Additive Manufacturing: Design, Processes and Applications 3031337921, 9783031337925

Citation preview

SpringerBriefs in Applied Sciences and Technology Panagiotis Stavropoulos

Additive Manufacturing: Design, Processes and Applications

SpringerBriefs in Applied Sciences and Technology

SpringerBriefs present concise summaries of cutting-edge research and practical applications across a wide spectrum of fields. Featuring compact volumes of 50 to 125 pages, the series covers a range of content from professional to academic. Typical publications can be: • A timely report of state-of-the art methods • An introduction to or a manual for the application of mathematical or computer techniques • A bridge between new research results, as published in journal articles • A snapshot of a hot or emerging topic • An in-depth case study • A presentation of core concepts that students must understand in order to make independent contributions SpringerBriefs are characterized by fast, global electronic dissemination, standard publishing contracts, standardized manuscript preparation and formatting guidelines, and expedited production schedules. On the one hand, SpringerBriefs in Applied Sciences and Technology are devoted to the publication of fundamentals and applications within the different classical engineering disciplines as well as in interdisciplinary fields that recently emerged between these areas. On the other hand, as the boundary separating fundamental research and applied technology is more and more dissolving, this series is particularly open to trans-disciplinary topics between fundamental science and engineering. Indexed by EI-Compendex, SCOPUS and Springerlink.

Panagiotis Stavropoulos

Additive Manufacturing: Design, Processes and Applications

Panagiotis Stavropoulos Laboratory for Manufacturing Systems and Automation Department of Mechanical Engineering and Aeronautics University of Patras Patras, Greece

ISSN 2191-530X ISSN 2191-5318 (electronic) SpringerBriefs in Applied Sciences and Technology ISBN 978-3-031-33792-5 ISBN 978-3-031-33793-2 (eBook) https://doi.org/10.1007/978-3-031-33793-2 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

To my father

Preface

Additive manufacturing (AM) is one of the fastest-growing and most promising manufacturing technologies, offering significant advantages over conventional manufacturing processes. However, the wide adoption of AM is hindered by the lack of experience in the utilization of its advantages, as well as the minimization of its drawbacks in the various stages of production. Toward this goal, the existing design and manufacturing strategies must be enriched accordingly to ensure the successful application of AM. In this book, a holistic framework for AM is analyzed, consisting of three pillars: design, processes, and applications, focusing on AM of metals. Practical guidelines are set toward a catholic AM-driven design framework capable of exploiting the AM process advantages [1]. The proposed approach utilizes the full design freedom potentials of AM with a linear design flow reducing design iterations and ultimately achieving first-time-right AM design process. Design guidelines for specific AM process groups are set, and connection with simulation and distortion compensation is considered, as well as algorithm-supported design optimization. The result is a modular, metric-supported methodology for the re-design of conventional products to more efficient ones, considering the buildability restrictions [1], cost and weight reduction, optimized material distribution based on the load case, and support minimization [2]. Regarding the second pillar, a process-centric approach is followed, covering the different optimization strategies for AM, aiming to increase product quality, energy efficiency, and build-time minimization. Initially, the available AM methods are classified based on their process mechanisms, analyzing the characteristics, advantages, and drawbacks of each [3], as well as criteria for the selection of the most suitable [4]. AM-specific modeling strategies are categorized according to process parameters and Key Performance Indicators (KPIs) for each AM process group [5]. Additionally, a simulation approach toward more effective process parameters [6] and path-planning selections is analyzed [7]. Issues regarding the successful monitoring, process control, and quality assessment [8] strategies are described utilizing a novel open-source toolkit [9]. Moreover, the connection of the aforementioned modules (modeling, sensorization, diagnostic, and prognostic functions [10]) toward digital

vii

viii

Preface

twin [11] and IoT applications [12], [13] is described. Finally, hybrid AM applications are discussed [14]. The roadmap for the industrial application of AM is paved by addressing the challenges hindering it, through a holistic solution framework, which is tailored to the needs of the industry, aiming to assist in the evaluation and eventual uptake of AM. An AM cell-based solution is proposed, and the development of a hybrid AM production line is also discussed, along with the aspects that must be taken into consideration for the enhancement of the quality, flexibility, and productivity toward the automated production of net parts of high complexity and low cost [14]. Moreover, the integration of the different phases of the development process into a model-based design platform for decentralized manufacturing utilizing Industry 4.0 capabilities is realized [15]. Additionally, the integration of different modules under the concept of a digital twin is described, aiming to meet diverse requirements, such as adaptivity, real-time optimization, and uncertainty management [16]. Finally, the entire supply chain of the AM equipment, operation, and end-of-life based on real data from the design and operation of a demonstration plant and aerospace cases are discussed toward less expensive, more energy-efficient, environmentally friendlier, and reconfigurable manufacturing alternatives utilizing the advantages of AM [17]. Patras, Greece

Panagiotis Stavropoulos

Acknowledgments Members of the Laboratory for Manufacturing Systems and Automation (LMS) of University of Patras and Prima Additive have been involved in this project. My sincere gratitude goes to Dr. Harry Bikas, Dr. Panagis Foteinopoulos, and Dr. Ioannis Stavridis for their support.

References 1. H. Bikas, A.K. Lianos, P. Stavropoulos, A design framework for additive manufacturing. Int. J. Adv. Manufact. Technol. 103, 1–15 (2019). https://doi.org/10.1007/s00170-019-03627-z 2. H. Bikas, J. Stavridis, P. Stavropoulos, G. Chryssolouris, A design framework to replace conventional manufacturing processes with additive manufacturing for structural components: a formula student case study. Procedia CIRP 57, 710–715 (2016). https://doi.org/10.1016/j. procir.2016.11.123 3. A.K. Lianos, H. Bikas, P. Stavropoulos, A shape optimization method for part design derived from the buildability restrictions of the directed energy deposition additive manufacturing process. Designs 4(3), 19 (2020). https://doi.org/10.3390/designs4030019 4. H. Bikas, P. Stavropoulos, G. Chryssolouris, Additive manufacturing methods and modelling approaches: a critical review. Int. J. Adv. Manufact. Technol. 83(1–4), 389–405 (2016). https:/ /doi.org/10.1007/s00170-015-7576-2 5. H. Bikas, N. Porevopoulos, P. Stavropoulos, A decision support method for knowledge-based additive manufacturing process selection. Procedia CIRP 104, 1650–1655 (2021). https://doi. org/10.1016/j.procir.2021.11.278 6. P. Stavropoulos, P. Foteinopoulos, Modelling of additive manufacturing processes: a review and classification. Manufact. Rev. 5, 2 (2018). https://doi.org/10.1051/mfreview/2017014

Preface

ix

7. P. Foteinopoulos, A. Papacharalampopoulos, P. Stavropoulos, On thermal modeling of additive manufacturing processes. CIRP J. Manufact. Sci. Technol. 20, 66–83 (2018). https://doi.org/ 10.1016/j.cirpj.2017.09.007 8. P. Foteinopoulos, A. Papacharalampopoulos, K. Angelopoulos, P. Stavropoulos, Development of a simulation approach for laser powder bed fusion based on scanning strategy selection. Int. J. Adv. Manufact. Technol. 108(9), 3085–3100 (2020). https://doi.org/10.1007/s00170020-05603-4 9. P. Stavropoulos, A. Papacharalampopoulos, J. Stavridis, K. Sampatakakis, A three-stage quality diagnosis platform for laser-based manufacturing processes. Int. J. Adv. Manufact. Technol. 110(11), 2991–3003 (2020). https://doi.org/10.1007/s00170-020-05981-9 10. A. García-Díaz, V. Panadeiro, B. Lodeiro, J. Rodríguez-Araújo, J. Stavridis, A. Papacharalampopoulos, P. Stavropoulos, OpenLMD, an open source middleware and toolkit for laser-based additive manufacturing of large metal parts. Robot Comput-Integr. Manufact. 53, 153–161 (2018). https://doi.org/10.1016/j.rcim.2018.04.006 11. A. Papacharalampopoulos, C.K. Michail, P. Stavropoulos, Manufacturing resilience and agility through processes digital twin: design and testing applied in the LPBF case. Procedia CIRP 103, 164–169 (2021). https://doi.org/10.1016/j.procir.2021.10.026 12. A. Papacharalampopoulos, C. Giannoulis, P. Stavropoulos, D. Mourtzis, A digital twin for automated root-cause search of production alarms based on KPIs aggregated from IoT. Appl. Sci. 10(7), 2377 (2020). https://doi.org/10.3390/app10072377 13. A. Papacharalampopoulos, J. Stavridis, P. Stavropoulos, G. Chryssolouris, Cloud-based control of thermal based manufacturing processes. Procedia CIRP 55, 254–259 (2016). https:/ /doi.org/10.1016/j.procir.2016.09.036 14. T. Souflas, H. Bikas, M. Ghassempouri, A. Salmi, E. Atzeni, A. Saboori, P. Stavropoulos, A comparative study of dry and cryogenic milling for directed energy deposited IN718 components: effect on process and part quality. Int. J. Adv. Manufact. Technol. 119(1), 745–758 (2022). https://doi.org/10.1007/s00170-021-08313-7 15. P. Stavropoulos, P. Foteinopoulos, A. Papacharalampopoulos, H. Bikas, Addressing the challenges for the industrial application of additive manufacturing: towards a hybrid solution. Int. J. Lightweight Mater. Manufact. 1(3), 157–168 (2018). https://doi.org/10.1016/j.ijlmm.2018. 07.002 16. P. Stavropoulos, D. Mourtzis, Digital Twins in Industry 4.0. In Design and Operation of Production Networks for Mass Personalization in the Era of Cloud Technology (Elsevier, Amsterdam, 2022), p. 277 17. P. Stavropoulos, A. Koutsomichalis, N. Vaxevanidis, Laser-based Manufacturing Processes for Aerospace Applications. In Materials Science and Engineering: Concepts, Methodologies, Tools, and Applications (IGI Global, 2017), p. 374

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Definition and Short Historical Review . . . . . . . . . . . . . . . . . . . . . . . . 1.2 AM Basic Steps and Process Families . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 AM Today, Challenges, and Shortcomings—Framework of This Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 2

2 Design for AM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Introduction to Design for AM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Design Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1 Geometric Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Build Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3 Process Capability Determination . . . . . . . . . . . . . . . . . . . . . . 2.3 Design Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.1 Accuracy (XY-Plane Versus Z-axis) . . . . . . . . . . . . . . . . . . . . 2.3.2 Anisotropic Mechanical Properties . . . . . . . . . . . . . . . . . . . . . 2.3.3 Shrinkage and Warping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.4 Surface Roughness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.5 Build Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Additive Manufacturability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 Re-design for AM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6 Design Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.7 Design for Post-processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.8 AM Relevant Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7 7 9 10 22 27 27 29 29 30 30 31 31 32 36 38 39 40

3 AM Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Process and Material Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 AM Part Quality Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 Surface Roughness and Layer-by-Layer Appearance . . . . . . 3.2.2 Porosity/Void Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.3 Anisotropic Microstructure and Mechanical Properties . . . .

45 45 49 50 52 53

4 5

xi

xii

Contents

3.2.4 Thermal Residual Stresses and Deformations . . . . . . . . . . . . 3.3 Process Optimization Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Heating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2 Scanner Head Speed and Scanning Strategy . . . . . . . . . . . . . . 3.3.3 Layer Thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 Simulation for AM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.1 Modeling Approaches and Scales . . . . . . . . . . . . . . . . . . . . . . 3.4.2 Simulations of Melt Pool-Related Phenomena . . . . . . . . . . . . 3.4.3 Thermo-Mechanical Simulations . . . . . . . . . . . . . . . . . . . . . . . 3.4.4 Multi-scale Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 Monitoring, Control, and Quality Assessment . . . . . . . . . . . . . . . . . . 3.5.1 Monitoring and Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.2 Quality Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6 Post-processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.1 Cleaning Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.2 Mechanical Material-Removal Post-processes . . . . . . . . . . . . 3.6.3 Thermal Post-processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.4 Electrothermal Post-processes . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.5 Chemical Post-processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.6 Laser-Based Post-processes . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7 Special Topics: Hybrid AM and Digitals Twins . . . . . . . . . . . . . . . . . 3.7.1 Hybrid AM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7.2 Digital Twins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

54 57 57 61 63 65 65 69 70 74 75 75 77 80 81 82 82 85 85 86 87 87 87 89

4 AM Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Introduction to AM Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Production of End-Use AM Parts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Business Perspective and AM Case Studies . . . . . . . . . . . . . . . . . . . . 4.3.1 Aerospace Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.2 Dental Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.3 Power and Energy Sector Case . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

95 95 101 103 103 109 114 119

5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Abbreviations

3D AHP AM BJ CAD CCD CDLP CFD CT DED DfAM DLD DMD DMLS DOD DR EBAM EBM EDM ESA FDM GD GFE HIP HMI KPIs LBM LENS LIBS LMD LMJ

Three-Dimensional Advanced Hierarchy Process Additive Manufacturing Binder Jetting Computer-Aided Design Charge-Coupled Device Continuous Digital Light Processing Computational Fluid Mechanics Computed Tomography Directed Energy Deposition Design for Additive Manufacturing Direct Laser Deposition Direct Metal Deposition Direct Metal Laser Sintering Drop on Demand Digital Radiography Electron Beam Additive Manufacturing Electron Beam Melting Electrical Discharge Mach European Space Agency Fused Deposition Modeling Generative Design Geometrical Feature Extraction Hot Isostatic Pressing Human–Machine Interface Key Performance Indicators Laser Beam Melting Laser Engineering Net Shaping Laser-Induced Breakdown Spectroscopy Laser Metal Deposition Liquid Metal Jetting xiii

xiv

LOM MAM ME MJ MJF MJP ML NDT PBF PCA PCRT ROI RP RTP SFTI SL SLA SLM SLS SME STL SVM TO VP

Abbreviations

Laminated Object Manufacturing Metal AM Material Extrusion Material Jetting Multi Jet Fusion Multi Jet Printing Machine Learning Non-destructive Testing Powder Bed Fusion Principle Component Analysis Process Compensated Resonant Testing Return on Investment Rapid Prototyping Room Temperature and Pressure Stress Formation Tendency Index Sheet Lamination Stereolithography Selective Laser Melting Selective Laser Sintering Small–Medium Enterprise Standard Triangle Language Support Vector Machine Topology Optimization Vat Photopolymerization

Chapter 1

Introduction

The definition and a short historical review of AM will be presented, followed by the importance of AM and the advantages it offers. The basic steps for the manufacturing of a product with AM will be presented, as well as a classification of the different AM process families. Moreover, the market role of AM and the prerequisites for its fruitful utilization will be discussed. Finally, the framework of this book will be analyzed.

1.1 Definition and Short Historical Review In Additive Manufacturing (AM) processes, parts are created in a layer-by-layer fashion by selectively fusing the material of the current layer on that of the previous one, based on information provided by 3D model data [1]. Rapid prototyping (RP), the predecessor of AM, was developed in 1981, and in 1984, the patent for Stereolithography was created. The main difference between the two is that RP is aimed at the manufacturing of prototypes, while AM mostly focuses on end-user parts [2]. The first metal AM application was Selective Laser Sintering, which became commercially available in 2006 [3]. The most important advantage of AM technology is that design complexity has an almost zero impact on cost [4, 5] (Fig. 1.1). The implications of this are of crucial importance, with the potential to revolutionize the way products are designed in many industrial sectors [6]. AM rendered viable in terms of cost the use of topology optimization, generative algorithms, and the utilization of cellulite structures, allowing for highly optimized products of high value, which can also combine complex assemblies in a single part [5]. Another vital edge of AM is that it constitutes the production of small product sizes, even of a single unique product, cost-effective, paving the

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 P. Stavropoulos, Additive Manufacturing: Design, Processes and Applications, SpringerBriefs in Applied Sciences and Technology, https://doi.org/10.1007/978-3-031-33793-2_1

1

2

1 Introduction

Fig. 1.1 Manufacturing cost as a function of design complexity in conventional manufacturing and AM

way for mass customization [7]. Last but not least is the environmental friendliness of AM, due to the lower material waste, as well as energy consumption, it offers in comparison with conventional manufacturing processes [4].

1.2 AM Basic Steps and Process Families The first step for manufacturing a product using AM is the creation of its design using a three-dimensional computer-aided design (3D CAD) program. The CAD file must then be imported into a slicing software, which breaks down the design into layers and the user selects the orientation and layer thickness. The rest of the process parameters are determined either in that step or in machine-specialized software. The build file of the part (usually G-Code) is then created. This file is ready to be imported into the AM machine, and the manufacturing of the product can start once the machine setup is complete. After the manufacturing of the part is finished, the post-processing takes place, which is different according to the AM process, as well as the desired product characteristics. Finally, the quality control of the ready product takes place (Fig. 1.2). In Table 1.1, AM processes have been classified into different process families along with the most common commercial names [3] and processable materials feedstock forms [4]. It has to be noted that Powder Bed Fusion and Directed Energy Deposition are the most commonly used process families for the manufacturing of metal parts, followed by Binder Jetting, although the latter requires much more extensive post-processing for metal end-use parts (Fig. 1.3).

1.2 AM Basic Steps and Process Families

3

Fig. 1.2 Basic steps for manufacturing a product with AM

Table 1.1 Classification of AM processes and typical commercial names [3] AM process group

Typical commercial names

Vat Photopolymerization (VP)

Stereolithography (SLA), digital light processing, solid ground curing, projection stereolithography

Powder bed fusion (PBF)

Electron beam melting (EBM), electron beam additive manufacturing (EBAM), selective laser sintering (SLS), selective heat sintering, direct metal laser sintering (DMLS), selective laser melting (SLM), laser beam melting (LBM)

Directed energy deposition (DED)

Laser metal deposition (LMD), direct metal deposition (DMD), direct laser deposition (DLD), laser engineered net shaping, electron beam freeform fabrication, weld-based additive manufacturing

Binder jetting (BJ)

Powder bed and inkjet head, plaster-based 3D printing

Material extrusion (MEx)

Fused deposition modeling (FDM), fused filament fabrication

Material jetting (MJ)

Multi jet modeling, aerosol jet, ballistic particle manufacturing, drop on demand (DOD), laser-induced forward transfer, liquid metal jetting (LMJ), multi jet printing (MJP), nanometal jetting, nanoparticle jetting, polyjet, printoptical technology

Sheet lamination (SL)

Laminated object manufacturing (LOM), ultrasonic consolidation

AM for the construction sector

Cement AM/3D printing, concrete AM/3D printing

4

1 Introduction

Fig. 1.3 Classification of AM process and processable materials feedstock forms [4]

1.3 AM Today, Challenges, and Shortcomings—Framework of This Book AM is a promising technology, capable of offering profitable solutions in today’s market. However, its uptake is hindered by the lack of experience in the utilization of its advantages, as well as the minimization of its drawbacks in the various stages of production [8]. This book is a modular guide integrating the stages of AM product development in a practical, reader-friendly approach, aiming for the wider adoption of AM, focusing on AM of metals. Toward this goal, the existing design and manufacturing strategies are enriched in a step-by-step approach forming a holistic framework divided into three pillars: design, processes, and applications, forming an essential aid for researchers, designers, engineers, simulation specialists, and post-graduate students. Specific design rules must be followed for the maximization of the added value of AM products [9]. Toward this end, an AM-driven design framework is analyzed, exploiting the AM process advantages through design optimization, connection with simulation, and distortion compensation. Furthermore, a metric-supported methodology is presented for the re-design of conventional products to more efficient AM ones, considering buildability restrictions [10], cost and weight reduction, optimized material distribution based on the load case, and support minimization [11]. Product quality is one of the most important issues of AM [8, 12]. In the second pillar of this framework, a process-centric approach is followed, covering optimization strategies for AM aiming for enhanced quality. Initially, the available AM methods are classified based on their process mechanisms, analyzing the characteristics, advantages, and drawbacks of each, as well as criteria for the selection of the most suitable according to specific needs [13]. Existing models for AM are categorized [14] according to process parameters and Key Performance Indicators (KPIs), and a simulation-based approach for KPIs enhancement is analyzed, considering

References

5

important process parameters [15] and scanning strategy selections [12]. Additionally, AM-specific monitoring, process control, and quality assessment strategies [14, 16, 17] are discussed. Finally, the connection of the aforementioned modules toward hybrid AM [18] and digital twin [19, 20] applications is described. The roadmap for the industrial application of AM is paved by addressing the challenges hindering it, aiming to assist in the evaluation and industrial uptake of AM [21]. The different phases of product development are integrated, encapsulating the entire supply chain of AM equipment, operation, and end-of-life toward a successful business plan [22]. Also, specific AM case studies are considered [17, 22]. Additionally, hybrid AM cell and production line setups are discussed, toward net products of high quality and low cost, ensuring flexibility, and productivity [21]. Finally, Industry 4.0 capabilities [23, 24] toward decentralized manufacturing are presented.

