in a variety of areas in support of digital textile printing. As this development has taken place, conditions are becoming increasingly suited to the implementation of a just-in-time print strategy. With enabling technology in place, we can begin to discuss what a JIT business might look like.

18.3 Just-in-time order processing The particulars of JIT order processing may differ depending on a number of variables including the product type and business model. No matter the specifics, a number of areas must be addressed for successful implementation to occur, including the capture of order information, management of designs and image data, fabric preparation, printing, and finishing, and order delivery.

324

Digital printing of textiles

18.3.1 Capturing order information The mechanism for capturing order information will vary tremendously depending on the nature of the business being served by a JIT print model. In many instances JIT printing will involve the manufacture of a sewn product for apparel, home, or accessory item. In this environment, the manufacturer may be responsible for a variety of steps including fabric development, cut and sew, and even the sale of the product to the final consumer. Alternatively, manufacturing may include a variety of supply chain partners, each responsible for a step in the manufacturing process. Depending on the manufacturing arrangement, the digital printer's `customer' may refer either to the end consumer or to the retailer/product developer. Whoever the `customer', it is important to remember that within a JIT strategy, the submission of an order to manufacturing is demand driven. In this `pull system' the demand for product will ultimately be created by product sales, and these sales may be obtained in a variety of ways including through traditional bricks and mortar retailing, or via catalog and on-line sales. However the product is offered for sale to the consumer, the manufacturer will likely receive just-in-time order information in electronic form in order to speed the process. Order information will outline both quantity and product specifications, including product style, size, print design, and colorway. In the fashion apparel and accessories markets, retailers are continuously striving to fill their stores and catalogs with new product and fresh seasonal looks. In this setting, product development is constant as design teams work close to delivery and press for shorter manufacturing cycles. The need for quick turns of smaller quantities of new product suggests that product development must coordinate with manufacturing to ensure just-in-time delivery and address any issues that arise during the manufacturing process. Digital printing is well suited to meeting these order demands as it offers tremendous manufacturing flexibility. In addition to quick turns of new product, there may also be a need for ongoing replenishment of basic products or as a result of unanticipated sales of seasonal items. The tracking of product inventory and sales at retail will likely provide the basis for determining these replenishment needs and will trigger order placement. With respect to online sales, manufacturers may receive their order information directly from the end consumer. Via this setting, consumers may place single or multiple unit orders for standardized product. The e-commerce setting is also well suited to a mass customization business model as it allows customers the opportunity to shop leisurely and view a wide range of products and styles within the privacy of their own home. In this setting it is possible to experiment with a variety of color options, prints, and style features and personalize their products by selecting from lists of customizable features. As 3D visualization technology advances, consumers will have increasing oppor-

Just-in-time printing

325

tunity to view their customized designs as finished product and even `try it on' their personal model, in the case of apparel. JIT is a manufacturing strategy that supports the implementation of both customized and standardized product orders. Keeping in mind the desire to reduce risk at retail and eliminate or reduce fabric inventory at the cut and sew level, digital textile printing is a favorable method for filling both single and multiple unit orders within the JIT setting.

18.3.2 Design and image management The management of design information is key to the successful implementation of a JIT digital print model. This design information must be ready to print if product is to be delivered in a timely fashion. With this in mind, the strategy for creating prepared-for-print (PFP) designs is somewhat dependent on the nature of the final product and on the specific manufacturing model being served. In general, it is possible to identify two main strategies for image preparation. The first involves the preparation of designs that are color reduced and separated as they would be for the conventional screen print process. These designs are typically created and/or edited using textile-specific software programs. Although the digital print process does not require a color reduced and separated image file, there may be a number of reasons to undertake this process. At the product development level, designers may be working from seasonal palettes that act as a unifying feature or color story within the retail environment. The process of specifying and integrating color from a palette is simplified for an image that is color reduced. Color separated images are also easily recolored and adapted in order to offer multiple colorways, coordinates, and the opportunity for product customization. With respect to production, the digital process may also be used to print short runs of fabric that are introduced into the marketplace as test runs. If the design is popular, it may be worth producing longer runs via screen printing. In this case, the design must be prepared for screen printing and the digital print must be representative of the larger screen print production run. The second strategy for image preparation involves the creation of designs that will only ever be produced using the digital print method. These designs may be created using textile-specific or off-the-shelf software and may take advantage of the wealth of design possibilities digital printing offers, including photo-realism, sophisticated tonal and textural effects, engineered printing, elimination of repeats, and unlimited image scale. As previously noted, it is even possible to combine surface design information with pattern information for the creation of printed fabric markers. Whatever the method for image preparation, the design file must be prepared and stored in digital format for the design information to be quickly converted into printed textile within the JIT setting. Once the design is in digital format it can be stored in a library or design archive for easy retrieval. This archive can be

326

Digital printing of textiles

used at the product development level for design inspiration and artists can access designs and repurpose them for new lines and seasons. Some businesses may also select to utilize such design libraries as an asset that can be offered to the customer as a customization resource.

18.3.3 Fabric preparation, printing, and finishing Once the order has been placed and the design has been created, selected, and/or customized, the design file can be placed in a queue for printing. In the JIT environment the digital printer must have access to specially pretreated greige fabrics for specific products and ink types. The breadth of this selection will depend on the manufacturing setup and the nature and range of customers being served. In the fashion apparel market, printers may be required to work closely with product development to identify and cultivate sources for a wide range of PFP fabrics that can be quickly delivered and converted as needed into pretreated goods for digital printing. Alternatively, printers may select to specialize and stock a selection of pretreated fabrics for the range of customers they serve. In either case, the printer must have an effective working relationship with the fabric mill(s) to ensure ready access to raw materials and, depending on the capabilities of the printer, fabric pretreatment may be outsourced or completed in-house. At this point it makes sense to batch orders for efficiency. The rules for batching orders will depend on the specifics of the printing and finishing setup. For example, some printers may offer printing with multiple ink types on a number of smaller-scale machines. In this instance, orders may be batched by customer and also by ink type and fabric and then sent to corresponding printers. Alternatively, the printer may operate one or more large-scale machines with one or two ink and fabric types. In this case, orders will likely be batched according to customer and fabric in order to minimize fabric changeover. The requirements for color fixation and wash-off will be driven by the ink chemistry. However the setup will depend on printing capacity and the number and type of printing machines. For example, the printing of short runs of a variety of fabrics using multiple small-scale printers may lend itself to the installation of one or more small-scale, open-width finishing units for production flexibility. In contrast, the printing of longer runs and/or a narrow range of fabric and ink types may lend itself to the use of in-line equipment and/or larger-scale open-width finishing capabilities. Some printers may select to establish finishing in-house, while others may outsource these capabilities. The outsourcing of fabric preparation and finishing may be a particularly attractive option where the digital printer provides other services (e.g. screen engraving or cut and sew operations) and has existing relationships with wet processing mills. Outsourcing of finishing operations may also be an option where applications of specialized chemical or mechanical finishes are required. It is important to

Just-in-time printing

327

keep in mind that JIT is about speed of delivery and so the cost benefit of outsourcing must be carefully analyzed to determine the route that provides the greatest value to all parties including manufacturer, retailer, and final consumer.

18.3.4 Order delivery Depending on the customer the finished goods may take the form of either digitally printed yardage or digitally printed sewn product. If the printer is responding to an order for digitally printed yard goods, they will deliver this product in roll form to the customer or to another supply chain partner that will cut and assemble the fabric into apparel, drapery, upholstered furniture or other textile product. However, the printer may also be the cut and sew manufacturer. In this case, the fabric will move directly from finishing into cutting. As technology and ink chemistry developments occur, it is plausible that the cutting operation may be placed as an in-line activity to printing. In the case of sublimation printing, it is currently possible to transfer the digitally printed image to cut pieces rather than a roll of fabric. Once the fabric is printed and cut, the pieces are moved out onto the floor for final product assembly. In order to adapt to quick changes in print, color, and style, sewing operations and other methods involved in assembly may be set up according to modules operated by teams of cross-trained specialists for shortcycle manufacturing. In the case of customized product, the cut parts must be identified as belonging to a single unit and must travel together from cutting through the assembly process. Currently available technology such as the Eton Unit Production System (UPS) can be used to convey and track parts for a single unit through the assembly process. In theory, order number and part identifiers could also be applied during the digital print procedure as another method to ensure that the parts for a custom product remain together. Once the product is assembled, it is ready for packaging and shipment to the retail outlet or final consumer.

18.4 Case studies The previous sections of this chapter have outlined a vision for just-in-time digital printing. How is the vision implemented in a real-life setting? This section will briefly examine the JIT printing activities of [TC]2's InkDrop Boutique and the Stork U See digital printing service.

18.4.1 The InkDrop Boutique The InkDrop Boutique is a small-scale digital textile printing service developed by the not-for-profit corporation [TC]2 to research and demonstrate the capabilities of digital textile printing for the sewn products industry. As of May 2005,

328

Digital printing of textiles

18.6 The InkDrop Boutique digital print area, provided courtesy of [TC]2.

the InkDrop team makes use of sampling-level digital printing equipment including four printers from Stork, Mimaki and MacDermid Colorspan, along with a batch steamer from Jacquard Products, to produce small quantities of digitally printed sewn product (see Fig. 18.6). They are supported by [TC]2's team of product development and sewing specialists (see Figs 18.7 and 18.8) who help them to manufacture a selection of customizable products for a range of customers including individual artists, designers, and museum stores. The goal of the InkDrop Boutique is to utilize their modest manufacturing resources to provide printing on demand and help their customers and visitors to learn about the benefits of digital printing as part of a just-in-time approach to manufacturing. InkDrop currently specializes in reactive dye printing and stocks a small selection of fabrics that are pretreated for that purpose. Their customers provide imagery in digital format, typically as a .tif file saved for the Windows platform. The design area currently uses Adobe PhotoShop as their main product development tool for the integration of imagery into standardized product templates for items including scarves, purses, totes, accessory cases and cushions (see Fig. 18.9). The use of digital product templates that contain cut and sew guidelines simplifies the product development process and helps to enable sample creation and speed to market (see Fig. 18.10). With respect to production, customers were initially offered minimum order quantities of 10 pieces. However, it eventually became evident that most customers could benefit from even smaller order quantities. With this in mind, minimum order quantities are currently determined in relation to the number of