References 1. A.S.T.M. Standard, Standard terminology for additive manufacturing technologies, in ASTM International F2792-12a (2012) 2. N. Hopkinson, R. Hague, P. Dickens, Rapid Manufacturing: An Industrial Revolution for the Digital Age (John Wiley & Sons, Chichester, England, 2006) 3. P. Foteinopoulos, Computational and empirical modeling of additive manufacturing processes. Dissertation, University of Patras, 2021. https://doi.org/10.12681/eadd/50503 4. H. Bikas, P. Stavropoulos, G. Chryssolouris, Additive manufacturing methods and modelling approaches: a critical review. Int. J. Adv. Manuf. Technol. 83(1–4), 389–405 (2016). https:// doi.org/10.1007/s00170-015-7576-2 5. P. Stavropoulos, P. Foteinopoulos, J. Stavridis, H. Bikas, Increasing the industrial uptake of additive manufacturing processes a training framework. Adv. Ind. Manuf. Eng. Available at SSRN 4169003 (2022) 6. G.A. Adam, D. Zimmer, Design for additive manufacturing—element transitions and aggregated structures. CIRP J. Manuf. Sci. Technol. 7(1), 20–28 (2014). https://doi.org/10.1016/j. cirpj.2013.10.001 7. H. Bikas, A.K. Lianos, P. Stavropoulos, A design framework for additive manufacturing. Int. J. Adv. Manuf. Technol. 103, 1–15 (2019). https://doi.org/10.1007/s00170-019-03627-z 8. P. Stavropoulos, P. Foteinopoulos, Modelling of additive manufacturing processes: a review and classification. Manuf. Rev. 5, 2 (2018). https://doi.org/10.1051/mfreview/2017014 9. H. Bikas, A.K. Lianos, P. Stavropoulos, A design framework for additive manufacturing. Int. J. Adv. Manuf. Technol. 103(9), 3769–3783 (2019). https://doi.org/10.1007/s00170-019-036 27-z 10. H. Bikas, J. Stavridis, P. Stavropoulos, G. Chryssolouris, A design framework to replace conventional manufacturing processes with additive manufacturing for structural components: a formula student case study. Procedia CIRP 57, 710–715 (2016). https://doi.org/10.1016/j.pro cir.2016.11.123 11. A.K. Lianos, H. Bikas, P. Stavropoulos, A shape optimization method for part design derived from the buildability restrictions of the directed energy deposition additive manufacturing process. Designs 4(3), 19 (2020). https://doi.org/10.3390/designs4030019 12. P. Foteinopoulos, A. Papacharalampopoulos, K. Angelopoulos, P. Stavropoulos, Development of a simulation approach for laser powder bed fusion based on scanning strategy selection. Int. J. Adv. Manuf. Technol. 108(9), 3085–3100 (2020). https://doi.org/10.1007/s00170-020-056 03-4

6

1 Introduction

13. H. Bikas, N. Porevopoulos, P. Stavropoulos, A decision support method for knowledge-based Additive Manufacturing process selection. Procedia CIRP 104, 1650–1655 (2021). https://doi. org/10.1016/j.procir.2021.11.278 14. P. Stavropoulos, A. Papacharalampopoulos, J. Stavridis, K. Sampatakakis, A three-stage quality diagnosis platform for laser-based manufacturing processes. Int. J. Adv. Manuf. Technol. 110(11), 2991–3003 (2020). https://doi.org/10.1007/s00170-020-05981-9 15. P. Foteinopoulos, A. Papacharalampopoulos, P. Stavropoulos, On thermal modeling of additive manufacturing processes. CIRP J. Manuf. Sci. Technol. 20, 66–83 (2018). https://doi.org/10. 1016/j.cirpj.2017.09.007 16. A. García-Díaz, V. Panadeiro, B. Lodeiro, J. Rodríguez-Araújo, J. Stavridis, A. Papacharalampopoulos, P. Stavropoulos, OpenLMD, an open source middleware and toolkit for laserbased additive manufacturing of large metal parts. Robot. Comput.-Integr. Manuf. 53, 153–161 (2018). https://doi.org/10.1016/j.rcim.2018.04.006 17. A. Papacharalampopoulos, C.K. Michail, P. Stavropoulos, Manufacturing resilience and agility through processes digital twin: design and testing applied in the LPBF case. Procedia CIRP 103, 164–169 (2021). https://doi.org/10.1016/j.procir.2021.10.026 18. T. Souflas, H. Bikas, M. Ghassempouri, A. Salmi, E. Atzeni, A. Saboori, P. Stavropoulos, A comparative study of dry and cryogenic milling for directed energy deposited IN718 components: effect on process and part quality. Int. J. Adv. Manuf. Technol. 119(1), 745–758 (2022). https://doi.org/10.1007/s00170-021-08313-7 19. P. Stavropoulos, D. Mourtzis, Digital twins in industry 4.0, in Design and Operation of Production Networks for Mass Personalization in the Era of Cloud Technology (Elsevier, Amsterdam, 2022), p. 277 20. A. Papacharalampopoulos, C. Giannoulis, P. Stavropoulos, D. Mourtzis, A digital twin for automated root-cause search of production alarms based on KPIs aggregated from IoT. Appl. Sci. 10(7), 2377 (2020). https://doi.org/10.3390/app10072377 21. P. Stavropoulos, P. Foteinopoulos, A. Papacharalampopoulos, H. Bikas, Addressing the challenges for the industrial application of additive manufacturing: towards a hybrid solution. Int. J. Lightweight Mater. Manuf. 1(3), 157–168 (2018). https://doi.org/10.1016/j.ijlmm.2018.07.002 22. P. Stavropoulos, A. Koutsomichalis, N. Vaxevanidis, Laser-based manufacturing processes for aerospace applications, in Materials Science and Engineering: Concepts, Methodologies, Tools, and Applications (IGI Global, 2017), p. 374 23. K. Ntouanoglou, P. Stavropoulos, D. Mourtzis, 4D printing prospects for the aerospace industry: a critical review. Procedia Manuf. 18, 120–129 (2018). https://doi.org/10.1016/j.promfg.2018. 11.016 24. A. Papacharalampopoulos, J. Stavridis, P. Stavropoulos, G. Chryssolouris, Cloud-based control of thermal based manufacturing processes. Procedia CIRP 55, 254–259 (2016). https://doi.org/ 10.1016/j.procir.2016.09.036

Chapter 2

Design for AM

Design for AM will be defined, and the main factors to be taken into account by a designer to fully exploit the AM process advantages will be discussed. Design aspects will be defined, and AM process-imposed limitations will be presented in detail. Design considerations will also be defined and the mutual relationship between these considerations and design choices will be discussed. Additive Manufacturability will be defined, and a framework for determining Additive Manufacturability will be presented. Approaches for re-designing existing components of conventional products to more efficient AM ones will be presented, followed by a summary of algorithmic design optimization approaches. Finally, design decisions that affect the effectiveness of downstream processes will be discussed, and relevant existing standards for AM design will be presented.

2.1 Introduction to Design for AM Owing to its inherent design freedom benefits, Additive Manufacturing (AM) is becoming more and more popular; however, most design engineers treat the field of AM as “expensive milling machines”, maintaining a design ethos stemming from established manufacturing processes. Furthermore, by taking a look at the design and production flow of AM nowadays, it is obvious that the design phase is completely disconnected from the machine programming and production phase [1]. The definition of design methods varies across industries, from the design of ideas and concepts to detailed technical and manufacturing drawings [2, 3]. Mechanical design of parts and assemblies has significantly progressed over the last 30 years, from manual drawings to parametric Computer-Aided Design suites (CAD) that enable enhanced design functions [4]. Over the last decade, engineering design

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 P. Stavropoulos, Additive Manufacturing: Design, Processes and Applications, SpringerBriefs in Applied Sciences and Technology, https://doi.org/10.1007/978-3-031-33793-2_2

7

8

2 Design for AM

has further evolved from a feature-based to a function-based mentality; state of art approaches nowadays dictate that the parts should be morphed based on their functional requirements and loading state [5]. Nevertheless, current design practices do not fully exploit the design freedom inherent in AM. Under the current perspective, the engineer designs a part as per their perception of good engineering practices by establishing a certain part geometry. Design optimization algorithms (such as Topology Optimization) may also be deployed at this stage. Only when the design is finalized, do the production aspects come to play, through the use of a machine (or manufacturer) specific software, which slices the part and generates the actual part program. In this step, there are usually options for surface modification/patterns and part nesting. Then, according to user input, the part is sliced into layers, and support structures are generated on the basis of certain criteria and algorithms. The final result and the efficient use of AM technology are highly dependent on the designers’ familiarity and experience with such processes. Design for Additive Manufacturing (DfAM) refers to a set of design rules and guidelines aiming to “Maximize product performance through the synthesis of shapes, sizes, hierarchical structures, and material compositions, subject to the capabilities of AM technologies” [6]. Design for AM is a very complex subject, where numerous factors and their interactions should be taken into account in an attempt to maximize the benefit occurring by utilizing AM for a specific part/use case. This is because the design of every component to be additively manufactured can be affected and in parallel has effects both in the AM process and the materials selected, as well as on the subsequent post-processing steps required to obtain a ready-to-use component. In general, design for AM should focus on designing a component in such a way that using AM can add value to it, compared with manufacturing using conventional processes. However, design opportunities stemming from the decision to use AM are not limited to the part itself; such opportunities for added value through AM may arise through different aspects of the component manufacturing and use, as summarized hereafter. • • • •

Through product design (improved product function and/or performance). Through process design (ease of manufacturing, reduced lead time, and/or cost). Through lifecycle (improved performance of functionality during product use). Through business models and digitalization (production design).

In order to be able to successfully implement design for AM and deliver value through its use of it, the designer should not only consider the strong points of each AM process but also has in mind their inherent drawbacks. All AM technologies come along with numerous manufacturing limitations [7]. When designing for AM, one should take into account numerous factors that can contribute to the final result. These factors can be “global” (i.e., applicable across all existing AM processes and orientating mainly to the layer-wise nature of AM) or process-specific, owing mainly to limitations related to the physical process mechanism. Nevertheless, the limit values for both types of factors are also dependent on the process, material, and even the specific machine and process parameters. These

2.2 Design Aspects

9

limitations should be taken into account by the designer during the early stages of a part’s design, as non-compliance is causing bottlenecks to the AM process. However, usually, these limitations emerge during the latest phases of a part design, which requires multiple design iterations until complete adherence to the AM rules for manufacturability. Supplementary features need to be added, or others suppressed so that the designed part can be successfully manufactured using the selected AM process, material, and machine. The following section will discuss existing DfAM guidelines and best practices. As design guidelines for AM are highly dependent on the process and process mechanism, materials used, and (to a lesser extent) machine, only broad guidelines can be given, and fine-tuning of the design limitations for a specific machine and/or material requires targeted and structured experimentation based on the limiting factors expected by the process mechanism, expert knowledge, and a large volume of data. To be able to classify limitations in a meaningful way, the terms design aspect and design consideration needs to be defined. A design aspect is defined as any particular feature which can be quantified during the design phase, that includes geometric features of the part (overhangs, bores, channels, walls, etc.), as well as relevant build parameters that need to be set in order to manufacture the part (layer thickness, build orientation, etc.). Design consideration is the effects of the process mechanism on the manufactured part. These considerations can be very specific properties of the process and quantified with certain KPIs (Fig. 2.1). Additive Manufacturability will be defined, and a framework for determining Additive Manufacturability will be presented. Design for AM approaches will be subsequently presented. There are two main approaches that could be exploited when it comes to designing a part for Additive Manufacturing (Fig. 2.2). The first approach involves re-designing or adapting a pre-existing part in order to make it suitable for Additive Manufacturing, while the second approach involves creating an optimized design from scratch, using algorithmic methods based on AM process capabilities. Both approaches need to be adapted to the specific AM process to be used, taking into account the design considerations/limitations applicable per process. Finally, design decisions that affect the effectiveness of downstream processes will be discussed, and relevant existing standards for AM design will be presented.

2.2 Design Aspects The design aspects linked to the design for AM processes are briefly presented in the following section. We can distinguish design aspects into two main categories: part’s geometric features and build parameters.

10

2 Design for AM

Fig. 2.1 Defining aspects and considerations for AM design [8]

2.2.1 Geometric Features Similar to conventional manufacturing, there are restrictions regarding the geometries that can be built using AM processes. The layer-by-layer principle followed by all AM machines has its limitations since each layer must be built almost directly above the previous one, for the build to not collapse [9]. The result is that not every geometry is possible as each geometrical feature must obey a certain geometrical continuity. Once this geometric continuity is overlooked in the design, the resulting part will suffer in its integrity (e.g., deformation, porous mass, reduced density). The design aspects that determine the quality of the outcome as well as guidelines for design best practices are presented in the following section. Overhanging Geometries Manufacturing overhanging geometries is a trite, yet challenging feature for the majority of the AM processes. A generic definition of an overhanging geometry is any geometry, whose orientation is not parallel to the build vector. The ability of the AM machine to manufacture a layer of material displaced to the previous defines its ability to create overhanging geometries. The magnitude of this layer parallel shift

2.2 Design Aspects

11

Fig. 2.2 Different design approaches to DfAM

sets the limit for the maximum overhang length and subsequently the maximum inclination angle that can be achieved. In general, overhanging geometries can be classified into four categories: overhangs, angled overhangs, bridges, and bores/ channels as illustrated in Fig. 2.3. Horizontal Overhangs Horizontal overhangs are one-sided abrupt geometrical changes, that resemble a cantilever beam. The horizontal distance that an AM machine can build without supports is limited and if exceeded, the whole build could fail. The limit of an

Fig. 2.3 Overhanging geometries’ categories. Adapted from [8]

12

2 Design for AM

overhang length is affected by numerous factors and the nature of the AM technology [10–12]. The AM process, the material used, and even the actual machine are variables in the equation that defines the maximum overhang length. Indicative overhanging lengths per AM technology as found in the state of art and state of practice review are presented in Table 2.1. When part specifications call for a greater overhang, the decision to be made is whether to alter the geometry of the part. The simplest way to resolve this issue is by replacing horizontal overhangs with angled ones. This effectively allows the creation of the overhang gradually over multiple layers of the part, thus reducing the overhanging distance. In case an angled overhang adaption is not feasible, due to the specifications and the geometry of the part, a support structure (further discussed in subsequent section) needs to be introduced to support the overhanging feature. Angled Overhangs Another category of overhanging geometries is the angled overhang. Due to the layerwise material deposition of all AM processes, angled overhangs can be correlated to horizontal overhanging distance as defined above; the maximum overhang angle is a product of the maximum overhanging distance over the thickness of a single layer (Fig. 2.4; Table 2.2). These numbers can diverge highly in certain parts whose surface quality is acceptable to be poor (on the down-facing areas of overhangs), and if the process parameters have been set up correctly [19]. For extrusion-based (MEx) AM technologies, extremely angled overhangs cannot be created, as the material cannot be deposited in mid-air [11]. For Powder Bed Fusion AM technologies, the powder surrounding the part acts as support and thus steeper angled overhangs can be realized. That is, there is a drawback regarding surface roughness as the surrounding powder is sintered unevenly on the downward-facing areas of the part (dross formation), as evident in Fig. 2.5. Bridging Similar to horizontal overhangs, a bridge is a horizontal geometry between two or more non-horizontal features that resemble a simply supported beam. A bridge is defined as any surface in the part geometry that is facing down between two or more features. Similarly to the previous restrictions, the designer must take into consideration the maximum length that the machine can bridge. If this length is exceeded, the part will not be successfully manufactured (Table 2.3). 1 Similarly to horizontal overhangs, when part specifications call for a larger bridge, the designer can introduce a “transition” angled overhang, reducing the bridging length (Fig. 2.6). If this is not possible, additional support structures may be introduced. Bores and Channels The ability to manufacture parts with hollow internal geometries is a major benefit for AM technologies since most of the time such features are impossible to achieve using conventional manufacturing methods, that enables profound geometrical flexibility

Overhangs

(mm)

[13]

[13]

[15]

10

Supports

0.5

5

7

[13]

N/A

CDLP

FDM

DLP

SLA

50

No need

Extrusion

Vat polymerization

Table 2.1 Overhang limits per AM technology

[13]

MJ [13]

MJP [13]

DOD

Material jetting

[16]

[13]

Binder jetting N/A

MJF

[13]

SLS

[13]

SLM

Powder bed fusion

N/A

EBM

[14]

LENS

N/A

EBAM

Direct energy deposition

[14]

LOM

Sheet

2.2 Design Aspects 13

14

2 Design for AM

allowing the creation of parts with significantly improved performance, such as increased heat convection capabilities, improved fluid flow [21, 22], or structural reinforcement/lattice structures [23, 24]. A characteristic example includes using AM to produce high-performance tooling for injection molding, by creating conformal cooling channels that follow complex paths very close to the surface of the tool, allowing much higher cooling rates and increasing productivity [21]. Bores and channels can be partially classified as overhanging geometries/bridges (Fig. 2.7); therefore, similar limitations regarding their buildability exist. It is worth mentioning that this category also includes non-vertical holes on the part (that are effectively short-length bores). Additional constraints should be taken into account since bores and channels are typically internal to the part; therefore, removal of unconsolidated material (powder, resin) or support structures presents a significant challenge. In addition, the AM process mechanism may result in unintended consolidation of material within the bore/channel, further limiting the capability of the process to produce very small and intricate cooling channels (Table 2.4). To improve the buildability of bores and channels, the up-facing surface can be designed as a non-round geometry; typically, a triangular shape with an overhang angle less than the maximum allowed by the process is used, resulting in a teardrop-like geometry. This transforms the bore/channel into a self-supporting structure and removes any need for post-processing if a non-round cross-section is permissible. Other common self-supporting cross-sections of bores/channels can be seen in Fig. 2.8. If a round cross-section is required, post-processing is needed. Particularly in the case of bores/holes, different approaches can be taken depending on the size of the bore/hole and the accuracy requirements. For very precise and small (less than 8 mm in diameter) holes, the approach is to remove them from the geometry to be created by AM and instead drill them from scratch at the post-processing stage. If accuracy requirements are not very strict, a bore/hole may be created in the AM part, but with a smaller diameter. This “pre-hole” will act as a locating feature for postprocessing, reducing drill forces, material waste, and time. When larger bores/holes are required, they are typically included in the AM part albeit undersized, and milling is used during post-processing to achieve the desired final diameter accurately.

Fig. 2.4 Stair-step effect

Angled overhangs

Supports

45°

55°

60°

No need

(Degrees)

[13]

[13]

[13]

FDM

N/A

CDLP

SLA

DLP

Extrusion

Vat polymerization

Table 2.2 Angled overhang limits per AM technology

NPJ [13]

MJ [13]

[14]

DOD

Material jetting

[17]

Binder jetting N/A

MJF SLS [18]

SLM

Powder bed fusion

N/A

EBM

[14]

LENS

N/A

EBAM

Direct energy deposition

[14]

LOM

Sheet

2.2 Design Aspects 15

16

2 Design for AM

Fig. 2.5 a Overhanging angles in a test part manufactured with SLM technology 0. b Overhanging angle-surface roughness [20]

Supports As already introduced, overhanging geometries can be successfully manufactured with the addition of support structures (Fig. 2.9). Support structures are sacrificial column-like geometries that are introduced during the build preparation phase (when the machine program is being generated), typically in an automated fashion by the respective software tool. Their purpose (as the name implies) is to offer support for the layers to be built on top, thus splitting the overhanging geometry into smaller bridges, making it manufacturable. While effective at guaranteeing a successful build, supports typically compromise the surface quality of the part; the reason is being that support structures and the part are overlapping [25]. In addition, supports increase the post-processing demands, as they need to be removed after the build is completed [11, 25]. To that end, we can distinguish between two types of supports: soluble supports which are typical of a different material to the main part, which can be dissolved by using water or a special solution in order to remove them, and hard supports which are built from the same material as the main part (albeit in a less dense way) and need to be mechanically removed from the part. Soluble supports can be met in Material Extrusion and Material Jetting processes, while hard supports are typically met in Vat Photopolymerization, metal Powder Bed Fusion, and sometimes in Material Extrusion. The latter type imposes an additional design restriction; tool accessibility for support removal should be ensured. These aspects are going to be presented in greater detail in the post-processing section of this chapter. In general, supports decrease the efficiency of the AM process; they add to the build time and require additional material to be built, while also dictating additional post-processing steps and equipment, increasing the overall time and cost needed, and contradicting the advantageous net-shape manufacturing nature of AM. Therefore, in order to increase AM Manufacturability, it is desirable that the part only has self-supported geometrical features if possible. A way of making the part self-supporting (besides the considerations already presented) is the substitution of temporary supports with permanent walls. The main benefit from this approach is that the newly added permanent wall does not need to be removed, thus reducing the overall required time and cost. In addition, it can make the part stronger, allowing

Supports

80

50

20

10

2

No need

(Degrees)

[13]

[13]

[13]

FDM

N/A

CDLP

SLA

DLP

Extrusion

Vat polymerization

[13]

MJ [13]

NPJ [13]

DOD

Material jetting

[13]

Binder jetting N/A

MJF

[13]

SLS [13]

SLM

Powder bed fusion

N/A

EBM

EBAM N/A

LENS 1

Direct energy deposition

N/A

LOM

Sheet

DED technologies can bridge long distances, as the part can typically be manipulated during the deposition process. This allows bridging a horizontal feature to be manufactured as a vertical one, effectively removing the bridging requirement

Bridging

Table 2.3 Bridging length limits per AM technology

2.2 Design Aspects 17

18

2 Design for AM

Fig. 2.6 a Transition angled overhang. b Intermediate support structure

Fig. 2.7 Bores or channels can be linked with overhanging geometry when manufactured in a non-vertical direction [8]

the removal of material from other areas, and making the design more efficient (Fig. 2.10). Re-orienting the part to change the build direction (therefore, what is considered an overhang as well as the magnitude of the overhanging angle) is a very efficient way of doing so, without introducing any additional design changes to the part. Orientation considerations will be presented in greater detail in the build parameters section of this chapter. Wall Thickness All AM processes have a minimum threshold on the wall thickness that is feasible to be manufactured. This is due to the building threshold determined by the fundamental building unit used by every AM machine—diameter of the laser beam, flow focal point, or nozzle—and the fact that the machine needs to make multiple passes to build a sufficient and solid feature. Another accountable parameter for thin walls is the height-to-thickness ratio, as oblong wall structures tend to collapse.1 Indicative minimum wall thicknesses for all AM processes are summarized in Table 2.5. An important path-planning aspect when designing thin walls close to the limits of the AM machine is that some geometries cannot be precisely depicted as the slicing software which generates the G-Code is not able to create the desired geometry [26]. Below the lower limit of the allowed thickness, the wall feature cannot be formed or when formed, will suffer from deformation [27]. As such, an integer multiple of 1

Typically 40:1 for metal AM, which needs to be abided for the skin not to collapse.