Just-in-time printing

329

18.7 Making patterns and cut files for InkDrop Boutique products, provided courtesy of [TC]2.

18.8 Sewing InkDrop Boutique products, provided courtesy of [TC]2.

330

Digital printing of textiles

18.9 A sample of InkDrop Boutique products, provided courtesy of [TC]2.

18.10 InkDrop Boutique design template for tote pockets and purses, provided courtesy of [TC]2.

Just-in-time printing

331

pieces that can be printed across a given fabric's width. For most product types, customers can also obtain multiple designs or colorways within a single fabric width. This strategy eliminates fabric waste and customers are able to obtain very small numbers that can be replenished on demand. This environment also makes it possible to experiment with a variety of designs and obtain a selection of unique products with minimal investment in setup time and inventory (see Fig. 18.11). Repeatability and accuracy of color reproduction is one of the great challenges for digital printing in a production setting. The Inkdrop Boutique uses a selection of color management tools to assist them with this task, including software from Stork, MacDermid Colorspan, DPInnovations, and Ergosoft. They are well versed in the creation of color profiles for specific fabric, ink, and printer combinations and typically create and apply profiles for groups of fabrics depending on similarity of weight and fiber type. Most of the artwork received is full color imagery that can only be replicated via the inkjet environment. They work closely with customers to ensure that color is reproduced according to their specific tolerances. Among their customer base, artists and designers are typically satisfied with a first print approach in which they obtain a printed sample for approval for which there has been no color adjustment on the original file. They have found this to be the quickest and least costly approach for new product development. Despite the use of color profiles for printing, museums often require greater color accuracy, and in these instances the InkDrop team will use an iterative approach to make color adjustments on the original file to reduce color casts and ensure accurate reproduction. Customers are asked to provide a printed paper version of their image to help guide the color correction process. As this strategy is more costly in terms of fabric, ink, and labor, an additional fee is charged for the service. However, the color-adjusted print file and a printed sample of the color-proofed image is kept for processing replenishment orders. Customer order sheets are used to digitally record manufacturing information related to both printing and product assembly, and the imagery and digital order information are easily retrieved for quick and accurate replenishment on demand. While most InkDrop customers develop a range of standardized products for their marketplace, some customers utilize the JIT digital print environment to support a mass customization business model. In this instance, product development and manufacturing strategies must be streamlined for successful implementation. The ultimate goal is to develop a `first print' strategy and to eliminate or minimize color correction and image preparation. From a product development standpoint, the creation of standardized product templates with predetermined interchangeable print features is very helpful. Product developers can maximize the value of their digital assets by creating image libraries and developing design options from archived imagery data. Within this scenario image layout and color reproduction can be predetermined and therefore,

18.11 Example of a printing marker corresponding to fabric width for tote pockets, provided courtesy of [TC]2.

Just-in-time printing

333

predicted and controlled to ensure that customers receive the product they envision without color or image adjustments and costly reprints. In some mass customization scenarios consumers supply the digital imagery to be printed (e.g. a photograph). The InkDrop staff has found this mass customization scenario the most challenging to implement as it is very difficult to control the quality of imagery supplied and to predict and control the color output according to customer and consumer preference. In this situation, they have found it is important to manage expectations and to provide `guidelines for imagery preparation' and `tips for best print results' to ease the process. It is also particularly imperative to optimize fabric preparation, printing and finishing conditions and to minimize variables that impact image quality for efficiency and predictability. Through their experiments with mass customization, the InkDrop staff have learned that personalized product can have great value to the final consumer and, although print strategies are not without pitfalls, the InkDrop team will continue to investigate and refine their process. While their manufacturing capabilities are modest, the InkDrop Boutique service has been of great benefit to their growing customer base that take full advantage of the just-in-time manufacturing flexibility that digital printing enables. It has also been a useful resource for the US sewn products and imaging industries as they investigate digital textile printing and its potential for the creation of unique products and businesses.

18.4.2 Stork U SeeÕ The Stork U SeeÕ standardization is an effort to further the adoption of digital textile printing by providing regional service centers for sampling and short-run production that enable color management from design concept through finished product. They currently operate four centers worldwide, with a facility in Thailand that offers production in addition to digital sampling. Stork U SeeÕ is described by the vendor as a `color communication quality standard' and their strategy for printing emphasizes four key areas including software, hardware, consumables and best practices. Stork U SeeÕ centers are certified to ensure interchangeability between print machines and locations for consistency of print results. The facility in Thailand houses a range of printing technology to meet sampling and production demands. As of May 2005, the plant is equipped with 18 printers including 12 Stork Sapphire and six Zircon II machines (see Figs 18.12, 18.13, and 18.14). The facility operates 24 hours per day, seven days per week, and runs two 12-hour shifts in an environmentally controlled building. The manufacturing environment is very clean and offers a comfortable work environment, quite unlike most conventional wet printing facilities that are often warm and humid. The printing area is staffed by six operators who oversee three to six machines during their shift, and Stork is able to offer printing on a range of fabric types including synthetic and natural fiber substrates. Machines are

334

Digital printing of textiles

18.12 Digital textile printing at the Stork U SeeÕ center in Thailand, provided courtesy of Stork Digital Imaging.

18.13 Digital textile printing at the Stork U SeeÕ center in Thailand ± closeup, provided courtesy of Stork Digital Imaging.

Just-in-time printing

335

18.14 Fabric inspection at Stork U SeeÕ center in Thailand, provided courtesy of Stork Digital Imaging.

equipped with closed, bulk ink systems for reliability and efficiency, and Stork is currently building a refrigerated room within the facility to ensure the ink supply is maintained at suitable temperatures. This print environment has been operating at production levels for three years and currently receives orders for runs up to 30,000 yards. In order to meet production demands some of the machines may be set up with more than one ink set for flexibility, while others are arranged with a single colorant type for speed. Open-width steaming equipment from Rimslow is utilized for fixation of samples and larger runs are generally sent to a local facility for post-processing on production-scale equipment. Stork's proprietary systems for color management are combined with workflow practices to ensure accurate reproduction of color and design according to customer specifications. While the InkDrop Boutique focuses their efforts on serving smaller-scale businesses, Stork has greater production capabilities and their customers include leading European and US fashion and swimwear retailers and product developers. Stork indicates that these customers are able to utilize the digital method to introduce unique product and are also able to reduce the risk of introducing prints into the marketplace through a continuous flow pipeline of JIT product and replenishment on demand.

336

Digital printing of textiles

18.5 Conclusion In Hall's (1987) discussion of manufacturing excellence, he notes that technological change isn't enough in and of itself. Manufacturing excellence is driven by changes in ` engineering, business, and people' (Hall, 1987, p. 14). In a sense, we must change the way we do business and provide employees with technology, engineering, and training in order to effectively implement a JIT strategy. Hall echoes Wantuck's (1989) belief that a new approach to manufacturing begins with the elimination of waste, the reduction in lead times and cost, and an investment in people, quality, and continuous improvement. The implementation of a digital printing strategy involves attention to each of these factors. The benefits will be evident to those organizations that value innovation in design and production. While implementation of digital printing for just-in-time manufacturing is not without challenge, the rewards may be tremendous. When combined with additional strategies that support short-cycle manufacturing and supply chain management, the benefits may be enhanced for all parties including manufacturer, retailer, and consumer.

18.6 References Hall, Robert W., 1987, Attaining Manufacturing Excellence, New York: McGraw-Hill. Wantuck, Kenneth A., 1989, Just In Time for America, Southfield, Michigan: KWA Media.

19 Design and workflow in digital inkjet printing H U J I I E , Philadelphia University, USA

19.1 Introduction The history of digital inkjet printing goes back to the nineteenth century, when the English physicist Lord Rayleigh investigated the physics of inkjet technology (Rayleigh, 1878). Beginning in the twentieth century, the first commercial non-impact digital printing started in the 1970s, for the carpet printing industry. In the United States, Milliken's Millitron system, and in Austria, Zimmer's Chromojet system, both became the standard method of modern production. The Millitron system, equipped with 10±20 jets per inch, utilized the computer injection dyeing system, in which the continuous streams of colorants are controlled with deflection by air jets. The Chromojet system is based on the drop-on-demand solenoid valve principle and the computer-controlled valves eject the colorants directly to the substrates. These printers are designed specifically for printing broadloom carpet and tiles, and printing quality is as good as 10±20 mesh of conventional printing technology (Dawson, 2003). The true commencement of commercial digital textile printing was not actualized until the late 1980s, in which sampling and digital strike-offs were implemented. Although many refinements and improvements in existing printing technology have been seen in the past 20 years (Dawson and Hawkyard, 2000), until recently there have been no large-scale implementations of industrial printing technology. It was not until the 1990s that digital printing technology grew in scope, in terms of applications and price points: from simple consumer desktops, through large formats to industrial applications (Pond, 2000) Since 2000, innovation and growth of new digital printing technologies has led to further implementation in the textile printing industry. Notably, several industrial production digital inkjet printers were introduced at ATME in Greenville, South Carolina, USA, in 2001, and at ITMA in Birmingham, UK, in 2003. At the outset of 2004, more than 100 units of industrial production digital inkjet printers were placed worldwide (Compton, 2004), securing diffusion and replication of emerging industrial expansion. Generally speaking, the evolution of technology is defined as consisting of three stages: (1) invention as discovery,

338

Digital printing of textiles

(2) innovation as the first commercial application, and (3) diffusion as widespread replication and growth (Grubler, 1998). If one considers the integration of digital printing technology in the carpet industry as representative of the innovation stage in the evolution of textile technology, it is about time that industrial expansion occurred. Digital printing technology has led to new opportunities in emerging textile design styles, which conventional industrial textile printing methods are unable to accomplish. The development of digital textile technology has influenced the workflow of manufactured printed textiles. Digital textile design has generated numerous new image-making possibilities, without the limited numbers of colors, or screen sizes of conventional printing. Digital textile printing expands the concept of design aesthetics by its ability to allow for more creative outcomes in the domain of textile printing. Moreover, this textile technology functions as a new conceptual framework for production and distribution. Changing attitudes have already begun to liberate the textile industry to become more cost-effective from the initial design concept to final distribution.

19.2 Evolution of textile printing workflow Historically, the manufacture of printed textile design requires a laborious timetable of diverse processes, and specialized professionals, from the creation of the first design to the final stage of production. This system has remained relatively unchanged since the industrial revolution. Today, in contrast, the diffusion of computer technology provides us with convenient access to production information, which ultimately has deconstructed existing paradigms, and in turn has generated new attitudes towards design and industry production. The main principle behind new textile industry production is the term `digital', in which all information processes can be controlled by digital computer technology. Most significantly, two new digital workflows, of digital strike-off printing and full digital textile production, have consequently led to a more streamlined and effective production environment.

19.2.1 Conventional textile printing workflow In the current textile printing industry, flatbed screen-printing and rotary screenprinting are the dominant production technologies and share more than 80% of the total printing mechanism. Furthermore, rotary screen-printing accounts for nearly 65% of total printing mechanisms (Stork, 2002). The workflow for these traditional textile-printing technologies consists of several segmented processes. Prior to introduction into the consumer market, final bulk production was developed through the following processes: (1) design concept and development, (2) design modification, and (3) engraving and strike-off.

Design and workflow in digital inkjet printing

339

Firstly, the majority of independent freelance textile designers sell their print designs to textile manufacturers. Although so-called in-house textile designers still exclusively design in manufacturing companies, there has been a shift towards purchasing freelance design work. This is in part due to the fact that companies have more diverse and creative design choices from various freelancers than from their own limited number of in-house designers. It is also due to the fact that existing specialized textile CAD software allows for immediate technical translation of textile designs into standard repeat and color reduction. Although CAD textile software has become an effective technical support system, the majority of textile designs are still created by the traditional hands-on methods on paper. Designers are responsible for visually indicating a limited number of spot colors and repeat sizes in their designs, which are based on these production methods. Once the jobbers, converters, and manufacturers acquire the initial designs, the design modification process begins. By using CAD systems, the designs are inevitably edited and translated into a limited number of spot colors and repeat sizes for a particular printing production method. Production technology determines the pricing and quality of final printed textile products. After the editing tasks, textile print designs are sent to engravers for creating screens. In engraving and strike-off processes, engravers use CAD systems to separate and convert textile designs into spot color separations. This is one of the most critical processes, in order to maintain the aesthetic integrity of the original designs. In addition to the CAD-operated color separation methods, reputable engravers still depend on manual skills which they have acquired through years of experience. In general, these color separations are stored as CAD data, and translated into separate films, which depict opaque black motifs, that represent each spot color for printing. Traditionally, these films are placed on the surface of screen meshes, which are pre-coated with photosensitive polymers and exposed to UV light, and eventually unexposed areas are washed away. Increasingly, digital printing technology has contributed to the implementation of direct filmless processes. The separation data from CAD systems applies opaque black ink or molten black wax directly onto the unexposed pre-coated polymer screen surface by inkjet printing to achieve as fine an image quality as 720 to 1019 dpi. After the screens are exposed to UV light, the ink and wax are removed (Moser, 2003; Int. Dyer, 1999a). Direct laser engraving technology has also become a viable process, where a carbon dioxide laser beam controlled by a CAD system removes positive images out of pre-coated polymer on the screen surface to achieve more than 2000 dpi. After the engraving process is completed and approved, the printing is executed at the print mill for sampling and proofing. Strike-offs are often sent back and forth between the engravers and design studios for approval. For the final production, print stylists or design mangers frequently go to the printing mills to approve the final strike-off before the bulk production is

340

Digital printing of textiles

19.1 Conventional textile printing workflow.

executed. Millwork is a lengthy and laborious process for the stylists or design managers, who often work (along with the mill technicians) on 24-hour shifts. This process requires an excellent knowledge of conventional printing methods, including a proficient understanding of matching the color, printing technology, repeat, and eyes to retain the original design aesthetics. This workflow is still the mainstream of the textile industry today and it is highly labor intensive and timeconsuming. A thorough knowledge of conventional textile printing technology is instrumental in the qualitative outcome of manufactured cloth. Informed professional design decisions are based on disciplined skills, and more importantly, from well-rounded professional work experience. The workflow is summarized in Fig. 19.1.

19.2.2 Digital textile strike-off printing workflow A new digital strike-off workflow has been integrated into the conventional printing and marketing process. Instead of using conventional printing techniques to create strike-offs, in which engraving processes are needed, manufacturers have begun to utilize large format digital printers to create digitally printed strike-offs. In this workflow process, manufacturers can use digital strike-offs for market testing and photography shoots, without going through the conventional engraving screen process. Thus, only the marketable designs proceed to be engraved and produced by conventional methods. In contrast, the conventional strike-off process in textile printing can take as long as 6±18 weeks (Spruijt, 1991), with a cost factor of $90±$110 for sampling as well as $260± $415 per screen creation (Int. Dyer, 1999b). In general, approximately only 50% of textile designs that are engraved go into print production yardage (Clark, 2003). By comparison, the digital strike-off workflow process can be reduced to 1±3 weeks and obviously eliminates extra engraving costs, which can sometimes save companies millions of dollars per year (Chapman, 2002). For the most part, this is the most dominant use of digital textile printing today and has continued to gain popularity. It is summarized in Fig. 19.2. In addition to strike-offs, short-run sample printing has become another popular outcome of digital inkjet printing technology. This is demonstrated in the domestic bedding and fashion industries, which both assemble printed samples for presentation at trade shows. A buyer can more easily be persuaded to come to a decision if dress samples and bedding ensembles are available for viewing. As soon as business is established, and orders are approved, the digital

Design and workflow in digital inkjet printing

341

19.2 Digital textile strike-off printing workflow.

design samples are sent to the engravers and mills for conventional production printing. The main objective of the digital strike-off workflow is that digitally printed strike-offs should accurately represent the final printing production quality and be producible in conventional printing methods. In order to achieve these goals, computer software plays an important role in analyzing the parameters of engraving information, drop formation, dithering information, and colorant reflectance data of both the digital printer and the color kitchen production environment. Sophisticated software also analyzes complex characterizations of fall-ons and mixes of the conventional textile wet printing process into its algorithm. By implementing color management systems, color accuracy and consistency can be achieved within an input, a monitor, and the printer's output in a designer's CAD station. In this way, textile design can be rapidly altered and quickly respond to the market needs before the final bulk production. This workflow process is considered most advantageous in locations where strike-off and sampling are on the same premises as designing and editing. Therefore, unlike technicians at mills where conventional printing takes place, designers can function as their own quality control technicians. The designers will be responsible for quick market trends and operating and maintaining their own digital printers in the design studios. Unlike conventional printing, the digital printers do not occupy a lot of space and are environmentally friendly. Digital printers can also be extremely efficient if they are located inside major cities, where the textile design markets are centred. In this way, it becomes highly effective to establish quality communication between printing mills and buyers, instead of at specific remote printing mill sites (Henry, 2004). It is foreseeable that in the near future, once digital production has a lower cost factor and higher printing speed, this printing and marketing process will have wider ramifications in full-scale bulk production.

19.2.3 Full digital textile printing production workflow ITMA 2003 in Birmingham, UK, has become a benchmark for starting an era of full digital textile printing production. Several robust industrial digital printers were introduced, including the DReAM machine by Reggiani Macchine, the Monna Lisa by Robustelli, the Artistri 2020 by DuPont, the TX-3 by Mimaki, and so on, and these are economically competitive to operate as fast as 150 square meters per hour for printing lengths up to 1000 square meters (Glover,

342

Digital printing of textiles

2004; Moser, 2003). The workflow of full digital textile printing is streamlined without having extra engraving tasks, and all stages of textile design production, from designing, editing and strike-off to production, can be executed under the CAD environment. Information technology includes digital data and asset management as well as electronic communications. It is also adaptable to current market demands of rapid trend cycles and short-run printing lengths with a quick and reliable response. Production quality digital textile printers allow for more original designs in a short to medium run production capacity, which is more economical than conventional printing production. This new workflow system is summarized in Fig. 19.3. This type of printing process has more flexibility in terms of style and mechanics. A novel design, such as a photographic image or a tonal image, can be articulated without mechanical limitations on the number of color screens or repeat sizes. In addition, the aesthetic image quality of textile designs which are processed through spot color separations in the conventional printing process can be enhanced by the full digital printing production workflow. Instead of translating design to the conventional two-bit raster-based screen separations, each color separation can be printed in the eight-bit tonal separations to retain more precise image integrity. These differ dramatically from the conventional printing methods, and offer a competitive edge for textile printers. In the current textile printing industry, it is apparent that there is a continuous shift of printing production from North America and Europe to other parts of the world where wages are lower. Asian countries, particularly China, share more than 50% of total printing production worldwide (Stork, 2002). Presently, several printed textile manufacturers in Europe and North America demonstrate full digital textile production workflow for their exclusive high-end fashion and furnishing markets. Integration of full digital textile printing production became one of their survival strategies in terms of differentiation from their peers (Bruni, 2004). On 1 January 2005, the quota system for the textile and clothing sector came to an end. In addition to other sectors of textile and clothing industries, the textile-printing sector will also be more competitive worldwide (Adiga, 2004). Textile printing operations in China are currently considered to be the largest and most successful, and eventually China will become the capital of textile

19.3 Full digital textile printing production workflow.

Design and workflow in digital inkjet printing

343

printing. The future of textile printing in China depends on finding the competitive niche in full digital printing production in order to ensure success.

19.3 New design styles In general, the assessment of printed textile design production is judged by its commercial viability and original creativity. Commercially successful printed textile designs can be sold continuously for many years. It is not rare for some production yardages to be continuously printed for over 20 years for greater commercial success. On the contrary, favorable value assessments of originality and creativity often result from publicity created by those who affiliate with journals, museums, academia, and publishers (Schoeser, 1986). To a certain degree, many manufacturers fulfill both parameters of good sales and reputable publicity for the success of their business. Although some printed textile designs accommodate both successful sales and well-received PR within one design, most manufacturers prepare two different marketing strategies to deal with these goals. Manufacturers, for example, spend time producing commercially successful textile designs that come from modifications of preconceived traditional textile design trends. These traditional designs are easily modified and processed into production yardage, utilizing previously mentioned digital printing technology. However, time and effort is also spent producing creative and original textile print designs, which are more characteristic of newer `contemporary' design aesthetics. The success of creative and original designs results from the novelty of the design aesthetic itself, which is largely determined by market trends and textile printing mechanical constraints (Eckert, 1997; Moxey, 1998). The trends of the market in a given time period are established by the symbiosis of political, economic and cultural factors in our society. The designers' inspirations are formed by their intuitions, visual stimuli, and research tasks influenced by a series of the symbiotic social phenomena including demographics, behaviors of consumers, pop culture, media, etc. (Wilson, 2001). Nonetheless, new design styles emerging from digital textile printing are influenced far more directly by the actual printing mechanics.

19.3.1 Textile printing mechanics and design styles Historically, mechanical methods of textile printing have defined the textile design styles. The history of design styles can refer to developments of the visual interpretation of three-dimensional reality into two-dimensional renderings. To illustrate, beginning more than 2000 years ago, block printing technology was the earliest example of three-dimensional simulation. In block printing, every block represents each color separation, and consists mainly of flat silhouette shapes. A three-dimensional effect of the motif is obtained by printing several separate layers of flat silhouette shapes to create light to dark

344

Digital printing of textiles

tonal definitions. The so-called `traditional floral design style' printed by screen printing technology today still retains the same original look as the historical block-printed designs. Looking back historically, engraved plate and roller printings, which originated in the eighteenth century, evolved into more sophisticated two-dimensional renderings of three-dimensional motifs. These detailed printing techniques represent the finest line and dot quality of historical printed textiles, which are still the goals for today's printing technology. These exceptional effects were rendered with tonal values created by cross-hatching and dry-point techniques, which consist of fine lines and dots. The so-called `toile design style', which originated from this historical printing method, set the standard for the look of traditional textile designs that are still on the market. In table and flatbed screen-printing technologies, which started at the beginning of the twentieth century, it is rather difficult to achieve perfect motif and screen registration alignment. The perfect butt-fittings of each motif into separate screens are almost impossible to print, especially in the early stages of this technological development. Eventually, the mechanical limitations of screen-printing began to dictate the print style. Screen-printed designs incorporated different scale motifs and fits, which prevented the problem of mismatching screen registrations. Instead of butt-fitting the motifs, the designers started to create intentional unfitted `looks' which were built into the design print aesthetic itself. For example, one of the adjacent motifs was either enlarged to have striations of wide trappings or reduced to have striations of unprinted areas. In current textile design, even though the precision of screen registrations has improved, some of the designs still maintain this `intentional unfitted style' for aesthetic purposes. In comparison, digital textile printing technology is free from mechanical constraints, which totally differentiates it from any other conventional textile printing processes. First and foremost, the elimination of screens and image transfer media benefits textile designs without step and repeat requirements. Secondly, any design digitally created on a VDU (visual display unit) can become an actuality, directly after it is printed on cloth with the CAD designing software, including proprietary textile design software as well as off-the-shelf Photoshop, Illustrator, etc. Precise color prediction and color matching can be established with color-management software and device calibration tools (spectrophotometers and colorimeters), including input (scanners, digital cameras, etc.), VDU, and output (print-outs). This instant digital imaging process provides textile designers greater opportunities to explore aesthetic novelty and to experiment with new imaging possibilities on fabrics to a greater degree than any previous textile printing technologies. Although historically there has been initial enthusiasm over new digital technologies, they have had a problematic infancy. When transfer-printing technology was introduced in the late 1960s, the textile printing industry was initially enthusiastic about the potential freedom of image creation without any

Design and workflow in digital inkjet printing

345

constraints. There were many experimental explorations of photographic images. However, the industry eventually realized that it was far from liberated, due to the fact that imagery could only be printed on limited synthetic substrates (Stark, 1990). Nonetheless, digital textile printing proved itself to be a technology truly without limitations. With current developments in colorations (inks), digital textile printing has the potential to produce images on almost any kind of fiber, just like conventional printing technology. Consequently, because of its characteristics of flexibility and freedom, digital textile printing technology generates new design styles that are impossible or extremely difficult to achieve by existing conventional printing technologies. Such styles can be categorized as follows: 1. 2. 3. 4. 5.