Bores and channels

5.0

4.0

3.0

2.0

1.5

1.25

0.75

0.5

0.4

Minimum diameter (mm)

SLA

DLP

N/A

CDLP

Vat polymerization FDM

Extrusion MJ

NPJ N/A

DOD

Material jetting

Table 2.4 Bores and channels dimensions’ limits for AM technologies Binder jetting N/A

MJF

SLS

SLM

Powder bed fusion EBM N/A

N/A

LENS

N/A

EBAM

Direct energy deposition

N/A

LOM

Sheet

2.2 Design Aspects 19

20

2 Design for AM

Fig. 2.8 Alternative self-supporting bores‘/channels’ cross-sections

Fig. 2.9 Temporary support structures added to make an overhang manufacturable

Fig. 2.10 Support reduction via a orientation modification, and b substitution with permanent walls

the fundamental tool path width must be used for the design. When the geometry’s width that must be manufactured is not an integer multiple of the fundamental tool path width, the slicer software will have to compensate for that issue. One of the most common ways to tackle this issue is to skip or overlap a certain line of the sliced surface. Another approach is to try and alter the tool path width. These solutions are, for the most part, insufficient as the integrity of the part is compromised or a generic parameter of the machine deviates from the optimum. This could cause a deviation in the dimensional accuracy and the mechanical properties.

Min. wall thickness

(mm)

2

1.6

1.0

0.7

0.6

0.5

0.4

0.3

0.25

0.2

0.1

SLA

DLP

CDLP

Vat polymerization FDM

Extrusion

Table 2.5 Minimum wall thickness for AM technologies

MJ N/A

NPJ N/A

DOD

Material jetting

Binder jetting MJF

SLS

SLM

Powder bed fusion EBM N/A

N/A

LENS

N/A

EBAM

Direct energy deposition

N/A

LOM

Sheet

2.2 Design Aspects 21

22

2 Design for AM

There are physical constraints when building thin walls as they are difficult to form and can be easily distorted [28]. The successful manufacturing of a thin wall is not always a hardware concern. The slicing software determines the G-Code for the AM machine’s fundamental building unit to follow, which can bottleneck the build [29]. A thin feature can be overlooked by the slicing software, although the machine can manufacture that feature in certain optimum scenarios [26]. Smallest Features Apart from the minimum wall thickness, which is considered a 1D thin feature, there are more small features that challenge the ability of the AM machine when comes down to manufacturability. A 2D thin feature that an AM machine can manufacture is usually referred to as the diameter of the smallest possible pin [26]. It could also refer to the side of a rectangular or complex curved geometry. This aspect should be considered during the design phase, as it defines the detail that can be introduced to the part. The smallest features typically attainable by existing AM technologies are summarized in Table 2.6.

2.2.2 Build Parameters Build parameters are selected at the slicing phase2 of the AM process. Besides commonly recognized process parameters (temperature, laser power, feed rate, etc.), some parameters are closely linked to geometric aspects. These parameters are also highly interconnected with the AM technology and the individual machine to be used. Layer Thickness Layer thickness is a factor that affects both the quality of the print and the build time needed to complete the part. With smaller layer thickness, a more detailed part is produced, and the “staircase effect” is minimized. Additionally, with smaller layer thickness, potential voids and gaps are eliminated, as the CAD file is being sliced with more precision and the geometry accuracy is maintained. On the counterpart, with thicker layers, the printing time is reduced. Regarding the staircase effect, another factor that is causing it is the slope angle. As the angle increases, the cosine is proportionally increasing the stair size [30]. Indicative layer thicknesses attainable by AM technologies are summarized in Table 2.7. A proposed solution to this matter is adaptive slicing (Fig. 2.11). The areas where detail is needed are sliced using a thin-layer height, whereas areas whose quality is not affected are sliced with a thicker layer height to contribute to an effective build regarding time and energy consumption [31].

2

This is where the CAD geometry will be translated to G-Code (or other manufacturer-specific machine code) for the AM machine to manufacture the part.

Smallest feature

2

1.6

0.7

0.6

0.5

0.3

0.25

0.2

0.1

(mm)

SLA

DLP

CDLP

Vat polymerization FDM

Extrusion MJ

NPJ N/A

DOD

Material jetting

Table 2.6 Smallest geometrical feature for AM technologies Binder jetting MJF

SLS N/A

SLM

Powder bed fusion EBM N/A

N/A

LENS

N/A

EBAM

Direct energy deposition

LOM

Sheet

2.2 Design Aspects 23

Layer thickness

µSLA

1

1000

400

380

150

130

120

100

70

50

32

25

20

16

8

6

3

SLA

(µm)

DLP

CDLP

Vat polymerization FDM

Extrusion

Table 2.7 Layer thickness for AM technologies

MJ

NPJ

DOD

Material jetting

Binder jetting MJF

SLS

SLM

Powder bed fusion EBM N/A

LENS

EBAM

Direct energy deposition LOM

Sheet

24 2 Design for AM

2.2 Design Aspects

25

Fig. 2.11 Adaptive slicing principle

Build Orientation As already mentioned, build orientation is one of the most crucial build parameters. The orientation of the part relative to the build direction determines which geometrical features are overhanging geometries and what their angle is. Subsequently, the build orientation determines the volume of support structures needed to successfully manufacture the part [10]. This in turn affects the surface quality of the part, as it determines which features will display a more pronounced stair-stepping effect, as well as which down-facing features will have a reduced surface quality. Moreover, it determines the axis on which the mechanical properties show anisotropic behavior, as typical for most AM processes. In metal PBF processes, orientation affects shrinkage (typically non-uniform along all axes) and residual stresses, which may lead to warping of the part. Orientation also affects the cross-sectional area of the part along the build direction, leading to two main effects depending on the AM technology used (Table 2.8). The first one is related to platform adhesion. The part must be restrained at the build plate; thus, the adhesion between the part’s base surface and the machine’s build plate is to be considered. Apart from securing the part, the part-build surface interface facilitates heat dissipation [32]. The second effect is related to the stresses that are developed at the rest of the part’s volume while its layers are manufactured. For the AM technologies that develop significant residual stresses due to the process mechanism (as analyzed in Chap. 3), it is desirable to maintain a small cross-section area to minimize heat accumulation which will lead to residual stresses and thus deformation/warping of the

DLP

CDLP

FDM

MJ

NPJ

Extrusion Material jetting DOD

SLS

SLM

EBM

LENS

Sheet

Adhesion

EBAM LOM

Direct energy deposition

Adhesion Adhesion Adhesion Adhesion Heat and heat and heat and heat and heat convection convection convection convection convection

MJF

Binder Powder bed fusion jetting

Adhesion Adhesion Adhesion Adhesion Adhesion Adhesion Adhesion N/A

SLA

Base Vat polymerization cross-section

Table 2.8 Base cross-section primary function per AM technology

26 2 Design for AM

2.3 Design Considerations

27

part (Table 2.9). This effect has a greater impact on more delicate and elongated structures such as CMF implants [33]. A thermal simulation for the heat concentration provides a picture for the design engineer regarding residual stresses [33]. Finally, related to productivity metrics, orientation can affect build time, and the volume of unused material that needs to be treated and re-used in subsequent builds (in VP, PBF, and BJT processes), as well as the ability to nest multiple parts in the same build job. As such, the optimization of the build orientation is a non-trivial challenge. For all the aforementioned reasons, orientation thinking is important when designing the part, as it enables the designed to correctly identify the areas of their design that could be problematic and subsequently adjust their approach to solve the potential issues.

2.2.3 Process Capability Determination As established by now, each AM process family has its own limitations that are mainly attributed to the process mechanism deployed. As such, typical values for each process family have been collected and presented herein. Nevertheless, exact limits are depending on multiple factors, such as the material and machine used, and even the process parameters selected by the user. Therefore, an appropriate test protocol should be deployed in order to experimentally obtain these values for the actual process/material combination to be used. Such values should be obtained for all applicable design aspects. To this end, different test artifacts have been proposed over the years [32, 34–43], some of which can be seen in Fig. 2.12. Despite the plethora of proposed test artifacts, most of them are tailored to a specific process family. As such, the designer should select the test artifact(s) that are conceived with their intended process family in mind and include the features that are important for the part they are designing. In addition, they should limit the size of the features to be tested to values that are of interest to be used on their part.

2.3 Design Considerations The term “design consideration” can be used to describe the results of design aspects and the process itself on the finished product, that includes geometric characteristics and mechanical properties of the part, as well as KPIs of the AM process. Presented below are the most important design considerations.

Develops residual stresses at cross-section

CDLP Moderate

FDM

DLP

SLA

No

Extrusion

Vat polymerization MJ

NPJ

DOD

Material jetting

Binder jetting

Table 2.9 AM technologies that require a small cross-section area throughout the build

Yes

MJF Yes

SLS

Significant

SLM

Powder bed fusion

Yes

EBM

Yes

LENS

Yes

EBAM

Direct energy deposition

Yes

LOM

Sheet

28 2 Design for AM

2.3 Design Considerations

29

Fig. 2.12 Test artifacts aimed toward characterization of the design limits of AM processes [44]

2.3.1 Accuracy (XY-Plane Versus Z-axis) An important design consideration is to distinguish between the machine’s accuracy on the XY-plane and the Z-axis. Most AM machines have a much higher accuracy on the XY-plane. The accuracy of the machine that will produce the desired part is crucial for the designer in the design phase. For pre-assembled mechanisms or assemblies in general, the dimensional accuracy with which the machine can manufacture has to be considered for the build to be a success. Accuracy and repeatability will dictate the required clearance and tolerances required for multiple parts to be able to be assembled. This is particularly critical when designing “build in place” joints and/ or mechanisms, where the clearance between parts that should experience relative movement should be enough to that the parts are not permanently fused together but at the same time small enough so that it does not introduce additional unwanted movement. As accuracy is different in different directions, such considerations should also be taken into account when selecting the build orientation.

2.3.2 Anisotropic Mechanical Properties AM technologies produce parts with anisotropic mechanical properties. The anisotropic behavior is rooted in the physical process mechanism of all AM processes and can be traced back to the layer-wise material deposition, which results in worse properties along the build axis for most AM process families. The physical mechanism that creates these discrepancies can be different depending on the process family and material used (for example, grain growth in metal PBF, layer adhesion in ME, etc.). Anisotropy can be generally mitigated to some extent via post-processing via heat treatment (as presented in Chap. 3). However, this adds an extra production

30

2 Design for AM

step, and it may not be feasible for components that cannot fit into a furnace, thus needs to be pointed out as a design consideration for AM. Having said that, there are two approaches to designing a part with a given load case. The first one is to orient the designed part in such a way that the loads are received in the direction the AM technology has the greatest mechanical strength. The other, more sophisticated approach is to shape and optimize the part with mechanical strength anisotropy in mind [45].

2.3.3 Shrinkage and Warping These design considerations apply mainly to AM processes utilizing a thermal-based process mechanism for the fusion of the layers. Due to the elevated temperature and the temperature gradients occurring during the AM process, some shrinkage and warping may occur. Such parts are typically cooled down gradually in an attempt to limit the impact of warping and shrinkage. Shrinkage can be mitigated by experimentally verifying the shrinkage percentage of the machine and material to be used and compensating by scaling up the part after the design phase. Shrinkage may be non-uniform across different directions, and modern slicing software allows for non-uniform scaling to accommodate this; as such, it is not a significant design consideration. What should be kept in mind though in terms of design is that parts with thicker geometries and flat and long surfaces, as well as parts with uneven wall thicknesses, will be prone to significant warping due to variable thermal shrinkage and stress along the cross-section of the part. It is thus recommended to maintain even wall thickness and consider even cross-sectional areas when orienting the part to minimize warping. In addition, large and flat surfaces that are not stiff should be avoided, and ribs can be added to them as needed to increase stiffness, mitigating warping.

2.3.4 Surface Roughness The roughness of the completed part is important, as it determines the post-processing steps needed in order to achieve the desired surface quality. The resulting surface roughness is not uniform throughout the entire surface of the printed part. This is caused by the geometry’s slope angle and the unintentional sintering under angled overhangs.

2.4 Additive Manufacturability

31

2.3.5 Build Time The build time refers to the total time required for an AM machine to manufacture the part. The build time and build orientation of the part are highly related. That is due to the fact that material deposition speeds on the XY-plane and Z-axis are not the same. The build unit (e.g., nozzle, laser) moves, thus building the part, with greater speed on the X- and Y-axis than the speed that the layers are adding up [46]. Changing the build orientation will affect the time needed for the AM machine to complete the part. Horizontally orientated parts will in general be printed faster than vertically orientated ones.

2.4 Additive Manufacturability Manufacturability is defined as the part’s ease to be manufactured and its design capacity for cost reduction [2, 47–49]. As such, we can define Additive Manufacturability as the part’s manufacturability when it is to be produced with an AM technology and the degree to which its design utilizes the advantageous aspects of the AM technology and abides by the buildability restrictions. Manufacturability of an AM part is not a duality of can-or-cannot be manufactured. The design aspects and features, that are manufacturable, vary across different AM technologies [50], due to their different build mechanisms, yet the same objectives of reducing manufacturing costs and optimizing the overall process remain. That is, the design efforts of alternations and adaptations for the specific manufacturing technology can be screened from the AM Manufacturability indicator (Fig. 2.13). The determination of Additive Manufacturability at the part design level starts from the identification of the part’s geometric features and their comparison with the AM technology’s capabilities. This cross-checks results in several features that are by definition non-feasible for manufacturing or require support structures (e.g.,

Fig. 2.13 Part manufacturability for AM [51]

32

2 Design for AM

Table 2.10 Framework of the AM Manufacturability determination # Step

Indicator

Part’s geometric features recognition

[from CAD/.stl file]

Cross-checking design features with AM capabilities

[Limit of overhangs, etc.]

Cross-checking design considerations with part specifications

[Surface roughness, porosity]

Magnitude of feature alternation to achieve manufacturable features

[Feasible-impractical-add supports]

Determine the actions for excess material removal

[Few-numerous supports]

Post-processing to reach part’s requirements (post-treatment)

[Minimal-major]

0

1

0

1

2

1

0

1

2

2

1

0

1

2

2

1

0

1

2

out-of-limit overhanging geometries). The magnitude of the design modifications required to correct the non-feasible to manufacturable features is the first leg of the part’s AM Manufacturability. This subset also includes the extent of the required support structures to secure the build while being manufactured [50, 52]. The second leg refers to the post-processing load that the part must go through to reach the desired state and meet the user’s requirements. The post-processing workload can be reduced with properly manufacturable designed features (e.g., overhanging geometries orientated within limits). The framework to determine a part’s Additive Manufacturability is shown in Table 2.10. The overall AM Manufacturability indicator is not one-dimensional as it describes the design’s easiness to be additively manufactured without alternations, the addition of support structures, and the extent of the post-processing load that the part requires to meet its final requirements. This affects the production performance and can be resolved with the implementation of DfAM rules to the part design (Fig. 2.14).

2.5 Re-design for AM When it comes to designing a part for Additive Manufacturing, the most straightforward approach involves re-designing or adapting a pre-existing part in the specifics of the AM process to be used. This approach utilizes conventional parametric design; the design engineer decides upon the most important features, generating the initial shape of the design, and then further determines the exact dimensions for the part to perform for the required specifications [16, 48, 53]. In this case, the designer is constrained by the pre-existing design of the part. Optimization algorithms can be applied but with

2.5 Re-design for AM

33

Fig. 2.14 DfAM rules’ affection on production and post-processing workload [51]

limited design freedom, and as such, the resulting design is often not optimal. In addition, the designer should manually take into account process-related aspects, as well as generic and process-specific DfAM guidelines. This approach requires a synthesis of design rules and guidelines, to either assist expert designers or be embedded in software tools executing design optimization routines. Such approaches have been structured and presented in the past [54]; most common geometrical features were identified and guidelines for designing each one as well as explanations and good practices were established and presented in a “design table” form. In [55], the implications involved in the introduction of AM to production have been discussed and compared with conventional manufacturing methods, making clear that there are advantages when using the appropriate design approaches to AM. One of the most promising value propositions of Additive Manufacturing technologies is the zero-cost geometrical complexity, assuming part manufacturability. Benefits such as design freedom, integrated design, production flexibility with no need for product-specific tools, and lead time reduction for short-series production are assets that AM can offer to the industrial sector (4.1), [56]. However, design for AM is similar in a way to design for composites; one must comprehend the unique challenges and opportunities that are linked to the technology to be used in order to take advantage of it. This can be achieved with the backpropagation of the process and manufacturing information, which morphs the shape and characteristics of the design, resulting in a part that abides by the DfAM rules from the first design iteration (Fig. 2.15). This allows not only more manufacturable designs to be generated, but also the design process is made in a more linear fashion, eliminating time-consuming design iterations. Re-designing a part for AM can be distinguished into three phases; (i) selection of potential parts and design simplification/consolidation, (ii) design optimization, and (iii) design validation. Typically, the selection of potential parts to be substituted by new designs made by AM is a decision based on empirical observations or previous knowledge and experience on the specific application, in conjunction with

34

2 Design for AM

Fig. 2.15 Conventional (orange) and proposed (blue) design process for AM [8]

AM process knowledge to ensure that AM adds value in that particular case (for example, identifying heavy parts with a high potential of weight reduction). Thus, in the first phase, the user can be in a position to investigate the potential for the replacement of a component and choose the most appropriate one. This decision is also based on specific requirements that have to be addressed and assure that the particular part can be produced via AM. The material used for manufacturing these parts should also be examined, to determine availability and mechanical or other properties. A systematized approach to check the validity of the aforementioned selection is presented in Sect. 3.1. Additionally, the design limitations due to the manufacturing process utilized are obtained. Material removal areas and possible filets or radii should be identified at this stage. Furthermore, connecting plates or brackets manufactured for joining the components with the rest of the assembly can be integrated into a singular design, thus saving weight and assembly time from the structure (part consolidation). Due to the additional geometric flexibility granted by transitioning to AM, the design of the part may be further simplified, removing geometrical features that have been introduced during previous versions of the part, designed to be manufactured via other processes. The key consideration at this stage is disconnecting the part function from its geometry; this allows the designer to retain only the design elements important for the functionality of the part and remove any other constraints which can lead to an improved design. However, in the case of direct part replacement, the constraints imposed from the operating environment of the investigated component must be respected. Mounting points and dedicated areas for the installation of other mechanical parts (such as bolts, bearings) should remain the same during the redesign phase, due to the fact that the choice of these items is based on interdependent criteria, and their rearrangement in the assembly may cause implications with their

2.5 Re-design for AM

35

Fig. 2.16 Re-design of an existing part for AM. Adapted from [54]

interface in the rest of the final product. A characteristic example can be seen in Fig. 2.16. During the second phase, a detailed study of the loads (application points, load paths, force magnitude) developed in the component from the operating conditions should be conducted. The designer can utilize this information along with finite element analysis (FEA) algorithms to identify inefficiencies in the design (for example, areas where material can be removed without adverse effects on the stiffness or strength of the part). For this, a knowledge of the attainable mechanical material properties should also exist, to determine stress thresholds. At this stage, optimization algorithms such as Topology Optimization (TO) can also be deployed, albeit in a limited way. Such optimization algorithms are presented in more detail in Sect. 2.4. The Topology Optimization algorithm can start removing material from the enabled design areas, by defining the areas of interest, i.e., areas for material removal, and frozen areas (bearings housing, mounting points, etc.). Normally, the optimization task is the generation of a structure maintaining the component’s stiffness threshold, while enabling a volume, stress, or weight restriction. Thus, the optimization parameter is the part compliance (deformation) which is represented as the sum of the model’s strain energy. This value has to meet a certain value during the optimization procedure. The result of phase 2 will be a new design, suitable for use in the existing operational conditions of the assembly (as imposed by the rest of the components), which will be lighter while at the same time maintaining the level of stiffness that is required. These iterations should also be aligned with pre-established performance targets;

36

2 Design for AM

weight, structural performance (displacement, max. stresses), and total volume are crucial measures of part performance. In the third and final phases of the conceptual design framework, the new design should be verified in terms of the performance targets and its operability inside the assembly. As soon as it is obtained that the new component can be effectively incorporated into the assembly, without sacrificing performance and mainly avoiding the re-design of other parts to be fitted in it, the user can allow the CAD file to be forwarded to a CAM tool. The CAM tool will subsequently slice the part, generate support structures, and generate the part program in order to build up the part. Based on the outcomes of the CAM tool, the designer can determine potential inefficiencies (such as excessive supports) and could use these findings to revise the design, improving manufacturability.

2.6 Design Optimization The second approach of design for AM involves creating an optimized design from scratch, utilizing algorithmic design methods; the designer decides upon the morphing equation, its parameters, and the specific space inside the build volume that the algorithm fills with structures of material [57, 58]. The algorithmic methods branch into two main and dissimilar subcategories, namely Topology Optimization (TO) and Generative Design (GD). Both approaches start from the build volume and add or subtract geometrical entities to morph the part. Topology Optimization tools are widely known and used for more than a decade. Starting from a «bulk» of material, the designer applies a series of constraints and boundary conditions and the algorithm optimizes the material distribution based on certain criteria. When designing structural parts, the most common criterion would be the minimization of part volume (and subsequently mass) while maintaining maximum stresses and their distribution at a certain, user-defined level. This iterative process greatly resembles exhaustive search algorithms, and as such, it is computationally intensive, requiring lots of time and resources to be executed. Applications of Topology Optimization (TO) to Additive Manufacturing have been discussed in [59], concluding that the method is mostly used to generate rough shapes that need to be refined before the actual parts are manufactured. Topology Optimization has been used as a method of decreasing the mass of a part while minimizing strain energy [60]. The promise for Topology Optimization in design for Additive Manufacturing, mainly focusing on functionally graded porosity, through the usage of lattice structures, has also been discussed in [61]. In [62] unconventional approaches to design are presented, enabled by the use of AM, focusing on bio-inspired cellular structures. In [63], a method of producing a part with AM via clearly established functions and constraints and the application of iterative Topology Optimization and re-design routine has been presented. However, validation of the parts’ design has been reported to be impossible if they have not been actually manufactured. A novel methodology applied to parts manufactured by laser-based AM processes was

2.6 Design Optimization

37

proposed in [64]. The authors identified global and local sets of rules in order to optimize AM production. However, the methodology is limited to simple “extrudedlike” 2.5D structures and not to complex 3D shapes. The use of lattice structures and their optimization have been investigated in [65] as an example of lightweight design. The paper shows that the use of the lattice structure should not be an objective by itself, yet the combination of the Topology Optimization and the lattice structure proves to be the optimal choice for the part under consideration and its purpose. Generative Design (GD) refers to an algorithmic design optimization method that creates a certain number of designs based on certain non-geometric requirements or constraints on product performance. The user then selects and fine-tunes the resulting designs that fit their constraints [66, 67]. Generative Design has been applied across different design problems in heterogeneous domains [68, 69]. He et al. [70] have presented a generative-based approach called truss layout optimization to create optimized additively manufactured components, including certain process-related build constraints. A method for Generative Design tailored in AM and inspired by termite nest building has been presented by Dokhia et al. [71]. Salta et al. [72] have discussed the viability of additively manufacturing an emergency shelter that was automatically adapted in terms of design using Generative Design algorithms. Generative Design approaches have also been used to design optimized support structures for AM [73]. Such tools have also been integrated into commercially available CAD software packages. Junk et al. [74] have collected and compared the market offerings in terms of such tools that can also integrate AM workflows. Generative Design approaches need the design engineer’s input at the end of the design stage to determine which features and which design variant are to be manufactured. Therefore, additional AM buildability knowledge is required to choose the optimum part variant at the end of the design process [68]. This feature and part variant selection are the first point of the design process where manufacturability concerns begin to appear, as certain geometries and features can make the product’s manufacturing unviable with AM. One very important limitation of both Topology Optimization and Generative Design methods is that the very specific optimization goals generate highly complex parts, making them impossible to be realized with conventional manufacturing. Albeit the fact that the resulting, highly complex parts are appealing for Additive Manufacturing, they do not take into account the specificities and limitations of the manufacturing process to be used. Design modifications are then required in a second stage to address problematic aspects of the design and increase the part’s manufacturability [75, 76]. This additional design modification stage highlights the need for a design method, where manufacturing specifications and the component’s functionality simultaneously act to shape the design and optimize the AM process to make it economically viable [77, 78]. An interesting addition to these types of tools would be a method to optimize the design for AM production, rather than function.

38

2 Design for AM

Such a method has been proposed in [79], focusing on achieving shape optimization, and has Additive Manufacturability criteria checked in an iterative fashion. The resulting geometry is not the most optimized in terms of weight reduction; however, the method claims to ensure manufacturability of the design with a small weight penalty over established TO approaches.

2.7 Design for Post-processing Despite the inherent benefits of AM processes, most of them face a common challenge; parts need to be post-processed before being able to be used. Post-processing includes any processing steps that follow the Additive Manufacturing of a part; this includes cleaning, mechanical material removal, (electro)thermal, and chemical treatments as well as laser-based post-processing, aimed to improve geometric characteristics, appearance, and mechanical behavior of the parts. Post-processing considerations are presented in detail in Sect. 3.6; nevertheless, certain design decisions can significantly improve the ease and effectiveness of post-processing that will be presented herein. The first design consideration linked to post-processing is related to residual unconsolidated material removal. Cavities and intricate geometries typically result in trapping unconsolidated powder or resin, which needs to be removed after the part is completed. Design features that can aid the removal of residual powder/resin are the introduction of radiused internal edges and maintaining a relatively simple internal profile without sharp direction changes. Hollow cavities should have at least one opening of sufficient size; ideally, multiple openings should be introduced on opposite sides of the cavity. Blind holes/bores should be substituted by through-holes whenever possible. For metal AM processes, thermal treatment can be deployed to alleviate anisotropic mechanical properties and reduce porosity. The designer should therefore ensure that the part will not be warped and distorted during the heat treatment process. This includes avoiding long, flat sections by the addition of strengthening ribs. These ribs should be added on the outer surface whenever possible, to aid residual powder removal as already anticipated. In addition, the designer should ensure that the build plate adhesion and support structures’ interface are strong enough to withstand thermal stresses. Support removal is also a challenge that can be solved through design. Non-soluble supports need to be mechanically removed. For polymer AM processes, due to the relatively low strength of the material, manual methods are deployed. For metal AM processes, power tools are typically required. In all cases, ensuring tool accessibility is paramount. If tool accessibility cannot be guaranteed, replacing temporary supports with permanent walls can be considered (see Sect. 2.2.1). AM processes cannot achieve very tight tolerances, flatness, and surface roughness. As such, mechanical material removal processes such as milling and/or grinding are deployed during post-processing. The designer should ensure that the AM parts

2.8 AM Relevant Standards

39

can facilitate those additional process steps. The parts should have enough additional material that can be mechanically removed without compromising the overall part dimensions. The part design should ensure sufficient strength and stiffness so that the part can withstand the mechanical loads imposed on them during post-processing. In addition, fixturing considerations should be kept in mind. Design elements can be added to enable the part to be easily clamped in a milling machine. These elements should also have sufficient strength and stiffness to ensure an effective milling operation, but should also be relatively easy to remove from the part once machining is completed. Additional referencing elements can be added, so that part setup in the milling machine can be facilitated. Such clamping and referencing elements are common practice in the metal casting industry, since cast components are also typically post-machined; therefore, similar approaches can also be followed here. These additional elements can be subsequently removed at the last step of post-machining, to further reduce component weight if not further required for functional purposes.

2.8 AM Relevant Standards Toward maturing AM technologies and making them more relevant for industrial uptakes, numerous standardization committees are working on establishing standards for AM. While most of these standards deal with certain procedures such as material handling, material characterization, heat treatment, there are a few that are linked to design and would be useful for designers to consult. These are: • ISO/ASTM 52900:2021 Additive Manufacturing—General principles—Fundamentals and vocabulary. • ISO/ASTM 52902:2019 Additive Manufacturing—Test artifacts—Geometric capability assessment of Additive Manufacturing systems. • ISO/ASTM 52910:2018 Additive Manufacturing—Design—Requirements, guidelines, and recommendations. • ISO/ASTM 52911-1:2019 Additive Manufacturing—Design—Part 1: Laserbased Powder Bed Fusion of metals. • ISO/ASTM 52911-2:2019 Additive Manufacturing—Design—Part 2: Laserbased Powder Bed Fusion of polymers. • ISO/ASTM 52912:2020 Additive Manufacturing—Design—Functionally graded Additive Manufacturing. • ISO/ASTM 52909:2022 Additive Manufacturing of metals—Finished part properties—Orientation and location dependence of mechanical properties for metal Powder Bed Fusion. • VDI 3405 Part 3 Additive Manufacturing processes, rapid manufacturing— Design rules for part production using laser sintering and laser beam melting. • VDI 3405 Part 3.2 (Draft) Additive Manufacturing processes—Design rules— Test artifacts and test features for limiting geometric elements.

40

2 Design for AM

• VDI 3405 Part 3.4 Additive Manufacturing processes—Design rules for part production using Material Extrusion processes. • VDI 3405 Part 3.5 Additive Manufacturing processes, rapid manufacturing— Design rules for part production using electron beam melting. • VDI 3405 Part 8.1 Additive Manufacturing processes—Design rules—Parts using ceramic materials.

References 1. H. Bikas, P. Stavropoulos, G. Chryssolouris, Additive manufacturing methods and modelling approaches: a critical review. Int. J. Adv. Manuf. Technol. 83(1), 389–405 (2016). https://doi. org/10.1007/s00170-015-7576-2 2. G. Chryssolouris, Manufacturing Systems: Theory and Practice, 2nd edn. (Springer-Verlag, New York, 2006). ISBN 978-0-387-28431-6 3. P. Patrick, Z. Zicheng, B. Richard, M. James, A framework for mapping design for additive manufacturing knowledge for industrial and product design. J. Eng. Des. 29(6), 291–326 (2018). https://doi.org/10.1080/09544828.2018.1483011 4. T. Mangera, F. Kienhöfer, K. Carlson, M. Conning, T. Purkis, G. Govender, Design for manufacture and assembly of a polycentric paediatric prosthetic knee. Proc. Instit. Mech. Eng. Part B: J. Eng. Manuf. (2019). https://doi.org/10.1177/0954405419840555 5. G. Colombo, F. Mandorli, Evolution in mechanical design automation and engineering knowledge management. Innov. Prod. Des. 55–78 (2011). https://doi.org/10.1007/978-0-85729775-4_4 6. W. Rosen, Computer-aided design for additive manufacturing of cellular structures. Comput. Aided Des. Appl. 4, 585–594 (2007). https://doi.org/10.1080/16864360.2007.10738493 7. H. Bikas, J. Stavridis, P. Stavropoulos, G. Chryssolouris, A design framework to replace conventional manufacturing processes with additive manufacturing for structural components: a formula student case study. Procedia CIRP 57, 710–715 (2016). https://doi.org/10.1016/j.pro cir.2016.11.123 8. H. Bikas, A.K. Lianos, P. Stavropoulos, A design framework for additive manufacturing. Int. J. Adv. Manuf. Technol. 103, 3769–3783 (2019). https://doi.org/10.1007/s00170-019-03627-z 9. G. Spiegel, W.B. Wright, IV. Slicing (2016), https://medium.com/3d-printing-in-o-p/iv-slicing72a9515f44bc. Accessed 04 Nov 2022 10. A.R. Nassar, E.W. Reutzel, Beyond layer-by-layer additive manufacturing – voxel-wise directed energy deposition, in Proceedings of the 26th Annual International Solid Freeform Fabrication Symposium - An Additive Manufacturing Conference. SFF 2015 (2015), pp. 273–283 11. J. Jiang, J. Stringer, X. Xu, R.Y. Zhong, Investigation of printable threshold overhang angle in extrusion-based additive manufacturing for reducing support waste. Int. J. Comput. Integr. Manuf. 31(10), 961–969 (2018). https://doi.org/10.1080/0951192x.2018.1466398 12. G. Allaire, C. Dapogny, R. Estevez, A. Faure, G. Michailidis, Structural optimization under overhang constraints imposed by additive manufacturing technologies. J. Comput. Phys. 351, 295–328 (2017). https://doi.org/10.1016/j.jcp.2017.09.041 13. B. Redwood, F. Schöffer, B. Garret, The 3D printing handbook: technologies, design and applications (2017) 14. J. Jiang, X. Xu, J. Stringer, Support structures for additive manufacturing: a review, journal of manufacturing and materials processing. J. Manuf. Mater. Process 2(4), 64 (2018). https://doi. org/10.3390/jmmp2040064 15. A. Alafaghani, A. Qattawi, M.A. Ablat, Design consideration for additive manufacturing: fused deposition modelling. Open J. Appl. Sci. 7(6), 291–318 (2017). https://doi.org/10.4236/ojapps. 2017.76024

References

41

16. F. Yang, Y. Tang, Y.F. Zhao, Manufacturability of overhang structures fabricated by binder jetting process, in ASME 2016 International Mechanical Engineering Congress and Exposition, Volume 2: Advanced Manufacturing (2016). https://doi.org/10.1115/imece2016-65927; E. Utley, An introduction to designing for metal 3D printing (2017), https://blogs.solidworks.com/ solidworksblog/2017/06/introduction-designing-metal-3d-printing.html. Accessed 17 Nov 2022 17. P. Chain, Supports in 3D printing: a technology overview [online]. Available at: https://www.3dhubs.com/knowledge-base/supports-3d-printing-technology-overview#bin der-jetting. Accessed 19 Nov 2022 18. M. Cloots, L. Zumofen, A.B. Spierings, A. Kirchheim, K. Wegener, Approaches to minimize overhang angles of SLM parts. Rapid Prototyp. J. 23(2), 362–369 (2017). https://doi.org/10. 1108/RPJ-05-2015-0061 19. P.J.S. Bartolo, A.C.S. de Lemos, A.M.H. Pereire, A.J.D.S. Mateus, C. Ramos, C. Dos Santos, D. Oliveira, E. Pinto, F. Craveiro, H.M. Bartolo, H. Almeida, I. Sousa, High value manufacturing: advanced research in virtual and rapid prototyping, in 6th International Conference on Advanced Research in Virtual and Rapid Prototyping (Leiria, Portugal, 2013) 20. D. Thomas, The development of design rules for selective laser melting. Dissertation, University of Wales, 2009 21. M. Lemay, Design for additive manufacturing: (1) internal channels (2018), https://www. linkedin.com/pulse/design-additive-manufacturing-1-internal-channels-matt-lemay. Accessed 22 Nov 2022 22. M. Pietropaoli, R. Ahlfeld, F. Montomoli, A. Ciani, M. D’Ercole, Design for additive manufacturing: internal channel optimization. J. Eng. Gas Turbines Power 139(10), 102101 (2017). https://doi.org/10.1115/1.4036358 23. C. Beyer, D. Figueroa, Design and analysis of lattice structures for additive manufacturing. J. Manuf. Sci. Eng. 138(12), 121014 (2016). https://doi.org/10.1115/1.4033957 24. B. Valmik, K. Prakash, P. Vinaykumar, K. Shreyans, G. Kiran, S. Rajkumar, A review on powder bed fusion technology of metal additive manufacturing, in 4th International Conference and Exhibition on Additive Manufacturing Technologies-AM-2014 (Banglore, India, 2014) 25. Renishaw article, Additive manufacturing and the need for supports (2019), https://www.ren ishaw.com/en/can-you-build-am-parts-without-supports--42173. Accessed 17 Nov 2022 26. L. Yang, H. Gong, S. Dilip, B. Stucker, An investigation of thin feature generation in direct metal laser sintering systems, in 25th Annual International Solid Freeform Fabrication Symposium (2014) 27. C.C. Seepersad, T. Govett, K. Kim, M. Lundin, D. Pinero, A designer’s guide for dimensioning and tolerancing SLS parts, in Annual International Solid Freeform Fabrication Symposium An Additive Manufacturing Conference (2012) 28. S. Kubiak, Guidelines for wall thickness in laser sintering (2017), https://www.additivemanu facturing.media/blog/post/guidelines-for-wall-thickness-in-laser-sintering. Accessed 21 Nov 2022 29. A. Papacharalampopoulos, P. Stavropoulos, H. Bikas, Path planning for filling 3D printed parts utilizing Hilbert curves. Procedia Manuf. 21, 757–764 (2017). https://doi.org/10.1016/j. promfg.2018.02.181 30. J. Kechagias, S. Zannis, G. Chryssolouris, Surface roughness modelling of the Helisys laminated object manufacturing (LOM) process, in 8th European Conference on Rapid Prototyping and Manufacturing (Nottingham, U.K, 1999), pp. 141–152 31. S. Sikder, A. Barari, H.A. Kishawy, Effect of adaptive slicing on surface integrity in additive manufacturing, in Volume 1A: 34th Computers and Information in Engineering Conference (2014). https://doi.org/10.1115/detc2014-35559 32. T. Mukherjee, H.L. Wei, A. De, T. DebRoy, Heat and fluid flow in additive manufacturing – part II: powder bed fusion of stainless steel, and titanium, nickel and aluminum base alloys. Comput. Mater. Sci. 150, 369–380 (2018). https://doi.org/10.1016/j.commatsci.2018.04.027 33. AM simulation: a guide to the best part orientation and support configuration, https://www. materialise.com/en/cases/am-simulation-best-part-orientation-support-configuration-cmf-imp lants. Accessed 27 Aug 2022

42

2 Design for AM

34. J.P. Kruth, B. Vandenbroucke, J. Van Vaerenbergh, P. Mercelis, Benchmarking of different SLS/SLM processes as rapid manufacturing techniques. Int. Conf. Polymers Moulds Innov (2005) 35. E.C. Santos, M. Shiomi, K. Oskada, T. Laoui, Rapid manufacturing of metal components by laser forming. Int. J. Mach. Tools Manuf. 46(12–13), 1459–1468 (2006). https://doi.org/10. 1016/j.ijmachtools.2005.09.005 36. B. Cavallini, J. Ciurana, C. Reguant, J. Delgado, Studying the repeatability in DMLS technology using a complete geometry test part, in Innovative Developments in Design and Manufacturing: Advanced Research in Virtual and Rapid Prototyping – Proceedings of VR@P4. (2009), pp. 349–354 37. W.M. Johnson, M. Rowell, B. Deason, M. Eubanks, Benchmarking evaluation of an open source fused deposition modeling additive manufacturing system, in Proceedings of the 22nd Solid Freeform Fabrication Symposium (Austin, TX, USA, 2011) 38. G.D. Kim, Y.T. Oh, A benchmark study on rapid prototyping processes and machines: quantitative comparisons of mechanical properties, accuracy, roughness, speed, and material cost. Proc. Instit. Mech. Eng. Part B: J. Eng. Manuf. 222(2), 201–215 (2008). https://doi.org/10. 1243/09544054JEM724 39. M. Fahad, N. Hopkinson, A new benchmarking part for evaluating the accuracy and repeatability of additive manufacturing (AM) processes, in Proceedings of 2nd International Conference on Mechanical, Production and Automobile Engineering (Singapore, 2012) 40. D. Scaravett, P. Dubois, R. Duchamp, Qualification of rapid prototyping tools: proposition of a procedure and a test part. Int. J. Adv. Manuf. Technol. 38, 683–690 (2008). https://doi.org/ 10.1007/s00170-007-1129-2 41. F. Xu, Y.S. Wong, H.T. Loh, Toward generic models for comparative evaluation and process selection in rapid prototyping and manufacturing. J. Manuf. Syst. 19(5), 283–296 (2000). https:/ /doi.org/10.1016/S0278-6125(01)89001-4 42. G. Papazetis, G.C. Vosniakos, Feature-based process parameter variation in continuous paths to improve dimensional accuracy in three-dimensional printing via material extrusion. Proc. Instit. Mech. Eng. Part B: J. Eng. Manuf. 233(12). https://doi.org/10.1177/0954405419838361 43. J. Allison, C. Sharpe, C.C. Seepersad, A test part for evaluating the accuracy and resolution of a polymer powder bed fusion process. ASME J. Mech. Des. 139(10) (2017). https://doi.org/ 10.1115/1.4037303 44. P. Shah, R. Racasan, P. Bills, Comparison of different additive manufacturing methods using computed tomography. Case Stud. Nondestruct. Test. Eval. 6(B), 69–78 (2016). https://doi. org/10.1016/j.csndt.2016.05.008 45. G. Chryssolouris, J. Kechagias, P. Moustakas, P. Koutras, An experimental investigation of the tensile strength of parts produced by laminated object manufacturing (LOM) process. CIRP Journal of Manufacturing Systems, in Proceedings of the 34th CIRP International Seminar on Manufacturing Systems (Athens, Greece, 2003) 46. C. Oswald, Additive manufacturing design considerations for production in aerospace (2019), http://app.emarketeer.com/resources/12517/My_Documents/AddManufacturingDesig nPart1-LAI-FINAL.pdf. Accessed 09 Sep 2022 47. J. Butt, S. Jedi, Redesign of an in-market conveyor system for manufacturing cost reduction and design efficiency using DFMA methodology. Designs 4, 6 (2020). https://doi.org/10.3390/ designs4010006 48. M.A. David, Design for manufacturability; how to use concurrent engineering to rapidly develop low-cost, high-quality products for Lean Production (Productivity Press, 2014), www. design4manufacturability.com/books.htm. Accessed 05 Oct 2022 49. V. Todi´c, D. Luki´c, M. Milosevic, G. Jovicic, J. Vukman, Manufacturability of product design regarding suitability for manufacturing and assembly (DfMA). J. Prod. Eng. 16, 47–50 (2012) 50. J. Willems, M. Megahed, Multi-Objective Optimization of PBF Workpiece Orientation (SimAM, Pavia, Italy, 2019) 51. A.K. Lianos, H. Bikas, P. Stavropoulos, A shape optimization method for part design derived from the buildability restrictions of the directed energy deposition additive manufacturing process. Designs 4(3) (2020). https://doi.org/10.3390/designs4030019