Millions of colors Extreme tonal effects and fine lines Photographic manipulation Special digital effects Large engineered images.

These will be discussed in the following sections.

19.3.2 Millions of colors Conventional textile printing is based on spot colors which are separately formulated on screens prior to printing. Conversely, digital printing is based on a preset process color of CMYK, whose combinations assign each pixel color of the images of the CAD textile design data. The textile design in CAD software can have the possibility of creation in 24-bit RGB colors. This is one of the advantages of digital textile printing, to visualize millions of colors in designs. Although retainable colors on designs in 24-bit RGB color space hold a much more extensive color gamut than the printable output color gamut, the creative possibilities in this design environment are enormous. Furthermore, several additional process colors become available, which are utilized in conjunction with the existing process colors of CMYK to increase the number of printable colors attainable. Moreover, color management software has been developed to control and manage more than four process colors to give precise color matching and prediction. Currently, in the exclusive high-end fashion printing market, some of the fashion accessories and dress prints use more than 30 colors representing original colorful creative looks. Depending on the volume of printing, these printing manufacturers utilize both digital and conventional printing technologies.

19.3.3 Extreme tonal effects and fine lines One of the most difficult tasks in conventional printing technology is the reproduction of smooth and clean graduating tonal effects. Throughout the

346

Digital printing of textiles

19.4 Example of extreme tonal effects and fine lines (Hitoshi Ujiie Design ß 2000).

history of textile printing technologies, one of the distinctive characteristics of the hand-engraved copperplate and hand-engraved roller printing technologies is the subtle and smooth shaded effects. To achieve these goals, the engravers and printers were renowned for their highly skillful craftsmanship, timeless efforts and experimentation (Storey, 1974). In contrast, digital textile printing has consistent depositions of microliters of colorants, controlled by the dithering algorithm of software, which enables it to produce smooth tonal effects and fine lines effortlessly on textile substrates. Figure 19.4 shows an example of extreme tonal effects achievable by digital printing technology.

19.3.4 Photographic manipulation Inputs of photographic images through scanners, cameras, and photographic stills from video cameras can be manipulated and printed digitally on cloth. This photographic manipulation incorporates the concept of `simulation and camouflage'. By integrating the historical idea of the `trompe l'oeil style', which refers to a design style that creates a `trick of the eye', printed photo-

Design and workflow in digital inkjet printing

347

graphic images can simulate the visual illusion as well. Historically, this `trompe l'oeil style' has been rendered with stylized painting techniques. The `simulation and camouflage style' is defined as the photo-realistic version of the `trompe l'oeil style'. The design exhibition `Skin: Surface Substance and Design' at the Cooper-Hewitt National Design Museum in New York in 2002 represented various photographic textile designs that simulated the illusion of real surfaces and/or objects on digitally printed textile substrates. These design images created provocative visual illusions, which generated the notion of unexpected and unusual textile design. The concept of `camouflage' on digitally printed textiles has been explored in military uniforms, developed not only for visual images but also camouflaged from infrared and other sensing devices. An example of photographic manipulation is shown in Fig. 19.5.

19.5 Example of photographic manipulation (Ion Design ß 2000).

348

Digital printing of textiles

19.3.5 Special digital effects The new printed textile designs can be characterized by a variety of special digital imaging effects. The image editing software provides a variety of digital effects by filtering and distorting tools. Any image can be manipulated to look digitally distorted and sometimes it gives an obviously prefabricated `digitally distorted and filtered' outcome. This could be considered controversial in terms of the authenticity of the originality of the imaging process, if the result represents too generic a digitally distorted and filtered look. Nonetheless, utilizing special image-editing software can be a great creative tool to visualize novel textile designs. Figure 19.6 shows an example.

19.3.6 Large engineered image A textile designer no longer needs to create images where the size and scale of the motifs are influenced by the mechanical and financial restrictions of the engraving process. Designers no longer need to be concerned with the concept of repeat and they can create one enormous engineered print for interior textile

19.6 Example of special digital effects (Hitoshi Ujiie Design ß 2003).

Design and workflow in digital inkjet printing

349

design and fashion design applications. The potential of repeatless imaging in printed textile design is immeasurable. Linking personalization and mass customization to a new business paradigm shift, where any choice of design can be printed on any desirable placement on demand, is the next wave of the future. Additionally, digital textile printing has stimulated the interest of artists, who visualize the images printed on textiles in the same way as paints on canvas. In 2002, the Textile Museum at Washington, DC, sponsored the exhibition `Technology as Catalyst: Textile Artist on the Cutting Edge', in which half a dozen artists and designers utilized digital printing technology on textiles to represent multiple creative possibilities. The Washington Post reviewed the exhibition as an `enhancement of an art form' as well as showing `unimagined freedom of scale'. One kind of digitally printed fabric has the potential to be explored as installation pieces, advertisements, stage sets, and art objects that allow for experimentation in the scale of figure to ground. An example of such a large engineered image is shown in Fig. 19.7. In the printed textile field today, implementation of digital printing technology will allow for a more comprehensive understanding of the textile

19.7 Example of large engineered image (Hitoshi Ujiie Design ß 2001).

350

Digital printing of textiles

design process. Digital literacy is mandatory for successful competition in the current textile market, whether for individual entrepreneurial success, or for industry standards. At the same time, this instant printing technology has created the opportunity for any digitally literate person to produce printed production, whether or not it is aesthetically qualitative or commercially viable. The role of textile design becomes a major part of the successful manufacture of printed textiles.

19.4 New definitions for the textile printing industry In general, the introduction of new technology influences society and changes our global structures and value systems. The classical industrial revolution in the eighteenth and nineteenth centuries introduced new machinery, labor systems and management styles in manufacturing. At the same time, it shifted sociocultural values, attitudes and lifestyles (Kranzberg, 1989). Diffusion of information technology in past decades also attests to the fact that many social and economic impacts were reflected in our society. Similarly, the evolution of digital textile printing technology has imposed new historical definitions and systems related to printed textile production and distribution.

19.4.1 On-demand production One advantage of digital textile printing is the digital management system, where the manufacturing and distribution processes are controlled and managed digitally. With the use of information technology, manufacturers communicate with individual consumers through the Internet and customers' printing needs are met immediately. Unlike the conventional textile printing environments, which require sizable inventory of screens, variety of colorants, and printed productions, by comparison digital manufacturing minimizes the inventory to digital information storage media. Manufacturers can also actualize the concept of just-in-time manufacturing by managing and synchronizing all stages of digital information. In the just-in-time digital textile printing manufacturing system, manufacturers can demonstrate one-to-one marketing to customers with electronic commerce, and the consumer's unlimited choices can be printed as personalized textile products. Moreover, digital manufacturing can link mass customization to a larger consumer domain. Manufacturers prepare a series of choices for design elements of the products through electronic commerce. According to the customer's choice of design elements, manufacturers can produce on demand. In conventional mass production, manufacturers provide tangible products from which consumers purchase. However, mass customization is the interactive exchange between the manufacturer and the consumer in the mass-market domain. Significantly, the focus has shifted individual customer's needs and preferences (King, 2002).

Design and workflow in digital inkjet printing

351

Seiren Viscotec in Japan provides one of the earliest examples of mass customization. On their website, the customer first determines his or her image selection method from two choices: (1) downloading the design software online to create the customer's own images on their personal computer, or (2) selection from prefabricated design elements and colors. Secondly, the customer chooses background colors and types for high quality T-shirts, where the images are placed. Finally, the digitally printed products ordered by the customer are delivered within two weeks. In both personalized and mass-customized manufacturing systems, digitally printed textile products can be categorized as valueadded products, which is one of the characteristics of digital textile printing today. Historically, the acceptance of new technology depends on end-users' psychological acceptance and readiness. The so-called `Generation Y', an upcoming consumer group, is one of the largest consumer groups in the USA. Its members, who were born between 1979 and 1994, possess the highest digital literacy and value products in which qualitative form and function appear to be components of successful marketing and key to the success of new manufacturing styles (Neuborne, 1999). Moreover, the success of these new manufacturing styles is attributed to the assurance and reliability of the interface and management in production, distribution and communication systems.

19.4.2 Cross-disciplinary application Currently, total annual worldwide textile print production has reached over 18.6 billion linear meters, and is increasing 1% a year (Stork, 2002). This figure represents all conventional printed textile production in the areas of clothing, decorative interior textiles and technical textiles. Aside from the textile industry, the graphic printing industry also utilizes textiles as printing substrates for their advertisements for printed products such as soft signage, pop materials, trade show graphics, flags, banners, etc. According to Web Consulting, digital printing technology has penetrated far more in the graphic design industry than in the conventional textile industry, in terms of digitally printed images on textile substrates. This is just one of the opportunities for the success of crossdisciplinary applications with new digital printing technologies. The use of the printing technology in a specific industry can be explored in diverse industries to maximize the creative possibilities. A digital textile printer can print images on different material surfaces with proper specifications of colorants and pretreatments, without requiring much downtime for changing the different colorations. This technology retains various cross-disciplinary opportunities, including researching more universal applications of specific types of printing colorants. Studies on the subject of `universal sets of colorations on digital textile printing', could possibly lead to novel engineering that enables us to print with one universal set of colorations for a variety of substrates. The same textile

352

Digital printing of textiles

design could presumably be actualized on various substrates of metal, ceramics, wood, plastics, etc., without changing types of colorants, and consequently increase opportunities for cross-disciplinary printing platforms. The cross-disciplinary approach is also manifested in the new business concept of `vast customization'. This term was introduced by I.T. Strategies in 2002. It defines the new business model as mass customization of a variety of products with digital printing technology. For example, printed products in the home furnishing market can be digitally printed, and mass customized, to create a visually cohesive interior. Such products can include bath rugs, comforters, mattress pads, shower curtains, towels, bed spreads, pillows, window treatments, sheets, table tops, etc. Furthermore, this cross-disciplinary approach has influenced the creative attitudes of various textile design communities. As textile designers broaden their preconceptions of textile design, they have the ability to create more crossdisciplinary environments with digital printing technology. Collaboration between architects, fashion designers, and graphic designers is already in the preliminary stages of fruition, due to innovative digital printing capacities. To illustrate, a textile designer could easily collaborate with an architect by creating an installation of printed fabric into an entrance of a specific building. Ultimately, the definition of textile design will be broadened into a larger category of surface design. Digital technology will allow surface designers to print virtually any image on any type of substrate surface. For this reason, textile/ surface design will most likely integrate more experimentation between the tactile surface and the printed image. A more innovative exploration between surface and print elements will introduce a `new' look in textile design. In the future, the creative possibilities will be endless.