References

43

52. O. Yilmaz, A.A. Ugla, Shaped metal deposition technique in additive manufacturing: a review. Proc. Instit. Mech. Eng. Part B: J. Eng. Manuf. 230(10), 1781–1798 (2016). https://doi.org/10. 1177/0954405416640181 53. D. Roller, An approach to computer-aided parametric design. Comput. Aided Des. 23(5), 385–391 (1991). https://doi.org/10.1016/0010-4485(91)90033-s 54. J. Kranz, D. Herzog, C. Emmelmann, Design guidelines for laser additive manufacturing of lightweight structures in TiAl6V4. J. Laser Appl. 27, S14001 (2015). https://doi.org/10.2351/ 1.4885235 55. S. Mellor, L. Hao, D. Zhang, Additive manufacturing: a framework for implementation. Int. J. Prod. Econ. 149, 194–201 (2014). https://doi.org/10.1016/j.ijpe.2013.07.008 56. A.K. Lianos, H. Bikas, P. Stavropoulos, A shape optimization method for part design derived from the buildability restrictions of the directed energy (2020) 57. S.L. Liu, Q.W. Chen, L. Tong, G. Cheng, An identification method for enclosed voids restriction in manufacturability design for additive manufacturing structures. Front. Mech. Eng. 10(2), 126–137 (2015). https://doi.org/10.1007/s11465-015-0340-3 58. M. Nashvili, M. Olhofer, B. Sendhoff, Morphing methods in evolutionary design optimization, in Proceedings of the 2005 Conference on Genetic and Evolutionary Computation - GECCO ’05 (Washington DC, USA, 2005). https://doi.org/10.1145/1068009.1068159 59. M.S. Edke, K.H. Chang, Shape optimization of heavy load carrying components for structural performance and manufacturing cost. Struct. Multidiscip. Optim. 31, 344–354 (2006). https:// doi.org/10.1007/s00158-005-0603-4 60. E.C. Tsirogiannis, G.C. Vosniakos, Redesign and topology optimisation of an industrial robot link for additive manufacturing. Facta Universitatis Series Mech. Eng. 17(3), 415–424 (2019). https://doi.org/10.22190/FUME181219003T 61. G. Chahine, P. Smith, R. Kovacevic, Application of Topology Optimization in Modern Additive Manufacturing, Solid Freeform Fabrication Symposium Austin, Texas, USA (2010) 62. D.W. Rosen, Design for additive manufacturing: a method to explore unexplored regions of the design space, in Solid Freeform Fabrication Proceedings (2007) 63. B. Vayre, F. Vignat, F. Villeneuve, Designing for AM. Procedia CIRP 3, 632–637 (2012). https:/ /doi.org/10.1016/j.procir.2012.07.108 64. R. Ponchen, O. Kerbrat, P. Mognol, J.Y. Hascoet, A novel methodology of design for AM applied to additive laser manufacturing process. Robot. Comput.-Integrat. Manuf. 30, 389–398 (2014). https://doi.org/10.1016/j.rcim.2013.12.001 65. G. Reinhart, S. Teufelhart, Load-adapted design of generative manufactured lattice structures. Phys. Procedia 12, 385–392 (2011). https://doi.org/10.1016/j.phpro.2011.03.049 66. K. Meintjes, Generative design – what’s that? – CIMdata (2017), https://www.cimdata.com/ de/resources/cimdata-blog/item/8402-generative-design-what-s-that/8402-generative-designwhat-s-that. Accessed 15 Sep 2022 67. P. Keane, Generative design: the road to production (2018), https://www.engineering.com/ story/generative-design-the-road-to-production. Accessed 15 Sep 2022 68. S. Khan, M.J. Awan, A generative design technique for exploring shape variations. Adv. Eng. Inform. 38, 712–724 (2018). https://doi.org/10.1016/j.aei.2018.10.005 69. S. Krish, A practical generative design method. Comput. Aided Des. 43(1), 88–100 (2011). https://doi.org/10.1016/j.cad.2010.09.009 70. L. He, M. Gilbert, T. Johnson, T. Pritchard, Conceptual design of AM components using layout and geometry optimization. Comput. Math. Appl. 78(7), 2308–2324 (2019). https://doi.org/10. 1016/j.camwa.2018.07.012 71. V. Dhokia, W.P. Essink, J.M. Flynn, A generative multi-agent design methodology for additively manufactured parts inspired by termite nest building. CIRP Ann. 66(1), 153–156 (2017). https:/ /doi.org/10.1016/j.cirp.2017.04.039 72. S. Salta, N. Papavasileiou, K. Pyliotis, M. Katsaros, Adaptable emergency shelter: a case study in generative design and additive manufacturing in mass customization era. Procedia Manuf. 44, 124–131 (2020). https://doi.org/10.1016/j.promfg.2020.02.213

44

2 Design for AM

73. Y. Zhang, Z. Wang, Y. Zhang, S. Gomes, A. Bernard, Bio-inspired generative design for support structure generation and optimization in additive manufacturing (AM). CIRP Ann. 69(1), 117– 120 (2020). https://doi.org/10.1016/j.cirp.2020.04.091 74. S. Junk, L. Burkart, Comparison of CAD systems for generative design for use with additive manufacturing. Procedia CIRP 100, 577–582 (2021). https://doi.org/10.1016/j.procir.2021. 05.126 75. M. Leary, L. Merli, F. Torti, M. Mazur, M. Brand, Optimal topology for additive manufacture: a method for enabling additive manufacture of support-free optimal structures. Mater. Des. 63, 678–690 (2014). https://doi.org/10.1016/j.matdes.2014.06.015 76. A. Panesar, M. Abdi, D. Hickman, I. Ashcroft, Strategies for functionally graded lattice structures derived using topology optimisation for additive manufacturing. Addit. Manuf. 19, 81–94 (2018). https://doi.org/10.1016/j.addma.2017.11.008 77. R. Molaei, A. Fatemi, Fatigue design with additive manufactured metals: issues to consider and perspective for future research. Procedia Eng. 213, 5–16 (2018). https://doi.org/10.1016/j. proeng.2018.02.002 78. T. Yunlong, F.Z. Yaoyao, A survey of the design methods for additive manufacturing to improve functional performance. Rapid Prototyp. J. 22(3), 569–590 (2016). https://doi.org/10.1108/ RPJ-01-2015-0011 79. Deposition additive manufacturing process. Designs 4(3), 19. https://doi.org/10.3390/design s4030019

Chapter 3

AM Processes

Initially, the available AM methods will be classified based on their process mechanisms based on the analysis of (a) their characteristics, advantages, and drawbacks and (b) the criteria for the selection of the most suitable AM method according to specific needs. The most important AM part quality issues will be analyzed in detail, along with the reasons that cause them. Process optimization strategies to improve AM product quality will be presented, focusing on the effect of crucial AM processes. Next, simulation for AM will be analyzed, detailing the different simulation types and how they can be practically applied in an industrial environment. Additionally, post-processing methods for AM will be described; moreover, AM-specific monitoring, process control, and quality assessment strategies will be presented. Finally, the connection of the aforementioned modules to Hybrid AM as well as to digital twin applications for AM will be discussed.

3.1 Process and Material Selection Initially, it has to be decided whether AM is a suitable choice for a given part. The most important parameters to be taken into consideration are the lot size and part complexity [1]. However, a cost model and a Return on Investment (ROI) should also be utilized for this selection [2]; this will be further analyzed in Chap. 4, under the business perspective of AM. Therefore, in this chapter, the decisions that have to be made—given that AM is the best manufacturing scenario for a certain product— will be analyzed, namely the selection of the material, the AM process family, and the specific machine that will be used. This is a multi-criterion decision-making problem that requires careful evaluation for its solution [2]. First, a general overview of this procedure will be presented, followed by additional details for each step. The material selection is the first step, in which the product requirements have to be considered. Once the material has been decided, the geometrical requirements have © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 P. Stavropoulos, Additive Manufacturing: Design, Processes and Applications, SpringerBriefs in Applied Sciences and Technology, https://doi.org/10.1007/978-3-031-33793-2_3

45

46

3 AM Processes

to be considered for the selection of the AM process family. The advantages and drawbacks of each AM process family need also to be considered in this selection process; this is qualitatively depicted in Fig. 3.1, along with the severity of common AM part quality issues. Once the AM process family has been selected, the final step is the selection of the specific AM machine. The maximum part dimensions, as well as the building speed and resolution capabilities of the machine, have to be considered; this allows for successful and cost-effective production. A synopsis of the basic steps of the material and process decision-making process can be seen in Fig. 3.2. Table 3.1 lists important considerations that have to be made concerning the material selection. More specifically, the intended use of a product is of critical importance because prototyping and end-use products have different requirements.

Fig. 3.1 Limitations and severity of common AM part quality issues of the different AM process families

Fig. 3.2 AM material-, process-, and machine-selection flowchart

3.1 Process and Material Selection Table 3.1 Material selection considerations [3]

47

1. Determine performance requirements via intent of use Prototyping and R&D? Validation or pre-production? End-use parts? 2. Translate performance to material requirements Stiffness/hardness Toughness/tear strength Temperature and chemical resistance Creep and durability 3. Select material Design space/operation environment Durability Longevity 4. AM machine selection Product requirements Material compatibility

This has a direct impact on the required mechanical properties and, consequently, on the material selection. Additionally, the choice of the AM process needs to fulfill the criteria of material compatibility and product requirements. Proceeding to the second step, Table 3.2 lists more detailed AM process considerations. More specifically, the available range of materials for each AM process should be investigated and the difference in mechanical properties among AM processes should be considered. Moreover, the process capabilities should be considered based on their accuracy, surface roughness, as well as their capability for the manufacturing of overhangs, bridges, and supports. A detailed summary of the capabilities of each process family regarding different design aspects can be seen in Fig. 3.3. Additionally, the build rate of the process families is a crucial criterion that directly affects the product cost and the post-processing requirements; these should be considered while taking into account the design requirements to achieve successful and cost-effective products. Figures 3.4 and 3.5 illustrate rules of thumb for the selection of a suitable AM process for prototyping and functional end-use parts, respectively. Furthermore, the utilization of the Advanced Hierarchy Process (AHP) is advised for a concrete decision-making process because it allows for a decision-making evaluation based on multiple criteria. To execute the method, different Key Performance Indicators (KPIs) must be selected; then, they should be weighted according to the level of importance for the specific case. Next, scores should be calculated for each KPI and the aforementioned weights should be applied [9].

48

3 AM Processes

Table 3.2 AM process selection considerations 1. Investigate material availability Available range of materials for each AM process Differences in mechanical properties across AM processes 2. Determine process capabilities Process accuracy (“resolution”) depends on the process mechanism Surface roughness and finish quality vary greatly between AM processes Overhangs, bridges, and support requirements also determine the design flexibility and post-processing requirements 3. Compare overall dimensions Some processes can manufacture large parts while others are limited in size. A rough estimation and exclusion of processes can be made at this stage 4. Investigate process build/deposition rate Some processes have faster production times than others 5. Determine post-processing needs Some processes require various post-processing steps to have a final product, increasing the production time cost

Fig. 3.3 Minimum/maximum values of design aspects that lead to improved quality and manufacturability for the different AM process families [2, 4–8]

3.2 AM Part Quality Issues

49

Fig. 3.4 Rules of thumb for the selection of AM process for prototyping

Fig. 3.5 Rules of thumb for the selection of AM process for end-use functional products

3.2 AM Part Quality Issues Currently, one of the most important issues hindering AM is that of part quality [1]. The most important quality issues can be summarized as follows. • • • •

Surface roughness and layer-by-layer appearance. Porosity/void formation. Anisotropic microstructure and mechanical properties. Thermal residual stresses and deformations.

Each of the aforementioned quality issues will be analyzed separately in the following sections.

50

3 AM Processes

3.2.1 Surface Roughness and Layer-by-Layer Appearance Surface roughness defines surface texture and can be quantified through the magnitude and quantity of the deviations that are normal to a given surface [10]. The parameters that affect surface roughness in AM can be classified into two groups: process/material-dependent and user selections-dependent. In the first category, the determining factors are the process mechanism, the machine resolution, and the precision, as well as the feedstock type. Figure 3.6 illustrates the capabilities of the different AM process families regarding surface roughness. It can be observed that Binder Jetting can achieve the best roughness values and is closely followed by laser PBF; meanwhile, electron PBF has the highest roughness values among the PBF family processes. The DED process family displays significantly higher roughness results than the BJ and PBF families. It is worth noting that powder-based feedstock leads to lower roughness values than the wire-based feedstock.

Fig. 3.6 Typical surface roughness ranges for metal AM process families

3.2 AM Part Quality Issues

51

The process mechanism, machine resolution, and feedstock type play a significant role; however, they only set the upper and lower limits in terms of the capabilities they offer. The final roughness result is determined by the user selections regarding the printing setup and the process-parameter values [11]. Layer-by-layer appearance can also manifest as surface roughness. Therefore, smaller layer thickness reduces this effect. Moreover, the orientation of the part in the building platform determines the orientation of its different surfaces with respect to the build direction [11]. The larger the deviation of a surface from the build direction, the higher the roughness of the given surface is. Another important parameter that is inexplicably connected with orientation is the supports [12]. Supports’ interfaces lead to increased roughness after their removal. Therefore, during the selection of the printing orientation, the user must strive toward both the minimization and the avoidance of supports on surfaces with delicate details. It should be mentioned that the PBF process families require fewer supports than the DED ones; however, other issues have to be considered, such as the optimization of the down-skin process parameters (laser power and scanning speed) to minimize roughness for overhanging geometries [13, 14] (Fig. 3.7). Finally, to achieve lower roughness for a given combination of the process– AM machine–feedstock type, the optimization of the process parameters is crucial. The most important process parameters affecting roughness are the layer thickness, laser power, scanning speed, power density, and overlap [16] (which will be further discussed in Sect. 3.3). One rather overlooked factor that can affect surface roughness is the tessellation error that occurs during the conversion of the CAD file to the Standard Triangle Language (STL) file, which is the most common format used by AM slicing software [17]. By increasing the number of elements used in the STL file, the intensity of this phenomenon is drastically decreased. However, the use of excessive elements should be avoided, as this would lead to very large STL files that would be more difficult to slice and process. It has to be noted that post-processing greatly improves the issue of surface roughness (Sect. 3.6); however, an initial low roughness greatly decreases post-processing cost and time. Fig. 3.7 Down-skin roughness of overhanging geometries in PBF [15]

52

3 AM Processes

3.2.2 Porosity/Void Formation Voids can be formed inside the part during AM (Fig. 3.8); void formation can be classified into the following categories [18]. 1. 2. 3. 4.

Processing-induced pores: lack of fusion and incomplete melting. Feedstock-related pores and impurities. Pores from powder compaction: void pockets between powder particles. Metallurgical pores.

The most common reason for pore formation is the first category, which mainly involves incorrect heat input in the melt pool [19] and insufficient material overlap (hatch spacing) [20]. More specifically, if a low energy density is used, the melting and fusion of the new layer with the previous one are not performed correctly, thus resulting in the formation of pores [19]. However, the use of very high energy densities can lead to the same effect due to the creation of a keyhole in the melt pool, which leads to air inclusions during solidification [21]. Additionally, if there is insufficient overlap in the hatch spacing, the fusion between the different paths of the same layer might not be complete, thus leading to the creation of pores. Therefore, careful optimization of laser power–speed and hatching is required to eliminate the first type of porosity formation. Another cause of porosity is the feedstock entrapped pores. Their source can be trapped atomization gas, the reaction of feedstock with moisture, or its oxidation [22]. The use of high-quality feedstock, its preheating, as well as careful handling and storage according to the manufacturer’s standards are required to avoid this cause of porosity [21]. Pores from powder compaction are formed when void pockets remain trapped between powder particles. To significantly improve this, the use of thinner layers, overlapping, and better compaction of the powder is suggested [18]. The formation of metallurgical pores is inherent in the dynamic phenomena that take place during Fig. 3.8 Pore formation in AlSi10Mg processed by SLM [25]

3.2 AM Part Quality Issues

53

Fig. 3.9 Hot cracking during deposition of IN718 [26]

the formation and solidification of the melt pool. This source of porosity cannot be fully predicted or entirely avoided due to its stochastic nature; however, higher energy densities tend to improve this issue [18]. Finally, it has to be mentioned that cracks might also appear in metal AM parts (Fig. 3.9); however, despite that this phenomenon decreases the final density, e.g., porosity formation, its source is different and it is attributed to the development of high thermal stresses during manufacturing [1]. These stresses are caused by excessive spatial non-uniformities of the thermal field (thermal gradients) and can be avoided via energy density and scanning strategy optimization [23], as well as the use of heat-diffusing supports [24]; this will be further analyzed in Sect. 3.2.4.

3.2.3 Anisotropic Microstructure and Mechanical Properties Due to the layer-by-layer nature of the AM processes, orientation dependence of mechanical properties is observed: the mechanical properties along the build axis are inferior to those along the two orthogonal axes [1]. To decrease the effects of anisotropy, careful selection of energy density and hatch spacing has to take place to allow for a sufficient fusion with the previously deposited layer and the adjacent deposited path [27]. Additionally, the orientation of the part in the build chamber has to be optimized to ensure that the optimal mechanical properties are obtained depending on the load requirements for a specific part (Fig. 3.10). However, to effectively minimize the effect of anisotropy, the microstructure should be studied and the impact of the various process parameters should be investigated. The source of the anisotropy lies in the grain evolution during manufacturing (Fig. 3.11). The microstructure can be controlled in situ utilizing re-melting, optimization of build envelope temperature, as well as cooling rates and heating profiles [29], thus improving layer-by-layer grain uniformity [29]. The role of the cooling rate is crucial in the uniformity of microstructure. Slower cooling rates lead to coarser grains, whereas faster rates lead to thinner ones [30]. Moreover, the design of the part plays an important role because thicker cross-sections cool slower than thinner ones,

54

3 AM Processes

Fig. 3.10 Effect of orientation on anisotropy: difference in mechanical strength of PLA parts [28]

Fig. 3.11 Heterogenous grain evolution during recrystallization in SLM

hence leading to coarser microstructures [31]. Furthermore, the cooling rates in the lower layers tend to be faster due to the higher heat conductivity of the baseplate material [32]. Finally, there are techniques in design for AM specifically aiming to minimize anisotropy, such as specialized lattice structures [33]. It should be noted that post-processing greatly improves the issue of anisotropy (Sect. 3.6); however, achieving a satisfactory initial anisotropy greatly reduces post-processing cost and time.

3.2.4 Thermal Residual Stresses and Deformations The uneven heating and cooling that take place during the manufacturing of parts in AM lead to the development of thermal stresses and deformations, both in the build direction (Z-axis) and in the horizontal plane (XY). This leads to the deterioration of

3.2 AM Part Quality Issues

55

part quality in terms of dimensional accuracy and mechanical properties and can even lead to total part failure. Additionally, in PBF, thermal deformations can potentially cause a crash of the re-coater during the spreading of a new powder layer. The cause for this phenomenon is the non-uniformity of the thermal field [1] both in the build direction (Z-axis) and in-plane (XY), and it can be classified into three categories: (i) induced stresses from upper layers to the below solidified layers, (ii) thermal contraction of the current layer, and (iii) thermal gradients in the XY-plane [1, 23, 34]. The first category occurs due to thermal gradients in the Z-axis. More specifically, the temperature of the upper solidified layers increases because they are closer to the heated top layer and they tend to expand. However, the lower layers, which have a lower temperature, stop this expansion; therefore, stresses are induced, which are compressive in the upper layers and tensile in the lower ones (Fig. 3.12). If these stresses are higher than the yield stress of the material, plastic deformations occur. After the part has cooled down, the plastic deformations are converted to residual stresses, and according to their magnitude, they can lead to deformations or cracks [34]. The reason for the development of the second category of thermal stresses/ deformations (i.e., ii) thermal contraction of the current layer) is the non-uniform temperature in the Z-direction [35]. The top layer has a high temperature during its creation, and as it cools down, it tends to shrink. However, the shrinkage is prevented by the layers below it because all layers have already fused, leading to the creation of tensile stress on the top layer and compressive on the layer below [34]. This is graphically depicted in Fig. 3.13. Finally, the third category of thermal stresses/deformations is created due to the uneven temperature distribution in the XY-plane. A pre-determined scanning strategy

Fig. 3.12 Induced thermal stresses in upper and lower layers due to Z-axis thermal gradient

56

3 AM Processes

Fig. 3.13 Thermal stresses due to the cooling down of the top layer

is followed for the heating of the top layer, which in most cases creates a number of outer contours followed by an infill pattern for the internal area of the layer. This uneven heating over time leads to the creation of thermal gradients in the XY-plane, the intensity of which defines the intensity of the induced thermal stresses in the XY-plane [1, 23] (Fig. 3.14). The combined effect of the aforementioned three categories of thermal stresses/ deformations leads to warping, i.e., the deformation in the XY-plane that is nonuniform along the Z-axis. Additionally, the same categories lead to shrinkage, due to which the final dimensions of the AM parts tend to be smaller than those of the CAD file used for the production of said parts [1, 35]. A more uniform distribution of temperature along the Z-axis is required for the reduction of the intensity of both the first and the second categories. This can be achieved by heating the base plate for the purpose of lowering the thermal gradient in the Z-direction. Additionally, the heating of the machine chamber helps in the mitigation of this effect [34]. Moreover, both of the previously mentioned actions

Fig. 3.14 Development of thermal stresses in the XY-plane due to thermal gradients caused by the heating sequence of the layer because of the simple hatch infill pattern

3.3 Process Optimization Strategies

57

decrease the intensity of the in-plane thermal gradients. However, the latter category can be more effectively tackled by evaluating the different scanning strategy alternatives in terms of the resulting XY-plane thermal gradients and by selecting the one that leads to the lowest. This extra scanning strategy selection criterion, which is presented in [23], can decrease the in-plane thermal non-uniformity by up to 20% for a given part (Fig. 3.15). It should be noted that the impact of the scanning strategy selection on XY-plane thermal gradients varies according to the design characteristics of a part, and it is of higher importance for parts with uniform X and Y dimensions. In addition, the existence of holes of increasing diameter increases the importance of the scanning strategy selection. However, after a diameter threshold has been reached, the importance of the scanning strategy selection greatly decreases due to the small XY-plane area of the part. Moreover, the significance of this selection is lower for parts of low thickness featuring large XY-plane surfaces, whereas it slightly decreases as the part thickness increases [1, 23]. Another strategy that is applied to reduce the negative effect of thermal stresses (both due to the Z-axis and XY-plane gradients) is the use of specialized supports aiming both to decrease the development of deformations during printing and to dissipate heat in order to lower thermal gradients [24]. This is mostly utilized for parts with long and narrow details along the Z-axis.