19.4.3 Entrepreneurship Initially, during the development of digital textile printing technology in the past decade, the marketplace was filled with positive and enthusiastic attitudes towards this new technology. The printer manufacturers, software companies, and ink suppliers were promoting their products to the textile industry, and in turn, the textile manufacturers were interested in learning about the digital printing systems. In the early days of development, the majority of digital printers and software were designed for the graphics industry. These were modified for the textile industry and were originally engineered solely for textile printing production. At the same time, the equipment manufacturers, OEM companies and suppliers had very limited understanding of the textile industry. Naturally, there was much confusion generated in the marketplace, which delayed the penetration of digital printing technology into the textile printing industry. During this period, textile machine manufacturers developed complete digital textile printing systems. However, the system and equipment costs, as

Design and workflow in digital inkjet printing

353

well as the conservative attitudes of the textile printing manufacturers, at least in the USA, created widespread reluctance to accept this new technology. Therefore, the early pioneers of digital textile printing were not the conventional textile manufacturers, but rather specialized or smaller entrepreneurial groups such as a printer for automobiles in Japan, silk printers in Italy, design companies, graphical service bureaus, system integrators, etc. (Tippett, 2001). This entrepreneurship is still the backbone in the development of digital textile printing today. As previously mentioned, the ability of short-run production with digital textile printing technology is encouraging textile designers. Today they have the opportunity to demonstrate viable personal work, which is digitally printed on fabric, to prospective clients. A new breed of independent entrepreneurs with access to digital printers will be able to produce their own limited yardage. Individual designers can create their own short-run printed textile collections unlike in traditional printing, which requires long print runs and high engraving costs. In New York City, entrepreneurial designers from around the world exhibit their textile design products in the ICFF (International Contemporary Furniture Fair). Year by year, increasing numbers of textile designers present original short-run digitally printed textile products. In the near future, the textile printing industry will become a new form of `cottage industry', with a few large printing mills and many small printing operations supported by entrepreneurship.

19.4.4 Pedagogy and industry Interaction between the global expansion of the digital textile printing industry and educational institutions is critical for the worldwide success of this new market. The new skilled and creative talents from educational institutions eventually become the next workforce. At the same time, educational institutions are responsible for placing their students in their specialized professional fields. This demand and supply mechanism of the industry and educational institutions helps to proliferate the entire field. The introduction of any new technology and principle requires pedagogic revision in educational institutions. Recent developments in digital textile printing and subsequent pedagogic revisions in textile education include textile engineering, science, management and design, but the most outstanding revision has occurred in the area of textile design. The introduction of new digital printing technology has expanded the design processes and the designer's roles, which previously had been separated from the production process. The new skilled textile designers of the future will be required to have more direct involvement from start to finish in the manufacturing process. For example, as mentioned in the digital textile strike-off printing workflow, textile designers are responsible for quality control of printed design, printing process, operation and maintenance of printers, as well as design aesthetics. At the same time, many

354

Digital printing of textiles

textile designers will be more involved with the short- to medium-run productions of their designs in the near future. Therefore, future textile design curricula should require technical and engineering components of digital imaging and digital textile printing as well as existing design aesthetics and marketing (Ujiie, 2002).

19.5 Future trends The first 10 years of our new millennium can be a pivotal decade for innovation in the textile printing industry. In the future, digital textile printing technology will establish its permanent position in the print textile industry. However, because technology is still evolving in the full range of mass-production printing technology, many issues still remain: printing reliability, printing speed, cost of machines and supplies, attainable color gamut, penetration of colorants to the substrates, fastness, etc. Moreover, digital textile printing is still at the direct printing stage, and special printing application styles can be considered within digital textile printing technology. However, the special printing styles that have been explored thus far have been applied only to conventional textiles. Specialty prints have increased the competitive edge in the textile industry, due to their attractive novelty in the marketplace (Dawson and Hawkyard, 2000). Burn-out, discharge, or any other novel printing style, when adapted to digital printing technology, will generate additional competitive edge to the global textile printing industry. In terms of further research sources, information can be found in the list of references. There are very few published books specifically on digital textile printing, because the technology is still evolving. The majority of the information can be found in journals and from attending conferences. The journals include the Journal of the Society of Dyers and Colourists, AATCC Reviews, International Dyer, and the IS&T (The Society for Imaging Science and Technology) journals. The conferences include the NIP (Non Impact Printing) Conference of the IS&T, the Digital Textile Printing Conference of Web Consulting, the Digital Printing of Textile Conference of the IMI (Information Management Institute), the NTC conference (National Textile Center) or its website: www.ntcresearch.org, and conferences and workshops at the Center for Excellence of Digital Inkjet Printing of Textiles at Philadelphia University.

19.6 References Adiga A (2004), `Hanging by a Thread', Time, 164, 24. Bruni F (2004), `Italy Fights to Remain the Home of Luxury Fabrics', The New York Times, Feb 24, Fashion and Style. Chapman K (2002), `Digital Printing Success: A True Story', AATCC Rev, 2, 5, 15.

Design and workflow in digital inkjet printing

355

Clark D (2003), `Applications of Digital Ink-Jet Printing on Textiles', AATCC Rev., 3, 1, 14±16. Compton M (2004), `New Opportunities and Real Life Success Stories in Digital Print Production', Digital Textile Printing Conference, Cary North Carolina US, [TC]2. Dawson T (2003), `Carpet and Yarn Printings', in Miles L, Textile Printing, Hampshire UK, SDC, 99±138. Dawson T and Hawkyard C (2000), `A new millennium of textile printing', Rev Prog Coloration, 30, 7±19. Eckert C (1997), `Design Inspiration and Design Performance', T I 78th World Conf and 5th Text Symp Textile and the Information Society, Thessaloniki Greece, Text Inst, 369±387. Glover B (2004), `The latest technology developments in ink jet printing from ITMA 2003', in Dawson T and Glover B, Textile Ink Jet Printing, Bradford, UK, SDC, 13± 29. Grubler A (1998), Technology and Global Change, Cambridge UK, University Press. Henry P (2004), `Computer sided design: Its pedigree and future contribution to the success of digital printing', in Dawson T and Glover B, Textile Ink Jet Printing, Bradford UK, SDC, 38±43. Int. Dyer (1999a), `The cheapest jobs are the best-prepared jobs', Int Dyer, 184, 9, 19±21. Int. Dyer (1999b), `Digital opportunity', Int. Dyer, 184, 9, 22±24. King M (2002), `Digital Textile Printing and Mass Customization', AATCC Rev, 2, 6, 9± 12. Kranzberg M (1989), `The Information Age', in Forester T, Computer in the Human Context, Cambridge Massachusetts US, MIT Press.19±32. Moser L (2003), `ITMA 2003 Review: Textile Printing', JTATM, 3, 3, 8, 12±14. Moxey J (1998), `A Creative Methodology for Idea Generation in Printed Textile Design', J Text Inst, 89, 3, 35±43. Neuborne E (1999), `Generation Y', BusinessWeek, 3616, 80. Pond S (2000), Inkjet Technology, Carlsbad California US, Torrey Pines Research. Rayleigh J (1878), `Instability of Jets', Proc London Math Soc, 10, 4. Schoeser M (1986), Fabrics and Wallpapers, London, Bell and Hyman. Spruijt J (1991), `The future textile printing', JSDC, 107, 75±76. Stark S (1990), `Technological Impact on Design Changes in 19th Century Printed Textile Fabrics', Ann Arbor Michigan US, Bell and Howell. Storey J (1974), `Manual of Textile Printing', New York, Van Nostrand Reinhold. Stork (2002), Developments in the textile printing industry 2002, Boxmear Netherlands, Stork Textile Printing Group, Tippett B (2001), `The Future of Textile Printing Will be Digital', NIP 17, Fort Lauderdale Florida US, IS&T, 418±422. Ujiie H (2002), `Textile Education in Digital Inkjet Fabric Printing', NIP 18, San Diego California US, IS&T, 254±257. Wilson J (2001), Handbook of textile design, Cambridge UK, Woodhead.

Index

abrasion 128, 131±4 absolute colorimetric (match) rendering intent 174, 184 absorbency, fabric 207, 265 acetate 234 acid dye inks 93, 205, 219, 250±1, 318±19 Artistri 2020 74, 75 dye±fibre interaction 233, 234 fastness and Tx series 109±10, 112 substrate preparation 210, 211 acids 210, 211 acoustic inkjet 41±2 acryl 234 additional inks 191±2 additives 237, 239, 258 adhesive print blanket (`sticky belt') 78±9, 103±5 agile manufacturing 296 air bubbles, trapped 44 air content 227, 229 alcohol 261, 262, 265, 266 alginate 208, 211, 212, 213, 240±1, 266±72 Algotex 11±12 alignment of print head 76 alkali 207, 208 effect of cationisation on alkali concentration 283±4 see also sodium carbonate all-in printing 201, 202 Amber PIJ printer 7 Amethyst inkjet printer 7 ammonium tartrate 210, 211 anionic dispersing agent 212 antifoam agents 221, 226 apparel manufacturers 305±6 apparel market 71±2 applications support 82 Aprion print head technology 7, 85, 86,

93, 94±5 aqueous inkjet ink 233±51 acid and direct dye ink formulation 250±1 additives 237, 239 disperse dye ink 245±50 dye±fibre interaction 233±5 organic solvents and surface energy 235, 236 production process of inkjet-printed textiles 240 reactive dye ink 240±5 reliability 237±40 time-dependent phenomena and surface-active components 235±7, 238 area coverage 189±90 Artistri Technology Center (ATC) 81 Artistri 2020 system 4, 7, 8, 9, 69±83, 322±3, 341 applications support, technical service and training 82 colour management 77±8 competitive environment 73±4 cost of printing 79±80 fabric handling 78±9 ink, pre-treatment and post-treatment 74±6 print head 76±7 print speeds and resolution 72, 73 artwork 18 assembly 321, 322, 327 ATME 337 ATP Color 11 automatic flushing cycle 160, 176 automatic nozzle recovery (ANR) unit 117±18, 119 automatic pattern generation 309±10 automation 294

Index back shooter TIJ 37, 38 banding 259 prevention in Tx series 102±6 batched orders 326 bend mode PIJ 34 binary deflection 31 binary halftoning systems 153 binders 219, 220, 221, 226, 230 selection 223±4 biocides 221, 226 black generation models 186 blanket 78±9, 92, 103±5 bleeding 230, 257, 258 prevention in Tx series 106±8 block printing 2, 343±4 BMP format 151, 152 body scanning 25±6, 161, 309±10 breakthrough technologies 306±7 Brother 48, 49, 50 bubble collapse 45 calibration display, input and output devices 168±73 DReAM system 90 of mixed colours 190±4 of single colours 188±90, 191 calibration chart 190±2 cameras, digital 149, 150, 172 camouflage and simulation 346±7 Canon 48, 51 Bubble Jet printers 4 laser technology 5 TPU 161 carpet printing, digital 3±4, 202, 337 cartridges, ink 75±6, 119±20, 160 cathode ray tube (CRT) monitors 169±71 cationic agents 209 cationic polymers 212 cationisation 209±10, 276±89 effect on alkali concentration 283±4 effect of cationic reagent concentration 280±1 effect on outline sharpness 281, 282 effect of thickener concentration and 285±6 effect on steaming time 281±3 effect on thickener concentration 284±5 pre-treatment for inkjet printing 279 printing processes 279 printing results with CMYK colours and reduced chemical levels 286±8 process 278±9

357

testing methods 279±80 washing 279 cavitation damage 45 cellulose/cellulosic fabrics 124, 125, 254±5 pre-treatments 207±10 see also cotton centres of textile production excellence 16±17 characterisation of display, input and output devices 168±73 chemical bonds 233 chemical cross-linking 230 China 342±3 3-chloro-2-hydroxypropyltrimethylammonium chloride 277, 278 Chromojet system 9, 202, 337 Chromotex system 9, 161, 202 Ciba Specialty Chemicals 85±6, 93±4, 95±7 CIE colour spaces 166, 180 CIELAB 78, 166, 180, 183, 267±72 CIELCH 166, 180 XYZ space 166, 167, 180, 183 closed loop colour management 182 CMY (cyan, magenta, yellow) 163, 166±7 colour gamut 173±4 CMYK (cyan, magenta, yellow, black) 78, 147, 163, 180, 195 cationised cotton and reduced chemical levels 286±8 colour gamut and textile printing 181, 182 coating a screen 20 ColorBooster 12±13 ColorGPS 187, 190±4, 197 colorimeters 170±1 Colorprint Twister 11 ColorSync 168 colour accuracy 263±4 colour book 78, 90 colour communication software 169±70, 175 colour data transpositions 164 colour density 227, 230 colour depth 267±72 colour difference 176, 195 Colour Engine 166, 183 colour fastness 101 cationisation 280, 281, 282, 283, 284, 287, 288 integrated fabric formation and colouration 127, 130±1, 140 Tx series 109±10, 111±13

358

Index

colour gamuts 78, 173±4 CMYK and 181, 182 comparison of six and eleven colour setups 195±6, 197 gamut mapping 187, 193, 194 largeness of 196 new design styles 345 number of inks and 193 pre-treatment and print quality 263±4, 267±72 testing 227, 230 that can be attained using CMY inks 157±9 Tx series 106, 107 colour management 163±98, 342 Artistri 2020 77±8 characterising display, input and output devices 168±73 colour communication 175 colour gamuts see colour gamuts colour reproduction performance of equipment operated with a CMM 175±7 future trends 177 general numerical colour specifications 166±8 ICC see ICC colour management JIT 319±20 InkDrop Boutique 331 integrated systems 323 Stork U See 333±5 new design styles and millions of colours 345 rendering intents 173±4, 183±4, 185 textile colours and common colour spaces 181±2 colour management systems (CMSs) 148, 164±6 colour reproduction performance of equipment operated with 175±7 Colour Matching Module (CMM or Colour Engine) 168, 183 colour profiles 90, 180±1, 230 ICC see ICC colour management colour reduction 325 colour reproduction 101, 273 Tx series 106, 107 see also colour gamuts colour separation 18±19, 149±50, 325, 339 colour spaces 166±8, 180 textile colours and 181±2 see also CIE colour spaces colour strength 279, 280, 281±2, 283, 284±5, 286±8

colour targets 172 colour uniformity 230 colour variations 21, 22 colourway 78 Como, Italy 16±17, 22 computer aided design (CAD) 17, 21±2, 23, 147, 339 editing and data storage systems 148±52 computer aided manufacture (CAM) 147 connectivity 298±9 consumer expectations 293 contingency planning 313±14 continuous improvement 313±14 continuous inkjet technology (CIJ) 3, 29, 30±2, 48, 50 continuous jetting reliability test 227, 229±30 conventional workflow 338±40 co-solvents 221, 224, 226 costs price and performance of textile inkjet printers 73±4 running costs 46±7, 101, 121 Artistri 2020 79±80, 82 DReAM 88 Tx series 116, 117 screen printing compared with digital printing 69±70, 79, 116, 117 cotton 124, 202, 234 cationised see cationisation polyester cotton blends 202, 212, 255 pre-treatment 207 pre-treatments for inkjet printing 207±10 case study 260±72 covalent bonds 233 craft production 295±7 creativity 23±4 crocking, colour fastness to 227, 231 cationisation 280, 281, 282, 283, 284, 287, 288 integrated fabric formation and colouration 127, 130±1, 140 cross-disciplinary approach 351±2 cut parts 295 printing on 307 cutting 321, 322, 327 d.gen International 13±14 data storage systems 148±52 data transmission system 175 Dawson International 201±2 defoamers 221, 226

Index degassing 75 delivery, order 327 demand, printing on 313±14, 350±1 Demand-Activated Manufacturing Architecture (DAMA) 252, 305 design 16±26 capture of digital design data 150, 163 comparison of traditional and digital methods 22±3 conventional textile printing workflow 338±40 cross-disciplinary approaches 351±2 digital process 21±2 educational institutions 353±4 entrepreneurship 353 impact of digital on 23±6 JIT 315±16, 317, 325±6 new design styles 343±50 extreme tonal effects and fine lines 345±6 large engineered images 348±50 millions of colours 345 photographic manipulation 346±7 special digital effects 348 textile printing mechanics and 343±5 new market 26 traditional methods for producing a textile collection 18±21 see also image preparation design competition, era of 302, 303 design libraries 325±6 design templates 328, 330 device dependent colour spaces 180 device independent colour spaces 180, 183 DewPrint machine 202 DGS 9 Diamond Sutra 2 diazo systems 20 dichromate systems 20 diffusion 256, 258 rectified diffusion 44, 75 digital grand format printing 5±6 digital islands 295, 308, 309 Digital Print Asia (DPA) 12 digital strike-off printing workflow 340±1 digital supply chain 307±9 future trends 309±10 Digital Technological Center (DTC) 95±7 activities 96±7 digital textile printing applications 70±2 costs compared with screen printing 69±70, 79, 116, 117

359

evolution of 1±15, 98±9, 337±8 carpet printing 3±4 DPI 2001 8 Drupa 2000 7 Drupa 2004 9 FESPA 1996 6 FESPA 1999 6±7 FESPA 2005 10±14 grand format 5±6 Heimtextil 2001 7 ITMA 1999 6±7 ITMA 2003 8±9 Seiren 5 SGIA 2004 10 sublimation 4 thermal inkjet 4 future of 120±2 industry needs 69±70 market needs 101 markets 70±2 origins 2±3 potential worldwide opportunity 69 vs traditional textile printing 16±26 2,3-dihydroxypropyltrimethylammonium chloride 277 direct dyes 233, 234, 250±1 discharge approach 231±2 discontinuities, technological 302 disperse dye inks 93, 205, 219, 245±50, 318±19 Artistri 2020 74±5 direct printing 245, 246±9 dye±fibre interaction 233, 234 fastness and Tx series 109±10, 113 flow rheological property 249±50 printing with 245, 246 sublimation transfer printing 245±6 substrate preparation 211, 212 types for inkjet 245±6, 247 dispersing agents 212, 215, 246±8 displacement 43±4 display monitors 163, 169±71 dithering 154±7 of whole pixels 175 DNG 152 Domino 48, 50 dot gain 157, 190, 256±7 dots per inch (dpi) 46 double heater TIJ 37, 39 DPI 2001 8 DPS 7 DReAM 9, 73, 74, 84±97, 161, 322±3, 341 DTC 95±7

360

Index

goals of the project 87±8 innovations 86±7 machine 88±95 blanket 92 drying phase 92±3 entry to the machine 92 inks 93±4 print heads 94±5 printing phase 92 project in textile printing context 84±7 drop ejector failure 45 drop formation 53, 54±7, 65 drop impaction 53±4, 57±66 future trends 65±6 rough surfaces 62±3, 64 smooth surfaces 65±6 homogeneous liquids 58±61 particle-laden drops 63±5 drop-on-demand inkjet (DOD) 29±30, 32±43, 318 comparison of PIJ and TIJ 43±5 DReAM 94±5 other technologies 30, 40±3 piezoelectric (PIJ) 29±30, 32±6, 48, 50±1, 318 thermal (TIJ) 4, 29±30, 36±40, 48, 51, 202 drop volume 45±6 drop volume control 36, 44±5 droplets 152±3 placing within a superpixel 154±7 Drupa 2000 7 Drupa 2004 9 dry cleaning fastness 227, 231 drying coated screen 18, 20 DReAM 92±3 drying rate 227, 229 DuPont 7, 8, 9 Artistri 2020 system see Artistri 2020 system Artistri Technology Center 81 durability of printed images 230±1 dust 117 dye±fibre interaction 233±5 dye load 243±5, 248±9 dye penetration 279, 280±1, 282, 283, 284, 285, 286, 287 edge raggedness 257 editing 148±52 educational institutions 353±4 electro-hydrodynamic extraction inkjet 42±3

electrophotography 2±3 electrostatic inkjet 41 electrostatic printing 2±3 Elema Oscilomink 3 Elmer's glue 223 Embleme 6 Encad 4, 7, 160 encoding 151±2 energy drop impaction 58±9 surface energy 235, 236, 237 engraving 18±21, 338±40 see also screen printing entrepreneurship 352±3 2,3-epoxypropyltrimethylammonium chloride 277 EPS 152 Epson 9, 48, 50 error diffusion dithering methods 155±6 Eton Unit Production System (UPS) 327 evaporation speed of solvents 237±40 exposure 18, 20 extrapolation, trend 301 fabric feed 78±9 accuracy 102±6 applicability of fabrics and fabric feed mechanism for Tx series 110±15 fabric structure 261, 262, 263, 264, 267, 268, 269±72 Fast T-Jet garment printer 10 fastness see colour fastness Fastran Process 202 ferment, era of 302, 303 FESPA 1996 4, 6 FESPA 1999 6±7 FESPA 2005 10±14 fibres dye±fibre interaction 233±5 matching ink chemistry and 254±6 see also under individual types of fibre file compression 151 file formats 151±2 fillers 209 film-forming agents 159 filterability, ink 227, 229 fine lines 345±6 finishing, fabric 320±1, 323, 326±7 fixation 213, 276±7, 320±1 see also print fixatives flatbed screen printing 252, 338±40, 344 flatbed scanners 149, 150 flexibility 298±9, 313 Floyd Steinberg method 155±6