3.3 Process Optimization Strategies The focus of this book is AM of metals and the most common metal-oriented AM process families are PBF and DED, followed by BJ, which has more limited applications. PBF and DED utilize a thermal-based process mechanism for the fusion of new layers; therefore, they have several common process parameters. The most important process parameters of thermal-based metal AM processes and their impact on crucial KPIs will be discussed in the present section because the quality of parts in AM is highly dependent on the optimization of process parameters. Process-parameter optimization aims to achieve minimized/maximized values of the KPIs that are dependent on corresponding process parameters. Tables 3.3 and 3.4 summarize the most important process parameters and KPIs, respectively. In this section, the impact of each process parameter of Table 3.3 on the KPIs of Table 3.4 will be discussed.

3.3.1 Heating The manner in which the heating for the fusion of a new layer with the previous one is performed has a very significant impact on part quality [16] because it affects almost all the KPIs of Table 3.4. More specifically, the heating temperature has to be adjusted to ensure the melting of the entire mass of the current layer, as well as a part of the layer directly below it [1]. This will ensure that the desired density is achieved,

58

3 AM Processes

Fig. 3.15 Intensity of thermal gradients in the XY-plane for different parts and infill patterns expressed as normalized stress formation tendency index [1, 23]

that no unmelted feedstock will remain, and that the correct fusion between layers is achieved. In this section, the focus is on the most common metal AM heat source, namely the laser; however, the basic principles of the heating process mechanism are the same for processes that use a different heat source, such as EBM. The role of the heat source is selective and controlled heating, leading to increased temperature and melting in order to fuse a new layer with the previous one. However, it should be noted that there is a connection between the laser power and scanning head speed. The combination of heating intensity (laser power) and scanning speed form

3.3 Process Optimization Strategies Table 3.3 Important process parameters in AM

59

• Heating – Heating temperature – Laser power (laser-based) – Melt-pool dimensions – Thermal gradients – Pre-heating – Re-melting – Base plate and chamber temperature – Cooling profiles • Scanner head related – Scanner head speed – Hatch spacing – Scanning strategy, raster angle, raster width • Layer thickness

Table 3.4 Important KPIs in AM

• Surface roughness • Mechanical properties – Part strength (tensile, impact, flexural) – Residual stresses • Microstructure – Porosity and void formation – Density • Topology and dimensional accuracy – Deformations, distortions – Part shrinkage • Energy consumption • Build time

the energy density, which is the determining factor for the maximum temperature, melt-pool dimensions, and the fluid-dynamics phenomena that take place [21]. These phenomena determine the surface roughness, part density, mechanical properties, and microstructure of the final parts; therefore, their optimization is of crucial importance [16]. Energy density and the different combinations of laser power and speed have to be defined taking into account the material used, the required part density, build time, and cost of the part, as well as the up-skin and down-skin considerations (Fig. 3.16). Low energy density leads to limited liquid formation and the balling effect (Fig. 3.17), whereas excessive energy density has a negative impact on surface quality, even leading to burnt and failed parts.

60

3 AM Processes

Fig. 3.16 Qualitative depiction of the effects of energy density (combination of laser power and scanning speed) in AM

Fig. 3.17 Balling effect in metal AM due to insufficient energy density and wetting

3.3 Process Optimization Strategies

61

Fig. 3.18 Steel parts manufactured in SLM without baseplate heating [34]

Feedstock preheating is important because it ensures possible moisture elimination [22]. Additionally, it increases the speed of the laser heating process because the feedstock is already at a higher temperature during its spreading (PBF) or spaying (DED). Particularly in PBF, the spreading of a preheated powder layer leads to smoother thermal gradients, thus improving microstructure homogeneity and reducing thermal stresses [23]. Re-melting can be utilized to further improve microstructure homogeneity, as well as to achieve almost fully dense parts [29]. However, it significantly increases processing time. The heating of the base plate is a common practice in AM of metals and it is usually used in plastics, as well. The combination of re-melting with a heated environment significantly improves microstructure, anisotropy, as well as the thermal stresses and deformation of the final parts [36]. Without the use of baseplate heating in metal AM, the possibility of thermal cracking and failed parts highly increases (Fig. 3.18). Cooling profiles are determined by the heat dissipation during the manufacturing of a part and are, therefore, a combined result of all the previously mentioned heating parameters [35].

3.3.2 Scanner Head Speed and Scanning Strategy One important aspect that is common in all AM processes is the movement of the scanner head. It is defined by the speed of the scanner head, the path sequence used for the scanning of a layer (scanning strategy), and the hatch spacing (overlap). In addition, the scanner head process-parameter group has an impact on all KPIs because it defines the area, the speed, and the sequence that a layer is heated; thus, it is directed in synergy with the heating process-parameter group. More specifically, the scanning speed has to be adjusted considering the desired laser power because the energy density is a result of the combination of laser power and scanning speed [36]. The scanning strategy is the path sequence followed for the scanning of a layer, and it comprises the contours, the infill, and the support structures. It has a direct impact on residual stresses, deformations, mechanical properties, anisotropy, and build time. Typically, a couple of contours of the outer surface of each layer are first

62

3 AM Processes

scanned, followed by the scanning of the internal surface of a layer utilizing a predetermined infill pattern. AM slicing software offers pre-defined scanning strategy/ infill patterns that the user can select [37]. Common infill patterns are the simple hatch and consecutive contours, as well as the ones depicted in Fig. 3.19. In metalbased AM, the most commonly used infill patterns are the simple hatch, stripes, and chessboard (Fig. 3.20). There are several criteria for the selection of scanning strategy (Fig. 3.21): (i) the minimization of anisotropy or obtaining high mechanical properties in a selected axis, (ii) build time decrease, (iii) specific design considerations, such as optimization of the internal topology regarding static and rotational stability, and (iv) the minimization of the thermal gradients caused by the heating sequence [1] (Fig. 3.14). When the simple hatch infill is used, the raster angle is the main factor that defines the directionality of the mechanical properties of the part, particularly in the material extrusion (MEx) process family [38, 39], whereas it plays a less significant role in

Fig. 3.19 Common infill patterns [23]

Fig. 3.20 Most commonly used infill patterns in metal AM

3.3 Process Optimization Strategies

63

Fig. 3.21 Different scanning strategy selection criteria [23]

metal AM [16]. Therefore, regarding MEx, if the load case requirements are on a specific axis, the best mechanical properties are obtained using a raster angle of 0°, whereas if not, the use of a 0°/45°/90°/ − 45° orientation leads to a close-to-isotropic behavior [23]. Build time is an important factor because it is directly connected with productivity, which is one of the issues hindering the wider industrial AM uptake. Scanning strategy can play an important role in the decrease in building time. To achieve a decrease in build time, the main factor of the scanning strategy that can be optimized is the rapid movements of the processing head, during which no material is being printed. The Hilbert curves’ infill strategy is specifically optimized toward this goal in [40], thus eliminating non-printing movements and simultaneously leading to parts with close-to-isotropic mechanical properties. Hatch spacing defines the overlap between two adjacent tracks in PBF, thus affecting the fusion between them, as well as the resulting microstructure [21] and surface roughness [41] (Fig. 3.22). A finer hatch spacing (high overlap) leads to excessive re-melting of the adjacent track, whereas a coarser hatch spacing (small overlap) leads to decreased adjacent track re-melting. Additionally, finer hatch spacing increases build time.

3.3.3 Layer Thickness The selection of these process parameters takes place during the slicing of the part design and it is one of the most basic process parameters. Their selection has a direct impact on both part quality (in terms of surface roughness and resolution) and on build time. More specifically, a smaller layer thickness will provide better resolution in the Z-direction (layer-by-layer appearance); however, it will significantly increase

64

3 AM Processes

Fig. 3.22 Resulting as-built solidified part surface using increasing hatch spacing from (a–d)

build time because more layers will need to be printed. The minimum possible layer thickness for the different AM families can be seen in Fig. 3.3. The technique that ensures the best characteristics of both thicker layers (faster build time) and thinner ones (resolution) is adaptive slicing, which is qualitatively depicted in Fig. 3.23.

Fig. 3.23 Adaptive slicing

3.4 Simulation for AM

65

3.4 Simulation for AM Simulation is the utilization of a computer for the solution of a model of a system; the model is a mathematical representation of the real-world system that includes assumptions and simplifications. Simulations in engineering are convenient software tools that allow for the optimization of process parameters aiming at KPI enhancement and the minimization of experimentation cost and time. Additionally, they offer a deeper understanding of the process mechanisms, leading to better intuition and problem-solving capabilities. Moreover, simulation is inextricably linked with advanced techniques for design for AM, such as Topology Optimization and genetic algorithms; therefore, it is a necessary tool for competent AM designers, as well.

3.4.1 Modeling Approaches and Scales For the creation of a simulation, the phenomena under study have to be modeled through mathematical expressions. There are three major types of modeling approaches: analytical, numerical, and empirical; each offers different advantages and limitations according to the AM process and process parameters/KPIs of interest. An analytical model is a set of mathematical equations describing the physics of a system to predict its outputs/behaviors based on certain inputs. This modeling approach provides exact solutions, capturing the physics of the systems, and is fast to compute. However, analytical models are highly complex for most dynamic realistic manufacturing scenarios and require extensive assumptions to become practical and solvable [16]. The empirical/experimental approach is straightforward and practical, setting up correlations between process parameters and KPIs of interest. Such models are practical, relatively easy to develop and provide accurate solutions for specific problems. However, they require costly and time-consuming experiments, they do not establish a connection to the physics of the process, and they are directly dependent on the specific conditions of the model calibration experiments that have been conducted [19]. Numerical models are based on the principles of numerical analysis and are used to approximately solve complex sets of equations—which describe a physical system—through an iterative process. Due to the large number of iterations involved, a computer system is required. Numerical simulations are capable of solving very complex, dynamic, and realistic problems, offering tremendous capabilities in process optimization and part design as well, meanwhile offering an overview of the physics of the process. Their disadvantage is that due to the complexity of the phenomena that take place in AM, they tend to have high computational demands and require specialized AM simulation software [1, 19]. Currently, the numerical modeling approach is the most widely used for the development of AM simulations. Figure 3.24 illustrates the number of studies using different modeling approaches for different AM process families [16]. It may be observed that the numerical approach is the most dominant in AM families in which the simulation of

66

3 AM Processes

Fig. 3.24 Number of studies using each modeling approach for different AM process families [16]

complex phenomena is required, such as thermo-mechanical and computational fluid mechanics (CFD) simulations. The same conclusion may be drawn from Fig. 3.25, in which the number of studies using each modeling approach for different KPIs [16] is depicted. The phenomena that take place in AM are implicit, multi-scale, and highly dynamic in time and space. Table 3.5 summarizes the simulation requirements corresponding to each AM phenomenon. It may be observed that for the creation of an all-encompassing holistic simulation, a coupled multi-disciplinary and multi-level simulation is required. However, such an approach would have prohibitive computational requirements. Instead, the study of the different scales and phenomena in separate simulations is the most common approach, using the optimized process parameters from one simulation level to the next. To further understand how to utilize simulation for AM efficiently, the connection of the phenomena that take place and their impact on quality have to be established. It should be noted that different phenomena require the simulation of different scales based on their physical mechanism. In Table 3.6, the phenomena taking place in AM are categorized in scales along with their corresponding KPIs. Utilizing this categorization allows for the study of specific part characteristics (KPIs) individually, through the simulation only of the dominant connected phenomena. Furthermore, Fig. 3.26 illustrates the different modeling approaches that are used according to the phenomenon scale.

3.4 Simulation for AM

67

Fig. 3.25 Number of studies using each modeling approach for different KPIs [16]

Table 3.5 Simulation requirements for the different phenomena in AM [1] Phenomenon

Simulation requirement

Highly dynamic temperature field over time and space

Transient thermal simulation

Coupled thermal stresses

Coupling with transient mechanical simulation

Melt-pool fluid dynamics

Coupling with computational fluid dynamics simulation

Movement of heat source

Movement of heating boundary condition

Material addition over time

Increase mesh size over time or mesh adaptation (numerical)

Phase changes (melting, evaporation)

Include the corresponding latent heat

Coexistence of different material states (powder, liquid, gas, solid)

Different thermal and mechanical material properties, contact boundaries between different states, and their interaction

Complex geometries and a very small heating Very fine mesh requirements (numerical) area

Lately, there has been an increasing interest in nanoscale simulations that is specially geared toward capturing the physics involved during the use of functionally graded materials. However, the most common industrially focused simulation approaches and commercial software focus on the micro- and macro-scales because the dominant phenomena mostly take place in these scales. To roughly simplify

68

3 AM Processes

Table 3.6 Connection of phenomena, scale, and relative impact on quality (KPIs) for metal AM [1, 2, 42, 43] Scale

Phenomena

KPIs

Nanoscale

Laser–material interaction, the interface of different materials at an atomic level

Crystallization, grains, and cluster evolution, especially for functionally graded materials

Micro-scale

Moving heating boundary, melt-pool fluid dynamics, and powder interaction, phase changes

Surface roughness, microstructure, porosity, grain size

Macro-scale

Heating following scanning strategy, layer addition

The final part is thermal stresses and deformations, the stair-case effect

Fig. 3.26 Different simulation approaches used in AM according to the modeled phenomenon scale [43]

simulation for AM, simulations can be classified in terms of scale and modeled phenomena into two major groups. (i) Melt-pool fluid dynamics phenomena: the dynamic transition from powder to liquid to solid, as well as the possible creation of keyhole due to vaporization. These phenomena determine microstructure, void formation, density, and roughness. Such simulations are performed for very small parts (commonly a single or a small number of heated tracks) and correspond to very short manufacturing time. (ii) Highly dynamic thermal changes. They are responsible for the induction of thermal stresses that lead to the deterioration of mechanical properties, the

3.4 Simulation for AM

69

possible loss of dimensional accuracy due to thermal deformations, and potentially failed parts due to thermal cracking or extensive deformation. The simulation of at least some layers is required to draw substantial conclusions, and the simulation of the entire part considering the exact scanner head movement is ideal. A holistic optimization of the different process parameters can be realized by understanding the purpose and the goals of each simulation scale, as well as how they can be utilized complementary to one another. A discussion about the simulation requirements of the aforementioned two important phenomena follows.

3.4.2 Simulations of Melt Pool-Related Phenomena The melt-pool fluid dynamics and the phase changes due to heating and cooling are the dominant phenomena in this scale. They are described by the continuum equations of the conservation of mass, momentum, and energy. ∇(ρU) = 0,

(3.1)

∂ (ρU) + ∇ · (ρU U) = −∇ p + ρg + ∇ · (μ∇U) + Su , ∂t

(3.2)

∂ (ρ H ) + ∇ · (ρU H ) = ∇ · (k∇T ) + ST , ∂t

(3.3)

where U is the velocity, g the gravitational acceleration, ρ the density, μ the viscosity, p the pressure, T the temperature, H the enthalpy of the material, k the thermal conductivity, and Su and ST are the momentum and energy source terms, respectively [44]. The term Su encompasses four types of forces [21]. i. The Darcy forces, which act as a momentum sink when the temperature of the part corresponds to the solid phase. ii. The surface tension that has a normal direction to the gas–liquid interface. iii. The surface tension that has a tangential direction to the gas–liquid interface (Marangoni forces). iv. The recoil pressure caused by the evaporation. The energy source term ST takes into account the latent heat during the melting phase change. The laser heat input is also incorporated within this term [21]. The aforementioned phenomena occur in the micro-scale. The computational resources required to allow for simulations of very small parts (one or two laser tracks) and manufacturing time are in the scale of approximately 100 μm and of a few milliseconds.

70

3 AM Processes

The information provided in such simulations is twofold. Initially, they allow for the optimization of highly important process parameters, namely laser speed and power, layer thickness, hatching distance, and powder particle characteristics, in terms of increased density, pore minimization, improved surface roughness, keyhole evolution, and minimization of the balling effect (Fig. 3.27). Additionally, they can be utilized to create a connection between micro-scale phenomena and melt-pool temperature. In this manner, it is possible to have an insight into the intensity of the aforementioned micro-scale phenomena based on the thermal field history (analyzed in Sect. 3.4.3).

3.4.3 Thermo-Mechanical Simulations The dominant phenomenon is laser-induced heating and heat conduction through the mass of the part, which is the cause of thermal expansions that lead to thermal stresses due to constraints set by the already solidified material (Sect. 3.2.4). The temperature throughout the mass of the material can be calculated over time by solving the differential equation of heat conduction in a 3D space [1]. ∇2T =

k ∂T , c p ∂t

(3.4)

where T is the nodal temperature, c p is the specific heat capacity, and k is the thermal conductivity of the material. For laser-based AM, which is the most common heating application in AM, the following set of equations describes laser-induced heating.  ∂ T  k = I. ∂z x=m

(3.5)

z=n

A Gaussian profile can describe the laser beam intensity, I , as a function of distance from the laser beam axis. I = I0 e

−2

 2 ld r

,

(3.6)

where ld is the distance in the x-axis of the node from the laser beam axis, r is the radius of the laser beam, and I0 is the intensity of the laser beam at the beam axis and the focal level; I0 is specified using the laser power, P, and the laser spot radius, r , as I0 =

2P , πr 2

(3.7)

3.4 Simulation for AM

71

Fig. 3.27 Melt-pool formation for different energy densities obtained by changing scanning speed in thermally coupled CFD simulation in PBF of metals [21]

72

3 AM Processes

Fig. 3.28 Temperature profile on the cross-section of the molten pool [36]

where P is the laser power. Thermal losses to the environment due to convection and radiation, as well as contact with the base plate, have to be accounted for as well. Figure 3.28 illustrates the temperature profile on an XZ-plane cross-section, where the melt pool is visible. This was calculated via the finite difference numerical thermal simulation developed in [23, 36]. Utilizing such a simulation, it is possible to know the melt-pool dimensions, maximum temperature, and thermal gradient intensity at any given time during the manufacturing of a part. The cause of thermal stresses, as analyzed in Sect. 3.2.4, is the thermal gradients caused by the moving heat source. More specifically, the temperature gradients cause displacements, which translate to thermal stresses and deformations in parts [1, 23, 45, 46]. This is expressed via the Generalized Navier equation. ¨ μ∇ 2 u + (λ + μ)∇(∇.u) + b∇T = ρ u, λ=

(3.8)

Eν , (1 + ν)(1 − 2ν)

(3.9)

αv E , 1 − 2ν

(3.10)

b=

where μ is the shear modulus, u is the displacement vector, λ is Lamé’s first parameter, T is the temperature, ρ is the material density, E is the Young modulus of the material, ν is the Poisson’s ratio of the material, and αv is the volumetric coefficient of thermal expansion of the material. The equations provided in this chapter form the basis for the development of a coupled thermo-mechanical simulation capable of calculating the temperature over

3.4 Simulation for AM Table 3.7 Typical values of parameters in PBF that are important for simulations

73

Parameter

Value

Laser spot diameter

50–100 μm

Scanning speed

0.5–2 m/s

Layer thickness

30–100 μm

Typical part dimensions

5–50 cm

space and time and of using it as input for the calculation of the resulting displacements. According to the state of the material and the constraints set by the solidified layers, the thermal stresses can be calculated; these stresses lead to deformations or even total failures due to cracks, depending on crack magnitude (Sect. 3.2.4). Thermo-Mechanical Simulation Challenges The development and utilization of practical thermo-mechanical macro-scale numerical simulations for metal AM are quite challenging in terms of computational cost and required running time. Taking as an example a typical metal AM process, e.g., PBF, the defining parameter for the mesh size is the laser spot size and the layer thickness. However, the typical part dimensions are several scales of magnitude greater (Table 3.7), leading to a very high number of elements/nodes. Additionally, the application of the heat source should change over time according to the corresponding scanning strategy. Layer addition and phase changes should also be incorporated. The combination of the above with the requirements of a coupled thermomechanical transient simulation constitutes macro-scale simulations for metal AM very computationally intensive. A strategy for the mitigation of this issue is the use of mesh adaptation over time [36]. More specifically, a thin discretization is used in the area close to the laser heating, which becomes gradually coarser for the parts of the mesh that are situated farther away from the laser. This allows for significant computational time conservation, while simultaneously preserving simulation accuracy because the spatial intensity of the phenomena dwindles the farther away from the laser. However, the heating source is moving; therefore, for this approach to be successful and accurate, constant mesh adaptation is required as the heating boundary condition moves. This requires the use of specific AM simulation software or the development of a tailored AM simulation package from scratch. Other manners of decreasing computational costs are the use of line heating [47]—as opposed to moving point heating—as well as layer lumping [48]. Both are employed as assumptions because they do not simulate the exact process mechanism. However, they can lead to results of acceptable accuracy, given that they are correctly calibrated and validated for the problem at hand. Additionally, the approach presented in [35] helps to decrease the computational costs in thermo-mechanical simulations, utilizing an empirical boundary condition that compensates for plastic deformations caused by thermal gradients; it requires only the solution of the thermo-elastic model. Finally, in the approach of [23], the stress formation tendency index (SFTI) has been developed, which encompasses the tendency for the development of thermal stresses

74

3 AM Processes

Fig. 3.29 Evaluation of three alternatives in terms of the intensity of thermal stresses and deformations using only a thermal simulation and the SFTI metric developed in [23]

based on the intensity of thermal gradients. Although the exact values of thermal stresses and deformations cannot be calculated with this approach, it enables the fast evaluation of different process-parameter scanning strategy alternatives. Thus, the alternative that leads to minimum thermal gradients and, consequently, to minimum stresses and deformations, can be identified in an industrial-friendly and practical manner in terms of computational time–cost and user experience (Fig. 3.29).

3.4.4 Multi-scale Simulations As per the previous analysis, it is apparent that the combination of all phenomena in a single simulation poses issues of practicality due to the required computational cost and time. However, a complete overview can be obtained regarding the optimization of all the crucial process parameters through the combination of different simulation levels. More specifically, a lower-scale model can be utilized for the optimization of certain parameters and for the creation of a correlation between the values of important process parameters with temperature. Once these have been established, the use of a practical macro-scale simulation—taking into account the exact part design and scanning strategy—can offer the desired overview for the complete process optimization, thus minimizing experimentation costs. There are various commercial AM specialized simulation software’s that either focus on micro-scale melt-pool simulations, macro-scale thermo-mechanical ones, or combine both of them both, offering a variety of process-parameter and part design optimization functions. Additionally, some allow for the use of paid cloud-computing services for the solution of computationally intense simulations.

3.5 Monitoring, Control, and Quality Assessment

75

3.5 Monitoring, Control, and Quality Assessment 3.5.1 Monitoring and Control Process-parameter optimization is required to ensure the successful manufacturing of parts of high quality. One step further toward this goal is the use of monitoring systems and adaptive process control, which aims to maintain quality measures (KPIs) within specific boundaries [49] through the real-time variation of process parameters. Such systems include sensors, data processing equipment, actuators, networks for the connection of the equipment, and algorithms to relate process variables to product attributes [50]. A comparison can be seen in Fig. 3.30 between controlled and uncontrolled AM systems regarding temperature distribution along the axis parallel to the laser beam. Monitoring systems are responsible for gathering data regarding a specific KPIrelated parameter and their development is a prerequisite for adaptive control. The selection of the monitoring system is based on the quality aspects that have to be optimized, as well as on the cost and requirements for a specific production plan. Monitoring systems can be classified into two types: direct and indirect. The first type can offer high accuracy; however, there might be practical limitations in its use. On the other hand, in the latter type, auxiliary quantities are measured and are correlated empirically with the quantities of interest [51]. The most important monitoring system types that are currently utilized for AM are listed in Table 3.8. The next step is the development of a system capable of real-time comparison of the monitored data to experimental/model-based data and the real-time modification of process parameters to meet the specified goals. The most common process parameters used for adaptive control in AM are the laser power and speed because they determine the energy density and melt-pool geometry, both of which are crucial for part quality (Sects. 3.3.1 and 3.3.2). The values of the controlled process parameters

Fig. 3.30 Comparison between the temperature variation over the distance of the laser spot between controlled and uncontrolled AMs [49]

76 Table 3.8 Important monitoring and inspection systems used in AM [2]

3 AM Processes

• Cameras and image processing • Thermal imaging – Charge-coupled device (CCD) cameras – Light spectroscopy • Melt pool spectroscopy – Laser-induced breakdown spectroscopy (LIBS) • Inspection—to be applied in real-time in future applications – Ultrasound and eddy current non-destructive testing (NDT) – X-rays and X-ray tomography

are constantly modified to ensure that the measured KPIs are within the desired range based on a model or experimental data that correlate the process parameters with the KPIs. The procedure for the creation of a model-based adaptive control scheme is depicted in Fig. 3.31. According to the phase of the process (melting, solidification, cooling down), the control algorithm investigates whether the corresponding criteria have been met and if not, the corresponding control law is applied [52]. As a rule of thumb for the development of a successful control scheme, the questions of the flowchart of Fig. 3.32 can be used. It has to be highlighted that through the utilization of IoT and Industry 4.0 technologies, the development of cloud-based control applications has been made

Fig. 3.31 Model-based adaptive control scheme of laser power for AM [49]

3.5 Monitoring, Control, and Quality Assessment

77

Fig. 3.32 Successful control flowchart [53]

possible. This allows for easier and more user-friendly process-control applications because it does not involve the user in the creation of the control scheme, model, or the required experimentation for their calibration. The only requirement from the machine user is to install the monitoring system on the AM machine and the equipment for the connection with the control service provider. In cloud-based control applications, the monitored data are transferred from the machine to the remote control unit, where the control signal is generated based on the control laws and optimized process-parameter values of the service provider. The signal is then transmitted back to the AM machine and the control is enforced. This procedure is repeated in pre-determined intervals, allowing for real-time control (Fig. 3.33). However, to ensure the safety of these cloud transactions, certification protocols have to be followed [53].

3.5.2 Quality Diagnosis Quality diagnosis systems can be classified into two categories: (i) on process (realtime), which utilize process mechanism monitoring systems to detect defects during manufacturing of the part, and (ii) post process that typically involve non-destructive testing techniques to detect anomalities of the produced part.

78

3 AM Processes

Fig. 3.33 Remote control utilizing certified expert agent [53]

Regarding the real-time systems, the data gathered from the sensors are analyzed utilizing feature extraction and empirical models, such as neural networks and machine learning, in order to detect defects and assess part quality in real time. Necessary components for a comprehensive quality diagnosis platform include (i) human–machine interface (HMI), (ii) real-time process monitoring and quality diagnosis systems, (iii) interfaces allowing for high integration ability and interoperability [54]. The architecture of a three-stage quality diagnosis platform for laser-based manufacturing processes using MWIR and NIR cameras is described below and illustrated in Fig. 3.34. Stage 1—Feature Extraction Feature selection and feature extraction methods comprise dimensionality reduction approaches and they can be classified into linear and nonlinear approaches. Their purpose is the elimination of redundant information from the incoming data (in this case, images) to reduce the required resources for their analysis. The Principle Component Analysis (PCA) is a commonly used method due to its automated feature extraction capabilities and relatively low algorithm complexity, which is based on the covariance of the values of the pixels [54]. The most important actions taken in this stage are the following. • • • • •

Read the experimental data. Determine the size of the datasets. Calculate the sample mean and standard deviations vectors. Standardize the data (centering and scaling of the data). Derive Covariance Matrix.

3.5 Monitoring, Control, and Quality Assessment

79

Fig. 3.34 Architecture of a three-stage quality diagnosis platform for laser-based manufacturing processes, utilizing PCA and geometrical feature extraction (GFE) in the first stage, SVM in the second, and HMM in the third [54]

• Compute the eigenvectors and eigenvalues. • Transform the data. • Derive the required components based on a cumulative variance threshold. Stage 2—Quality Assessment Models Defect classification and the prediction of part quality take place in this stage. Defect detection can be achieved either directly, by utilizing a physics-based procedure, or indirectly, through empirical models such as neural networks and machine learning. In either case, the dimensionally reduced data of each frame are fed into the model, which then classifies them as acceptable or not acceptable. Machine learning (ML) algorithms are the most common approach utilized for QA in AM. However, utilizing this approach initially requires a supervised classification procedure for the prediction method to be implemented. There is a plethora of ML models to choose from. In [54], several ML algorithms have been tested regarding their performance for laser welding and AM applications; it was found that a support vector machine (SVM) with linear kernel yields relatively good results concerning the classification success rate for dimensionally reduced PCA melt-pool image data. Stage 3—Decision on the Overall Quality Although the quality of each frame is detected and classified, a decision needs to be made regarding the overall quality evaluation for each continuous path, as well as for

80

3 AM Processes

an entire layer. The use of statistical models can provide the means for this goal. In the approach of [54], the hidden Markov models are utilized for the decision-making process, combined with the maximum likelihood criterion, for the quality assessment of a path. The same logic can be applied for the evaluation of the decision for each layer, which consists of a sum of paths. Regarding the non-destructive testing methods, there are various inspection techniques for the inspection of the surface, utilizing visual enhancements (magnification) or fluorescent liquid penetrants. Such approaches are not limited to AM products and they are applied for quality inspection of various manufacturing parts. The same can be stated regarding dimensional inspections, which can be undertaken using conventional methods. Of particular interest in AM are the internal inspection of parts, since these techniques can offer information regarding the structural integrity of the parts. The most common methods of internal part inspection are radiography, electromagnetic, ultrasonic, and computed tomography. From the aforementioned methods, the most commonly used in AM are digital radiography (DR) and computed tomography (CT), as they provide the most reliable results. CT is preferred for internal geometry validation, while DR is better at gross defect detection. The existence of pores and cracks can be detected and the successful fusion between the layers can be validated; however, neither of the methods has enough sensitivity to detect defects at the layer level [55]. Process Compensated Resonant Testing (PCRT) is a technique that has significant potential for non-destructive testing in AM products. The mechanism of PCRT is based on the measurement of resonant frequencies of a component, which are influenced by a part’s stiffness, geometry, and mass. The stiffness of a part and, as a result, its resonant frequencies differ when structural defects exist in a part. The frequency change is proportional to the change in stiffness and the severity of the defect for a given shape and mass and can therefore be used not only for their detection but also for the quantification of their severity [56].

3.6 Post-processing Post-processing for AM can be divided into two categories: essential and nonessential. All actions that constitute a minimum requirement and must always be performed are included in the first category. • Cleaning of the part, and removal of the excess material from the build chamber (e.g., powder-based processes). • Removal of the part from the base plate. • Removal of the supports. Once these actions have been completed, a great variety of other post-processes can take place, depending on the part requirements, intended use (end-use part or prototype), material, and AM process [56]. A common family of post-processing operations is related to the improvement of part geometry (dimensional accuracy and

3.6 Post-processing Table 3.9 Categorization of AM post-processes based on process mechanism [2]

81

Cleaning processes

Electrothermal processes

• Manual cleaning of the components

• Electrical Discharge Machining (EDM)

• Automatic powder removal

• Electrostrengthening

• Cleaning of the part in an ultrasonic bath

• Electrospark deposition

• Blasting of AM components • Induction heating with a cryogenic CO2 jet Mechanical material removal

Chemical processes

• Cutting

• Chemical etching

• Abrasion

• Chemical machining

• Milling

• Chemical brightening

• Turning (used for axisymmetric parts) Thermal processes

Laser-based processes

• Sintering

• Laser shock peening

• Annealing

• Laser ablation

• Stress relieving

• Laser polishing

• Quenching • Tempering • Aging • Hot isostatic pressing

outer surface) [58]. In addition, improving the structural integrity and the mechanical properties of parts (such as porosity and residual stresses) [59] is essential for metallic AM. It should be noted that post-processing plays an important role in the final part cost, and as such, it needs to be taken into account. A detailed categorization of the post-processes used in AM based on their process mechanism is summarized in Table 3.9. An analysis of the most prominent and common post-processing methods for metallic AM will be further discussed.

3.6.1 Cleaning Processes This is an essential step, particularly for powder-based AM processes such as PBF and binder jetting, in which the unused powder is thoroughly removed from the built part. Special safety measures have to be taken because powder can be harmful to humans. There are several industrial solutions dedicated to cleaning, including automated powder removal systems, ultrasonic baths, and cryogenic blasting. In Table 3.10, the benefits and limitations of such approaches have been summarized [56].

82

3 AM Processes

Table 3.10 Benefits and limitations of different cleaning post-processes Method

Benefits

Vibrations and gravity

• • • • •

Limitations

Least expertise required • No view on dead spaces requires visual and internal inspection Particle removal by excitation • Need to know the ideal Simple equipment needed orientation Ease of use The vibration will detach particles • Need to know natural frequencies of all available material (vibrations cleaning)

(Electro)magnetism • Individual or clusters of powder could be controlled at the same time

• Creating Maglev may be challenging • No previous knowledge

Ultrasonic

• Ideal for micron-sized particles • Different frequencies target different particle sizes • Demi water enhances particle removal

• Reflection from acoustically hard materials • Can remove micron-sized particles • Long cleaning times

Chemical flush

• Non-corrosive • Low boiling point

• Evaporates at room temperature and pressure (RTP) • Toxic

3.6.2 Mechanical Material-Removal Post-processes Mechanical material-removal post-processes are utilized very often for the postprocessing of AM-built parts. Their main applications include support removal, surface finish improvement, and enhancement of dimensional accuracy. The most common material-removal processes utilized for post-processing in metallic AM are milling, turning, grinding, and abrasive flow machining [10]. Material support and baseplate removal, which are essential steps in metal AM, can be performed either manually or by utilizing an automated system. The advantages and limitations of both options are presented in Table 3.11. The mechanical material-removal processes are the most common and easily accessible ones and can be found in any machine shop; they are a well-known subject to the readers of this book, and as such, they will not be analyzed one by one any further.

3.6.3 Thermal Post-processes Particular emphasis must be placed on the thermal-based post-processes because they allow the minimization—or even complete elimination—of the anisotropy, which is inherent to all AM processes due to the nature of their process mechanisms and the residual stresses (for thermal AM processes). Moreover, thermal post-processes

3.6 Post-processing

83

Table 3.11 Benefits and limitations of different mechanical material-removal post-processes Method

Benefits

Limitations

Manual support • Quick decisions if a near-net shape has removal support components that can or cannot be removed • Low expertise operators • Low equipment cost

• Requires a significant amount of manual work • Labor and time to manufacture the part increases

Automated • Consistent, high-quality results support removal • Reduced cycle time for removing supports

• Higher equipment cost • Qualified employees • The expertise of the operators

allow the calibration of the mechanical properties of a product (ductility, tensile strength, yield strength, elongation at break), as well as of its grain size [60], to the exact needs of a particular application within the limits of the material used for its production. Sintering Sintering is the process of fusing particles into a solid mass by using a combination of high pressure and high temperature without reaching the liquefaction point of the material. For metal binder jetting, sintering is an essential post-processing step for load-bearing parts because the bonding that the binder offers to the metal part cannot withstand high mechanical loading (“green part”). Sintering causes part shrinkage; this needs to be accounted for in the design phase. Annealing, Stress Relieving The aforementioned post-processes are widely used in metal AM since they effectively address structural integrity challenges that are inherent in the AM process mechanism like. • • • •

Residual stresses reduction. Porosity reduction. Reduction of anisotropic microstructure. Increased ductility.

The process steps that are common to both annealing and stress relief are (i) heating, (ii) holding the part in the heated environment, (iii) cooling, and (iv) the possibility of multiple applications of the previous; however, there are differences regarding the maximum temperature, holding time, and cooling rates between them because they lead to different effects in the part (Fig. 3.35). The main difference between stress relieving and annealing is that stress relief does not change the chemical or mechanical properties of the material because its objective is to reduce the residual stresses. Therefore, the metal part is heated to a temperature below its lower critical point and the cooling process is slow. The maximum temperature can be

84

3 AM Processes

Fig. 3.35 Annealing and stress relief: effect on tensile strength, ductility, and residual stresses according to the maximum heating temperature

significantly lower than the critical recrystallization temperature in order not to affect the microstructure of the part. In annealing, the material is heated above its recrystallization temperature for a set amount of time. Under these conditions, atoms migrate in the crystal lattice and the number of dislocations decreases, thus causing increased ductility and lower strength and hardness. New uniform grains replace those that have been deformed by internal stresses (recrystallization phase). This is followed by controlled and slow cooling, leading to parts with decreased anisotropy, reduced chance of crack propagation, and improved corrosion behavior [59]. Tempering, Quenching This thermal post-process is utilized for the strengthening and hardening of ironbased alloys through heating, rapidly cooling, and reheating. In tempering, the part is heated to a point that it becomes ductile; if heated above a certain point, the grain (molecular) structures are changed. In quenching, hardness and strength are increased at the cost of brittleness and potential cracking. If those processes are combined at varying speeds and temperatures, the desired grain structure and mechanical properties can be achieved, which can differentially vary within the volume of the part. For example, a combination of high hardness and strength surface can be achieved, while retaining a highly ductile part interior [60].

3.6 Post-processing

85

Aging Metal aging is used on solution heat-treated metal alloys and can be achieved either artificially or naturally. In natural aging, metal precipitates are formed in supersaturated alloying elements; these precipitates block dislocations in the metal and increase its strength and hardness while reducing its ductility. In artificial aging, this procedure is accelerated in a solution heat-treated metal alloy. The process that is followed is heating below the recrystallization temperature; meanwhile, the temperature is high enough to speed up precipitate formation. Then, the metal alloy is cooled rapidly to prevent any further change in the metal precipitates [60]. Hot Isostatic Pressing (HIP) HIP can effectively close the porosities that have been observed in the as-built samples. During HIP, the metal is compressed in a chamber at elevated pressure and temperature using an atmosphere of an inert gas (usually Argon) in the chamber. Compression is achieved through the high operating pressure of the inert gas. HIP provides a coarser microstructure than that of the as-built part, and at high temperatures, it can eliminate the anisotropy of the AM-produced part and can reduce crack propagation. However, HIP also reduces the strength of the part, compared to that of its as-built state [59].

3.6.4 Electrothermal Post-processes This family of post-processes utilizes heat that is created through electric current as its process mechanism; therefore, it can only be used on materials that are electric conductors, such as metals. Their applications include support and baseplate removal, surface finish improvement, and enhancement of their dimensional accuracy. The most common electrothermal post-processes used in metal AM are wire-EDM and die sinking-EDM. They are preferred over mechanical material-removal post-processes for materials with low machinability, such as Inconel 718, 618, and hard steel alloys.

3.6.5 Chemical Post-processes Chemical etching is the main post-process utilized within this process group; it is not particularly common because it is specialized for specific industries only. It requires low tooling compared to other process groups and low-machinability materials can be effectively processed. Other advantages are that it can process complex geometries and provides a very clean surface at the end of production. However, chemical postprocesses are not used on very rough surfaces because they will not have the desired effect. In such cases, machining takes place in the first stage; then, chemical postprocesses are employed to provide the final finish. Chemical etching is mainly applied

86

3 AM Processes

to alloys that are hard to machine but have a high chemical reactivity, as well as in cases of medical implants; more specifically, chemical etching is employed for the removal of adhered/unmelted powder particles and the fabrication of porous structures and micro-scale features.

3.6.6 Laser-Based Post-processes This post-process family utilizes laser heat as its energy source and has a unique integration advantage: one single laser head/source has both a geometry-related function (ablation) and a structural integrity-related function (peening). In laser ablation, the surface texture is improved in terms of roughness using a laser source. This application is also referred to as laser polishing. Additionally, the dimensional accuracy of the part can be improved by making corrections to the part. Finally, another application of laser ablation is the fabrication of micro-scale features, which is referred to as laser etching [61]. This procedure can be highly optimized in terms of process time through the use of a polygon head; the processing speed is more than ten times higher compared to that of conventional laser heads (Fig. 3.36). In laser peening, the process mechanism that affects the structural integrity of a part is the pressure waves that are created due to plasma formation from laser radiation. The use of a water tamper increases the generated pressure substantially (up to one order of magnitude). Laser peening replaces the tensile stresses created in AM parts during the AM process with compressive residual stresses. Moreover, it partially eliminates voids that are situated close to the part surface and enhances the fatigue lifetime and strength of metallic AM parts. This post-process is best suited when localized part enhancement is required, thus eliminating the need for processing the entire part.

Fig. 3.36 Polygon head for increased speed processing during laser etching

3.7 Special Topics: Hybrid AM and Digitals Twins

87

3.7 Special Topics: Hybrid AM and Digitals Twins 3.7.1 Hybrid AM Post-processing is an integral part of AM, particularly for metallic AM end-use parts that are required to meet specific requirements in their mechanical properties. However, the most common post-processing application is surface finishing because the layer-by-layer nature of AM processes, as well as support and base plate removal, have a negative impact on roughness. Therefore, to minimize the time and, consequently, the cost of AM, the integration of common post-processes into the AM machine itself was considered, thus leading to hybrid AM machines [62]. More specifically, hybrid AM is a combination of an AM process and other non-AM process/processes [63]; essentially, it integrates certain post-processing steps. Two types of hybrid AM can be defined based on whether the different processes are performed sequentially or in-envelope [64]. In the first type, the manufacturing of the part using AM is completed and then the hybrid AM machine starts performing other processes, such as machining or laser-based post-processing. In the second type of hybrid AM, the processing heads for AM and other processes are interchanged or are operated in parallel during the manufacturing of the part. This type of hybrid AM has certain limitations; AM process compatibility cannot be achieved with powder bed-based AM processes. Therefore, mostly powder- or wire-based DED AM is employed. Except for operation sequencing, other important considerations have to be made, including hybrid AM evaluation and process planning for hybrid AM [65] (Figs. 3.37 and 3.38). There is a selection of commercially available hybrid AM machines; however, a custom solution can be developed to be tailored to the specific needs [63]. The use of robotic systems is a common industrial approach for such a solution and offers several advantages, such as the ease of integration of different subtractive and AM process heads, as well as real-time monitoring and process control systems. It is geared toward a fully automated, comprehensive solution (Fig. 3.39).