Index flushing cycle 160, 176 foaming 227, 229 forecasting, technology 300±4 freelance design 339 full digital textile printing production workflow 341±3 gamma response curves 170 gamma value 169 gamut mapping 187, 193, 194 Generation Y 351 Gerber Scientific 5, 6 GIF 152 gradation 108±9 graininess 257, 261±3 grand format printing 5±6 graphic printing industry 351 grey levels 154±7 grey scale 46 grey tracking 171 Haloid 2 heat-aging stability test 227, 229 heat presses 231 Heimtextil 2001 7 Heracle dual gantry inkjet printing system 13±14 Hertz method 31 Hewlett-Packard 4, 48, 51 hiding approach 231±2 high-speed single±pass inkjet printer for corrugated cardboard 9 HimeTex 2000 98 historical analogies 303±4 Hollanders Printing Systems 12±13 home furnishings market 71 homogeneous liquid drop impaction rough surfaces 62±3 smooth surfaces 58±61, 65±6 hot air ovens 231 humectants 221, 224, 226 hydrophilic non-ionic polymers 209 hydroscopic organic compounds 224 hydroxyl groups 124, 125 IBM 4 ICC colour management 167±8, 180±98 advantages 184±5 basics 182±4 current technologies 187±94 ICC profile generation 190±4 linearisation wizard 188±90, 191 print environments 187±8 disadvantages 185

361

future trends 196±7 requirements and problems for ICC profiling 185±7 results 194±6, 197 Ichinose 4, 7, 161 ICM 168 Idanit 6 ideal model of dot gain 257 image editing software 348 image preparation 147±62, 344 control of printing machine 159 future trends 160±1 image design 148±52 capture of digital design data 150, 163 encoding, compression and storage of pattern data 151±2 image formation 152±9 colour gamuts using CMY inks 157±9 drop placement within a superpixel 154±7 JIT 315±16, 325±6 machine performance monitoring 159±60 image quality 82, 259 pigment inks 230 pre-treatment and 256±8 print head selection 45±6 Tx series 101±6, 121 Imaje 48, 50 in-store customisation 161 Inca Digital 9 industrial revolution 295 ink cartridges 75±6, 119±20, 160 ink coverage 79±80 ink fill-up test 227, 229±30 ink filterability 227, 229 ink latitude 44, 47±8 ink limit 192 ink mists 117, 118 ink royalty 47 InkDrop Boutique 327±33 inkjet technology 5, 29±52, 84±5, 147±62, 163, 214 CAD, editing and data storage systems 148±52 companies currently active in print head technology 48±9, 50±1 compared with screen printing 22±3 continuous (CIJ) 3, 29, 30±2, 48, 50 control of printing machine 159 development of 2±3 drop formation 53, 54±7, 65

362

Index

drop impaction 53±4, 57±66 drop-on-demand (DOD) 29±30, 32±43, 318 factors to consider 45±8 future trends 49, 160±1 limitations 85, 214±15 machine performance monitoring 159±60 pixel and image formation 152±9 stability and reliability and Tx series 116±18, 119 strengths and weaknesses 43±5 inks 204±6, 274, 318±19 additional 191±2 aqueous see aqueous inkjet ink Artistri 2020 74±6, 80 DReAM 93±4 `ideal solution' ink 319 ink-textile substrate interaction 256±8 matching fibre material and ink chemistry 254±6 pigment see pigment inks `universal' ink type for all fabrics 83, 310 universal set of for all substrates 351±2 waste ink 80 inoperative jets 160, 176±7 input devices 171±2 integrated fabric formation and colouration 123±43 evaluation of print fixatives 128±9, 134±8 experimental 126±9 final evaluation 129, 138±42 print fixative as sizing agent 124, 128, 131±4, 142 results 129±42 size evaluation 126±7, 129±31 integrated systems approaches 321±3 intentional unfitted style 344 interchangeable parts 295 inter-colour bleeding see bleeding International Contemporary Furniture Fair (ICFF) 353 International Textile Machinery Exhibition (ITMA) 1999 6±7, 98 2003 8±9, 337, 341 Internet 26 ionic bonds 233 ISO 13660 standard 135, 259, 260 IT8 colour targets 172±3 Italy 14, 16±17, 22, 25

Jacquard loom 16 jaggies 295 jet faults 160, 176±7 JetPrint material handling system 11 JPEG 151, 152 Jumbo Fast T-Jet 10 just-in-time printing (JIT) 312±36, 350 case studies 327±35 InkDrop Boutique 327±33 Stork U See 333±5 enabling the process 314±23 cutting and product assembly 321 fabric preparation and finishing 320±1 printing 316±20 product development 315±16 supporting technology, processes and materials 315±21 technology development and integration 321±3 just-in-time concept 312±13 order processing 323±7 principles of JIT strategy 313±14 printing on demand in digital environment 313±14 K/S values cationisation 279, 280±5, 286±8 disperse dye inks 243±5 reactive dye inks 243±5 Kimoto 14 knitted fabrics 254, 261, 262, 263, 264, 265±72 knurled rollers 103 Kodak 49 kogation 45, 177 Konica 7, 73, 74 Korea 24 Kornit Digital 10 931 dual platen printer 10±11 Kubelka-Munk theory 243, 279 L*a*b* colour system 78, 166, 180, 183, 267±72 lamp technology 302±3 large engineered images 348±50 laser technology 5 latent acids 210, 211 latex polymers 224, 226 LCD monitors 171 LCH (lightness, colour, hue) 166 Lectra Systems 12, 323 Leggett and Platt (L&P) Virtu series 8 Lexmark 4, 48, 51, 160

Index life cycles product 299±300 technology 302±3 light, colour fastness to 127, 130, 287, 288 limited editions 25 line density 129, 136±7, 139±42 line quality analysis integrated fabric formation and colouration 129, 134±8, 139±42 pre-treatment and print quality 261, 262, 267, 268 line raggedness (rag) 129, 135±6, 139±42 line width 345±6 effect of cationisation 280, 281, 282 effect of thickener concentration and 285±6 integrated fabric formation and colouration 129, 137±8, 139±42 linearisation wizard 188±90, 191 lint 117 liquid ink fault tolerant (LIFT) inkjet 43 lithographers 25 long-term strategy 313±14 lookup tables (LUTs) 78, 167, 168, 173 low strike-through resistance fabrics 110±15 Lucas±Washburn equation 106±7 manipulation, photographic 346±7 manufacturing flexibility 298±9, 313 markers JIT printing 321, 322, 332 supply chains and 304, 306, 307±8 markets 26, 70±2, 81, 121 mass customisation 25, 293±311, 350±1 development from craft production and mass production 295±7 digital printing supply chains 307±9 future trends 309±10 JIT printing 331±3 limitations 297±8 product life cycles 299±300 technology forecasting 300±4 time, technology and connectivity 298±9 traditional supply chains 304±7 mass density, ink 227, 228 mass production 295±7 Match Print II software 9 maximum spreading ratio 59, 60, 61, 63 McCue patent 6 mechanical binding 230 meniscus 228

363

Metro Media Technologies (MMT) 5 Milliken Millitron system 3±4, 202, 337 Mimaki 7, 9, 12, 73, 161 Tx series see Tx series minimum thread radius 55±7 mixed colours, calibration of 190±4 modulus 128, 131±4 monitors 163, 169±71 Monna Lisa system 9, 73, 74, 341 monochlorotriazine (MCT) reactive dyes 242, 244 Monti Antonio 14 mottle 257 moveable member TIJ 38±9, 40 MS printing systems 11 multifunctional dispersing agents (MFDA) 215 multiheater TIJ 37, 39 multilevel halftoning 153 multiple deflection 31 Murray Davies formula 189 Mutoh 73 Natura process 6 necking 54±7 novel products/markets 121 nozzle excitation PIJ 34, 36 nozzle packing density 43±4 nozzles automatic nozzle recovery 117±18, 119 cost per nozzle 46 increasing by introducing a stagger head 119 nylon 211, 234 on-demand production 313±14, 350±1 on-line sales 324±5 one-at-a-time production 313±14 optical density 189±90, 227, 230 optical dot gain 256 order processing, JIT 323±7 capturing order information 324±5 design and image management 325±6 fabric preparation, printing and finishing 326±7 order delivery 327 order to delivery time 296, 297 organic solvents 235, 236, 237 evaporation speed 237±40 Osiris 48 outline sharpness see line width outsourcing 326±7 pad-batch method 278±9

364

Index

page-wide array systems 49 paper, for transfer printing 213 paper printing, inks for 206 particle diameter/drop diameter ratio 64, 65 particle-laden liquids drop formation 53, 54±7, 65 drop impaction 54, 63±5, 66 particle size 222, 227, 228 particle volume fraction 64, 65 partitive colour mixing 158 PDF 152 pedagogy 353±4 penetrants 221, 226 penetration, dye 279, 280±1, 282, 283, 284, 285, 286, 287 penetration length 106±7 perceptual (picture) rendering intent 174, 184, 185 Perfecta 4, 6, 7 performance of printing machine monitoring 159±60 and price 73±4 permanence of printed images 230±1 pH 227, 228 phase separation 227, 229 Philyasystem 14 photo emulsion 18, 21 mixing 19 photographic manipulation 346±7 physical dot gain 256 piezoelectric inkjet (PIJ) 29±30, 32±6, 48, 50±1, 318 compared with thermal inkjet 43±5 pigment dispersion 221, 222±3 pigment inks 93, 202±3, 206, 215, 218±32, 318±19 Artistri 2020 75 formulation for digital textile printing 221±6 overview 219±21 post-treatments 231 pre-treatments 212±13, 231 tests and test methods 227±31 basic physical properties 228 durability and permanence of images 230±1 ink and media interaction properties 230 ink and print head interaction properties 229 overview 227 regulation and safety 231 shelf lifetime properties and jetting

reliability tests 229±30 white ink 231±2 pin-holes 21 pixels 152±9 drop placement within a superpixel 154±7 PNG 152 polyamide fibres 255 polyester 211±12, 234, 255 polyester/cotton blends 202, 212, 255 polyethylene glycols (PEG) 225, 226 polymeric pigment dispersion 223 porous layer feed PIJ 34, 36 porous substrates 58 positive 18 preparing 20 post-treatments 181±2, 213±14, 219, 231, 240, 255±6, 274 Artistri 2020 74±6 predictive models of drop impaction 60, 61, 66 preparation, fabric 320±1, 323, 326±7 pre-pulses 36, 45 pre-treatments 159, 201±17, 219, 240, 252±75 Artistri 2020 74±6 DReAM 90 effect on print quality 258±74 case study 260±72 pre-treatment effects on print quality 259±60 print quality measurement 259 fabric pre-treatment sequence 206±7 future trends 215, 272±4 ink systems 204±6 for inkjet printing 207±13, 254±8 cotton and cellulosics 207±10 ink-textile substrate interaction 256±8 matching fibre material and ink chemistry 254±6 nylon 211 paper for transfer printing 213 polyester 211±12 polyester/cellulose blends 212 pre-treatments of substrates 258 wool and silk 210±11 jet printing machines 214 JIT 320, 326 limitations 214±15 pigment printing 212±13, 231 reasons for 203±4 Tx series 105±6 price, and performance of printers 73±4

Index print environments 187±8 print fixatives evaluation of 128±9, 134±8 as sizing agents 124, 128, 131±4, 142 print heads 29±52 Aprion print head technology 7, 85, 86, 93, 94±5 Artistri 2020 76±7 companies currently active in print head technology 48±9, 50±1 drop formation 53, 54±7, 65 drop impaction 53±4, 57±66 factors to consider in selection 45±8 cost 46±7 image quality 45±6 ink latitude 44, 47±8 productivity 47 future trends 49 inkjet technologies 29±45 continuous 3, 29, 30±2, 48, 50 drop-on-demand 29±30, 32±43, 318 piezoelectric 29±30, 32±6, 48, 50±1, 318 strengths and weaknesses 43±5 thermal 4, 29±30, 36±40, 48, 51, 202 lifetime 46±7 print head productivity cost 47 print quality 258±74 case study 260±72 colour gamut and colour depth 263, 264, 267±72 graininess analysis 261±3 line quality analysis 261, 262, 267, 268 pre-treated fabrics 265±72 un-pretreated fabrics 260±4 effects of pre-treatment on 259±60, 272±4 measurement 259, 260 problems 258±9 printability 227, 230 printer characterisation 172±3 printer driver 159 printing pastes 106±8 printing speed 70, 82, 121, 252, 316, 318 Artistri 2020 72, 73 factors influencing 154 price and performance of inkjet printers 73±4 process colours 79, 80±1, 147, 319±20, 345 DReAM 93±4 product assembly 321, 322, 327 product development 315±16, 317, 324

365

Product Development Management (PDM) packages 309 product life cycles 299±300 production conventional workflow 339±40 on-demand production 313±14, 350±1 process for inkjet-printed textiles 240 scale of 318 productivity 47, 101, 121 Tx series 118±20 Profile Connection Space (PCS) 168, 183 proofing 184±5 protein fibres 255 see also silk; wool pure-liquid drops see homogeneous liquid drop impaction push mode PIJ 34, 35 2-pyrrolidone 224, 226 quality 70 image see image quality print see print quality Quick Response 296 raggedness edge 257 line 129, 135±6, 139±42 Rainbow (RB) printer series 11±12 Ratti group 23 RAW 150, 152 rayon 62, 234 reactive dye inks 93, 124, 202, 204±5, 219, 240±5, 318±19 Artistri 2020 75 cationised cotton see cationisation dye±fibre interaction 233, 234 dye load and colouration 243±5 fastness and Tx series 109±10, 111 pre-treatments 207, 210 printing with 240±1 and reaction mechanism 241±2 stability 242, 243±4 reactive moieties 241±2 reactivity 204 rebounding 59 rectified diffusion 44, 75 Reggiani 9, 85±6, 273 DReAM see DReAM DTC 95±7 register marks 20 regression analysis 301±2 regulation 227, 231 relative colorimetric (proof) rendering intent 174, 184

366

Index

reliability of inks 237±40 rendering intents 173±4, 183±4, 185 resistance 23, 24±5 resolution 46, 101±2, 230 respect for people 313±14 retailing 324 retraction 59 reverse osmosis 204 Reynolds number 58, 237 RGB (red, green, blue) colour space 163, 166±7, 180, 195 transformation into XYZ values for sRGB colour space 168 rheology modifiers 221, 225, 226 Rhome printers 6 RIP (raster image processor) software 89±90, 159 Robustelli Monna Lisa system 9, 73, 74, 341 Roland 73 roller screen printing 123, 252 roof shooter TIJ 37 rotary scanners 149 rotary screen printing 252±3, 338±40 roughness, surface 58 drop impaction on a rough surface 62±3, 64 RPL Supplies 4 run length encoding (RLE) 151 running costs see costs Russell Corporation 305, 306 S-curves 302, 303 safety 227, 231 Salsa 7 samples 21, 22 short-run sample printing 340±1 satellites 57, 257, 259 saturation (graphic) rendering intent 174, 184 Sawgrass 6 scale of production 318 scanners 148±50, 171±2 Scitex Iris 3 Scitex Vision 85±6, 93 DReAM print heads 94±5 Scotchprint 2000 3 screen printing 80, 123, 344, 346 costs compared with digital inkjet printing 69±70, 79, 116, 117 designer's perspective on digital vs traditional textile printing 16±26 spot colour process 79 traditional method 18±21

workflow 338±40 sedimentation 250 Seiko Printek (SPT) 76 Seiren 5, 120, 161, 212 Viscotecs system 5, 351 self-dispersed pigment dispersion 223 settling rate 227, 229 sewing threads 310 SGIA 2004 10 shear mode (shared wall) PIJ 33±4, 76 shear rate 249±50 short-run sample printing 340±1 side shooter TIJ 37, 38 Siemens 3 silica 126±42, 266±72 silicone 266±72 silk 210±11, 234 silk accessories market 70, 71 simulation and camouflage 346±7 single ink channels, calibration of 188±90, 191 size distribution 222, 227, 228 sizing agents (sizes) 123 evaluation 126±7, 129±31 print fixatives as 124, 128, 131±4, 142 smooth surfaces, drop impaction on 65±6 homogeneous liquids 58±61 particle-laden drops 63±5 sodium alginate 208, 211, 212, 213, 240±1 sodium bicarbonate 208 sodium carbonate 124, 126±42, 207, 208, 240±1 sodium dithionite (sodium hydrosulphite) 214 sodium meta nitrobenzene sulphonate 209 soft proofing 184±5 soft signage market 72 software 319 printer driver 159 RIP 89±90, 159 solid level (total solids) 223, 227, 228 Sony 49 special digital imaging effects 348 Spectra 48, 51 spectrophotometers 170±1, 172 speed, printing see printing speed speed to market 298, 300 splashing 58 spot colours 79, 80, 93, 319 spreading ratio 58±61, 63 squeeze mode PIJ 34, 35 sRGB colour space 167±8

Index stability reactive dye ink 242, 243±4 stable and reliable performance of Tx series 116±18, 119 stagger head 119 start-up failure 237±40 steaming 214 effect of cationisation on steaming time 281±3 Stork 3, 6, 7, 12, 323 Stork U See 12, 333±5 stretch fabrics 110±15 strike-off 338±40 digital strike-off printing workflow 340±1 strike-through 101 Tx series 108±9 low strike-through resistance fabrics 110±15 sublimation transfer printing see transfer (sublimation) printing substantivity 204 substitution, era of 302, 303 substrate preparation 201±17 fabric pre-treatment sequence 206±7 future trends 215 ink systems 204±6 jet printing machines 214 limitations 214±15 post-treatments 213±14 pre-treatments for inkjet printing 207±13 reasons for pretreatment 203±4 subtractive colour mixing 158 Sun Chemical 9 superpixels 153±4 drop placement within 154±7 supply chains 293, 294, 304±10 complexity of 313 digital printing supply chains 307±9 future trends 309±10 traditional supply chains 304±7 trends in supply chain strategies 295±7 surface-active components 235±7, 238 surface design 352 surface energy 235, 236, 237 surface tension 2, 222±3, 227, 228, 237, 238 surface tension driven inkjet 43 surfactants 221, 224±5, 226 suspended heater TIJ 37±8, 40 swimwear market 70, 71

367

table screen printing 252, 338±40, 344 [TC]2 316, 317 InkDrop Boutique 327±33 technical service 82 technological discontinuities 302 technology forecasting 300±4 tenacity 128, 131±4 tensile properties 128, 131±4 test patterns/images 160, 172, 177, 195 TexPrint 187, 188±90, 191 TGA 152 thermal excitation 31±2, 49 thermal inkjet (TIJ) 4, 29±30, 36±40, 48, 51, 202 compared with piezoelectric inkjet 43±5 thermo-mechanical inkjet 42 thickeners effect of cationisation on thickener concentration 284±5 effect of thickener concentration and cationisation on outline sharpness 285±6 for inkjet printing pre±treatments 207, 208, 210, 211, 212, 258 integrated fabric formation and colouration 126±42 thread radius, minimum 55±7 TIF(F) 151, 152 time compression 296, 297, 298, 300 time-dependent phenomena 235±7 time-series estimation 301 toggle-switch type cartridge 119±20 tonal effects, extreme 345±6 Tootals 201±2 total solids 223, 227, 228 training 82 transfer (sublimation) printing 4, 201±2, 252, 344±5 disperse dye inks 205, 245±6, 247 paper for 213 transition rate and length 57 transparency 25±6 trend analysis 301±4 trend extrapolation 301 triacetate 234 `trompe l'oeil style' 346±7 turnaround time 253 TWAIN protocol 150 Twister 11 two-phase method 201, 202, 207 Tx series 98±122 market needs for digital textile printing 101

368

Index

marketing profile 99±100 technical issues and solutions 101±20 colour reproduction 106, 107 fastness 109±10, 111±13 feed mechanism and applicability of fibres 110±15 high resolution images 101±6 inkjet stability and reliability 116±18, 119 prevention of bleeding 106±8 productivity 118±20 running cost 116, 117 strike-through 108±9 Tx-1600S 99 Tx2-1600 99, 114, 115 Tx3-1600 9, 99, 100, 114±15, 117±18, 341 unattended operation 120 upholstery 321, 322 urea 124, 126±42, 207, 209, 240±1 US Screen Printing Institute (USSPI) 10 van der Waals bonds 233 variable dot size 102 vast customisation 352 VideoJet 48, 50 9600 machine 3 vinylon 234 vinylsulphone (VS) reactive dyes 241±2, 243 Virtu system 8 viscosity 54±5, 203, 222±3, 254 disperse dye inks 249±50 and penetration length 106±7 testing 227, 228 viscosity control agents (rheology modifiers) 221, 225, 226 Viscotecs digital system 5, 351 vision 163 Vutek 5 warp sizes 123 washing 276, 279, 320±1 colour fastness to 227, 231 cationisation 280, 281, 282, 283±4, 287, 288

integrated fabric formation and colouration 127, 130 post-treatment 214 screen printing 20±1 waste, elimination of 313±14 waste ink 80 water/alcohol wicking ratio 261, 262, 265, 266 weaving 123, 124 see also integrated fabric formation and colouration `web-safe' colours 175 Weber number 58, 237 wet expansion/shrinkage fabrics 110±15 wettability 203±4 white ground staining 279, 280, 281, 282, 283, 284, 285, 287 white ink 231±2 wicking 256, 258, 261, 262, 265, 266 wide-format colour inkjet printers (WFP) 98 wool 210±11, 234 workflow 337±55 evolution 338±43 conventional textile printing workflow 338±40 digital textile strike-off printing workflow 340±1 full digital textile printing production workflow 341±3 future trends 354 new definitions 350±4 new design styles 343±50 woven fabrics 254, 261, 262, 263, 264, 265±72 Xaar 48, 51 Xerox 3, 48, 51 yard goods 327 yarn size 263, 264 yield stress 225 Zimmer 4 Chromojet system 9, 202, 337 Chromotex system 9, 161, 202