3.7.2 Digital Twins Digital twins are digital replications that allow the seamless data exchange of data between physical and virtual systems [3]. Digital twins require the integration of process simulation, monitoring, real-time control, and IoT applications [66, 67]; therefore, advanced knowledge and understanding are required, as well as considerable time and cost to develop. However, they provide invaluable advantages in terms of overview, coordination, and optimization, allowing for the breakthrough toward Industry 5.0 [66]. More specifically, they allow for a holistic process overview for all stakeholders through real-time remote monitoring, promoting better team collaboration and financial decision-making [68]. Additionally, risk assessment is

88

3 AM Processes

Fig. 3.37 Hybrid AM evaluation flowchart [3]

Fig. 3.38 Life-to-value product optimization through hybrid AM. Adapted from [65]

enhanced, leading to better anticipation of potential problems. Consequently, this can translate as minimization of unplanned downtime, accidents, as well as maintenance costs through preventive maintenance; all of the aforementioned constitute factors that contribute to improved production time [69]. Finally, they are powerful process optimization tools because they are based on optimized simulations that are continuously updated and trained through real-time data. Figure 3.40 illustrates a schematic representation of the digital twin application for a robotic-based hybrid AM application.

References

89

Fig. 3.39 Customized hybrid AM solution utilizing robotic cell, subtractive, additive process heads, real-time monitoring, and control [63]

Fig. 3.40 Schematic representation of digital twin application for robotic-based hybrid AM application

References 1. P. Foteinopoulos, Computational and empirical modeling of additive manufacturing processes. Dissertation, University of Patras, 2021. https://doi.org/10.12681/eadd/50503

90

3 AM Processes

2. P. Stavropoulos, P. Foteinopoulos, J. Stavridis, H. Bikas, Increasing the industrial uptake of additive manufacturing processes a training framework. Adv. Ind. Manuf. Eng. (2022). Available at SSRN 4169003 3. H. Bikas, S. Koutsoukos, P. Stavropoulos, A decision support method for evaluation and process selection of additive manufacturing. Procedia CIRP 81, 1107–1112 (2019). https://doi.org/10. 1016/j.procir.2019.03.261 4. H. Bikas, A.K. Lianos, P. Stavropoulos, A design framework for additive manufacturing. Int. J. Adv. Manuf. Technol. 103(9), 3769–3783 (2019). https://doi.org/10.1007/s00170-019-036 27-z 5. Q. Meinders, 3D Printing Manual Guide. University of Twente, Rapid Prototyping Laboratory (2017), Available in https://www.utwente.nl/en/rpl/Data%20sheets/Manuals/am-userguide.pdf 3D Printing Manual Guide. Accessed 09 Nov 2022 6. Z. Wu, S.P. Narra, A. Rollett, Exploring the fabrication limits of thin-wall structures in a laser powder bed fusion process. Int. J. Adv. Manuf. Technol. 110(1), 191–207 (2020). https://doi. org/10.1007/s00170-020-05827-4 7. H. Lansdown, Digital Modelmaking: Laser Cutting, 3D Printing and Reverse Engineering (The Crowood Press, Marlborough, 2019) 8. D.G. Ahn, Directed energy deposition (DED) process: state of the art. Int. J. Precis. Eng. Manuf.-Green Technol. 8(2), 703–742 (2021). https://doi.org/10.1007/s40684-020-00302-7 9. P. Foteinopoulos, A. Papacharalampopoulos, P. Stavropoulos, Block-based analytical hierarchy process applied for the evaluation of construction sector additive manufacturing. Procedia CIRP 81, 950–955 (2019). https://doi.org/10.1016/j.procir.2019.03.233 10. G. Chryssolouris, Manufacturing Systems: Theory and Practice (Springer Science Business Media, New York, 2006) 11. P. Foteinopoulos, V. Esnault, G. Komineas, A. Papacharalampopoulos, P. Stavropoulos, Cement-based additive manufacturing: experimental investigation of process quality. Int. J. Adv. Manuf. Technol. 106(11), 4815–4826 (2020). https://doi.org/10.1007/s00170-020-049 78-8 12. D. Obilanade, C. Dordlofva, P. Törlind, Surface roughness considerations in design for additive manufacturing-a literature review. Proc. Des. Soc. 1, 2841–2850 (2021). https://doi.org/10. 1017/pds.2021.545 13. A. Townsend, N. Senin, L. Blunt, R.K. Leach, J.S. Taylor, Surface texture metrology for metal additive manufacturing: a review. Precis. Eng. 46, 34–47 (2016). https://doi.org/10.1016/j.pre cisioneng.2016.06.001 14. J.J. Babu, M. Mehrpouya, T.C. Pijper, G. Willemsen, T. Vaneker, An experimental study of down facing surfaces in selective laser melting. Adv. Eng. Mater. 2101562 (2022). https://doi. org/10.1002/adem.202101562 15. Additive Metal Manufacturing, Design for Additive Metal Manufacturing (2016), https://add itivemet.com/wp-content/uploads/2016/07/design.pdf. Accessed 09 Nov 2022 16. P. Stavropoulos, P. Foteinopoulos, Modelling of additive manufacturing processes: a review and classification. Manuf. Rev. 5, 2 (2018). https://doi.org/10.1051/mfreview/2017014 17. S. Hällgren, L. Pejryd, J. Ekengren, 3D data export for additive manufacturing-improving geometric accuracy. Procedia CIRP 50, 518–523 (2016). https://doi.org/10.1016/j.procir.2016. 05.046 18. A. Sola, A. Nouri, Microstructural porosity in additive manufacturing: the formation and detection of pores in metal parts fabricated by powder bed fusion. J. Adv. Manuf. Proc. 1(3), 10021 (2019). https://doi.org/10.1002/amp2.10021 19. S. Bland, N.T. Aboulkhair, Reducing porosity in additive manufacturing. Met. Powder Rep. 70(2), 79–81 (2015). https://doi.org/10.1016/j.mprp.2015.01.002 20. J.A. Slotwinski, E.J. Garboczi, K.M. Hebenstreit, Porosity measurements and analysis for metal additive manufacturing process control. J. Res. Nat. Inst. Stand. Technol. 119, 494 (2014). https://doi.org/10.6028/jres.119.019 21. W. Yuan, H. Chen, T. Cheng, Q. Wei, Effects of laser scanning speeds on different states of the molten pool during selective laser melting: Simulation and experiment. Mater. Des. 189, 108542 (2020). https://doi.org/10.1016/j.matdes.2020.108542

References

91

22. I.E. Anderson, E.M. White, R. Dehoff, Feedstock powder processing research needs for additive manufacturing development. Curr. Opin. Solid State Mater. Sci. 22(1), 8–15 (2018). https:// doi.org/10.1016/j.cossms.2018.01.002 23. P. Foteinopoulos, A. Papacharalampopoulos, K. Angelopoulos, P. Stavropoulos, Development of a simulation approach for laser powder bed fusion based on scanning strategy selection. Int. J. Adv. Manuf. Technol. 108(9), 3085–3100 (2020). https://doi.org/10.1007/s00170-020-056 03-4 24. J. Jiang, X. Xu, J. Stringer, Support structures for additive manufacturing: a review. J. Manuf. Mater. Proc. 2(4), 64 (2018). https://doi.org/10.3390/jmmp2040064 25. C. Weingarten, D. Buchbinder, N. Pirch, W. Meiners, K. Wissenbach, R. Poprawe, Formation and reduction of hydrogen porosity during selective laser melting of AlSi10Mg. J. Mater. Process. Technol. (2015). https://doi.org/10.1016/j.jmatprotec.2015.02.013 26. Y. Chen, F. Lu, K. Zhang, P. Nie, S.R.E. Hosseini, K. Feng, LZ, Dendritic microstructure and hot cracking of laser additive manufactured Inconel 718 under improved base cooling. J. Alloy. Compd. 670, 312–321 (2016). https://doi.org/10.1016/j.jallcom.2016.01.250 27. Y. Kok, X.P. Tan, P. Wang, M.L.S. Nai, N.H. Loh, E. Liu, S.B. Tor, Anisotropy and heterogeneity of microstructure and mechanical properties in metal additive manufacturing: a critical review. Mater. Des. 139, 565–586 (2018). https://doi.org/10.1016/j.matdes.2017.11.021 28. H. Gonabadi, A. Yadav, S.J. Bull, The effect of processing parameters on the mechanical characteristics of PLA produced by a 3D FFF printer. Int. J. Adv. Manuf. Technol. 111(3), 695–709 (2020). https://doi.org/10.1007/s00170-020-06138-4 29. X. Tan, Y. Kok, W.Q. Toh, Y.J. Tan, M. Descoins, D. Mangelinck, C.K. Chua, Revealing martensitic transformation and α/β interface evolution in electron beam melting three-dimensionalprinted Ti-6Al-4V. Sci. Rep. 6(1), 1–10 (2016). https://doi.org/10.1038/srep26039 30. X. Tan, Y. Kok, Y.J. Tan, G. Vastola, Q.X. Pei, G. Zhang, C.K. Chua, An experimental and simulation study on build thickness dependent microstructure for electron beam melted Ti– 6Al–4V. J. Alloy. Compd. 646, 303–309 (2015). https://doi.org/10.1016/j.jallcom.2015.05.178 31. S.M. Kelly, S.L. Kampe, Microstructural evolution in laser-deposited multilayer Ti-6Al-4V builds: part I. microstructural characterization. Metall. Mater. Trans. A 35(6), 1861–1867 (2004). https://doi.org/10.1007/s11661-004-0094-8 32. X. Tan, Y. Kok, Y.J. Tan, M. Descoins, D. Mangelinck, S.B. Tor, C.K. Chua, Graded microstructure and mechanical properties of additive manufactured Ti–6Al–4V via electron beam melting. Acta Mater. 97, 1–16 (2015). https://doi.org/10.1016/j.actamat.2015.06.036 33. Y. Wang, S. Li, Y. Yu, Y. Xin, X. Zhang, Q. Zhang, S. Wang, Lattice structure design optimization coupling anisotropy and constraints of additive manufacturing. Mater. Des. 196, 109089 (2020). https://doi.org/10.1016/j.matdes.2020.109089 34. K. Kempen, B. Vrancken, S. Buls, L. Thijs, J. Van Humbeeck, J.P. Kruth, Selective laser melting of crack-free high density M2 high speed steel parts by baseplate preheating. J. Manuf. Sci. Eng. 136(6), 061026–061032 (2014). https://doi.org/10.1115/1.4028513 35. P. Stavropoulos, P. Foteinopoulos, A. Papacharalampopoulos, G. Tsoukantas, Warping in SLM additive manufacturing processes: estimation through thermo-mechanical analysis. Int. J. Adv. Manuf. Technol. 104(1), 1571–1580 (2019). https://doi.org/10.1007/s00170-019-04105-2 36. P. Foteinopoulos, A. Papacharalampopoulos, P. Stavropoulos, On thermal modeling of additive manufacturing processes. CIRP J. Manuf. Sci. Technol. 20, 66–83 (2018). https://doi.org/10. 1016/j.cirpj.2017.09.007 37. A. Papacharalampopoulos, H. Bikas, P. Stavropoulos, Path planning for the infill of 3D printed parts utilizing Hilbert curves. Procedia Manuf. 21, 757–764 (2018). https://doi.org/10.1016/j. promfg.2018.02.181 38. M. Lorenzo-Bañuelos, A. Díaz, I.I. Cuesta, Influence of raster orientation on the determination of fracture properties of polypropylene thin components produced by additive manufacturing. Theoret. Appl. Fract. Mech. 107, 102536 (2020). https://doi.org/10.1016/j.tafmec.2020.102536 39. A.R. Torrado, D.A. Roberson, Failure analysis and anisotropy evaluation of 3D-printed tensile test specimens of different geometries and print raster patterns. J. Fail. Anal. Prev. 16(1), 154–164 (2016)

92

3 AM Processes

40. A. Papacharalampopoulos, P. Stavropoulos, H. Bikas, Path planning for filling 3D printed parts utilizing Hilbert curves, in 15th Global Conference on Sustainable Manufacturing, vol. 21 (Haifa, Israel, 2017), pp. 757–764. https://doi.org/10.1016/j.promfg.2018.02.181 41. M. Xia, D. Gu, G. Yu, D. Dai, H. Chen, Q. Shi, Influence of hatch spacing on heat and mass transfer, thermodynamics and laser processability during additive manufacturing of Inconel 718 alloy. Int. J. Mach. Tools Manuf 109, 147–157 (2016). https://doi.org/10.1016/j.ijmach tools.2016.07.010 42. G. Singh, A.M. Waas, V. Sundararaghavan, Understanding defect structures in nanoscale metal additive manufacturing via molecular dynamics. Comput. Mater. Sci. 200, 110807 (2021). https://doi.org/10.1016/j.commatsci.2021.110807 43. P. Stavropoulos, V.C. Panagiotopoulou, Developing a framework for using molecular dynamics in additive manufacturing process modelling. MDPI Model. 3(1), 189–200 (2022). https://doi. org/10.3390/modelling3010013 44. Z.S. Saldi, Marangoni driven free surface flows in liquid weld pools. Dissertation, Delft University of Technology, 2012, https://doi.org/10.4233/uuid:8401374b-9e9c-4d25-86b7-fc4 45ec73d27 45. D. Ding, Z.S. Pan, D. Cuiuri, H. Li, A tool-path generation strategy for wire and arc additive manufacturing. Int. J. Adv. Manuf. Technol. 73(1–4), 173–183 (2014). https://doi.org/10.1007/ s00170-014-5808-5 46. J.P. Kruth, G. Levy, F. Klocke, T.H.C. Childs, Consolidation phenomena in laser and powderbed based layered manufacturing. CIRP Ann. Manuf. Technol. 56(2), 730–759 (2007). https:/ /doi.org/10.1016/j.cirp.2007.10.004 47. J. Irwin, P. Michaleris, A line heat input model for additive manufacturing. J. Manuf. Sci. Eng. 138(11), 111004 (2016). https://doi.org/10.1115/1.4033662 48. X. Liang, D. Hayduke, A.C. To, An enhanced layer lumping method for accelerating simulation of metal components produced by laser powder bed fusion. Addit. Manuf. 39, 101881 (2021). https://doi.org/10.1016/j.addma.2021.101881 49. A. Papacharalampopoulos, P. Stavropoulos, J. Stavridis, Adaptive control of thermal processes: laser welding and additive manufacturing paradigms. Procedia CIRP 67, 233–237 (2018). https://doi.org/10.1016/j.procir.2017.12.205 50. National Academies of Sciences, Engineering, and Medicine, Manufacturing Process Controls for the Industries of the Future (The National Academies Press, Washington, DC, 1998). https:/ /doi.org/10.17226/6258 51. P. Stavropoulos, D. Chantzis, C. Doukas, A. Papacharalampopoulos, G. Chryssolouris, Monitoring and control of manufacturing processes: a review. Procedia CIRP 8, 421–425 (2013). https://doi.org/10.1016/j.procir.2013.06.127 52. A. Papacharalampopoulos, J. Stavridis, P. Stavropoulos, G. Chryssolouris, Cloud-based control of thermal based manufacturing processes. Procedia CIRP 55, 254–259 (2016). https://doi.org/ 10.1016/j.procir.2016.09.036 53. A. Papacharalampopoulos, H. Bikas, C. Michail, P. Stavropoulos, On the generation of validated manufacturing process optimization and control schemes. Procedia CIRP 96, 57–62 (2021). https://doi.org/10.1016/j.procir.2021.01.051 54. P. Stavropoulos, A. Papacharalampopoulos, J. Stavridis, K. Sampatakakis, A three-stage quality diagnosis platform for laser-based manufacturing processes. Int. J. Adv. Manuf. Technol. 110(11), 2991–3003 (2020). https://doi.org/10.1007/s00170-020-05981-9 55. J.M. Waller, B.H. Parker, K.L. Hodges, E.R. Burke, J.L. Walker, Nondestructive evaluation of additive manufacturing state-of-the-discipline report (No. JSC-CN-32323) (2014), https://ntrs. nasa.gov/citations/20140016447. Accessed 09 Nov 2022 56. L. Koester, H. Taheri, T. Bigelow, L. Bond, Nondestructive testing for metal parts fabricated using powder based additive manufacturing. Mater. Eval. 76 (2018), https://ndtlibrary.asnt. org/2018/NondestructiveTestingforMetalPartsFabricatedUsingPowderBasedAdditiveManu facturing. Accessed 09 Nov 2022 57. E. Maleki, S. Bagherifard, M. Bandini, M. Guagliano, Surface post-treatments for metal additive manufacturing: progress, challenges, and opportunities. Addit. Manuf. 37, 101619 (2021). https://doi.org/10.1016/j.addma.2020.101619

References

93

58. N.N. Kumbhar, A.V. Mulay, Post processing methods used to improve surface finish of products which are manufactured by additive manufacturing technologies: a review. J. Instit. Eng. (India): Series C 99(4), 481–487 (2018). https://doi.org/10.1007/s40032-016-0340-z 59. X. Peng, L. Kong, J.Y.H. Fuh, H. Wang, A review of post-processing technologies in additive manufacturing. J. Manuf. Mater. Proc. 5(2), 38 (2021). https://doi.org/10.3390/jmmp5020038 60. T. Lyman, H.E. Boyer, P.M. Unterweiser, Metals Handbook (American Society for Metals, Cleveland, Ohio, 1948) 61. P. Stavropoulos, P. Stournaras, K. Salonitis, G. Chryssolouris, Experimental and theoretical investigation of the ablation mechanisms during femptosecond laser machining. Int. J. Nanomanuf. 6, 55–65 (2010). https://doi.org/10.1504/IJNM.2010.034772 62. J.P.M. Pragana, R.F.V. Sampaio, I.M.F. Bragança, C.M.A. Silva, P.A.F. Martins, Hybrid metal additive manufacturing: a state–of–the-art review. Adv. Ind. Manuf. Eng. 2, 100032 (2021). https://doi.org/10.1016/j.aime.2021.100032 63. P. Stavropoulos, P. Foteinopoulos, A. Papacharalampopoulos, H. Bikas, Addressing the challenges for the industrial application of additive manufacturing: towards a hybrid solution. Int. J. Lightweight Mater. Manuf. 1(3), 157–168 (2018). https://doi.org/10.1016/j.ijlmm.2018.07.002 64. P. Stavropoulos, H. Bikas, O. Avram, Hybrid subtractive–additive manufacturing processes for high value-added metal components. Int. J. Adv. Manuf. Technol. 111, 645–655 (2020). https:/ /doi.org/10.1007/s00170-020-06099-8 65. P. Stavropoulos, H. Bikas, O. Avram, A. Valente, G. Chryssolouris, Hybrid subtractive–additive manufacturing processes for high value-added metal components. Int. J. Adv. Manuf. Technol. 111(3), 645–655 (2020). https://doi.org/10.1007/s00170-020-06099-8 66. G.L. Knapp, T. Mukherjee, J.S. Zuback, H.L. Wei, T.A. Palmer, A. De, T.J.A.M. DebRoy, Building blocks for a digital twin of additive manufacturing. Acta Mater. 135, 390–399 (2017). https://doi.org/10.1016/j.actamat.2017.06.039 67. A. Gaikwad, R. Yavari, M. Montazeri, K. Cole, L. Bian, P. Rao, Toward the digital twin of additive manufacturing: integrating thermal simulations, sensing, and analytics to detect process faults. IISE Trans. 52(11), 1204–1217 (2020). https://doi.org/10.1080/24725854.2019. 1701753 68. P. Stavropoulos, A. Papacharalampoloulos, K. Tzimanis, Design and implementation of a digital twin platform for AM processes. Procedia CIRP 104, 1722–1727 (2021). https://doi. org/10.1016/j.procir.2021.11.290 69. P. Stavropoulos, A. Papacharalampopoulos, C.K. Michail, G. Chryssolouris, Robust additive manufacturing performance through a control oriented digital twin. Metals 11(5), 708 (2021). https://doi.org/10.3390/met11050708

Chapter 4

AM Applications

In this chapter, a more holistic view of Additive Manufacturing (AM) will be presented, integrating all the previous modules in a practical way. Additionally, AM will be presented from a business perspective, focusing on the different phases of product development, encapsulating the entire supply chain of AM equipment, operation, and end-of-life toward a successful business plan. Furthermore, applications of AM in industrial cells and lines will be discussed. Also, specific AM case studies will be considered. Finally, Industry 4.0 capabilities and decentralized manufacturing applications will be presented.

4.1 Introduction to AM Applications In Additive Manufacturing, parts are created in a layer-by-layer fashion by selectively fusing the material of the current layer upon that of the previous one, based on information provided by 3D model data [1]. AM is the successor of rapid prototyping, differing in the fact that it specifically aims to manufacture end-use parts, rather than just prototypes [2]. Its first application took place in 1999 for the creation of a scaffold for a human bladder and in 2006 the first metal AM application was made commercially available in the form of Selective Laser Sintering [3]. The interest in AM has been steadily increasing, leading to its rapid recent growth [4] (Fig. 4.1) and improvement of all its aspects [5, 6], increasing the maturity of AM toward its wider industrial uptake [7, 8] for the production of end-use parts (Fig. 4.2). The maturity of AM in important industrial sectors, namely medical, aerospace, industrial goods, and automotive, is already high (Fig. 4.3), offering specific advantages in each industrial sector Fig. 4.4. Similar is the picture drawn by Gartner

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 P. Stavropoulos, Additive Manufacturing: Design, Processes and Applications, SpringerBriefs in Applied Sciences and Technology, https://doi.org/10.1007/978-3-031-33793-2_4

95

96

4 AM Applications

Fig. 4.1 AM market value is steadily increasing [9]

Fig. 4.2 Percentage of end-use parts manufactured by additive manufacturing [10]

(Fig. 4.5), showing that many AM applications are currently in the slope of enlightenment and a few are already entering the plateau of productivity, highlighting that one of the most important challenges of AM today is the increase of its industrial uptake. AM continues to progress toward becoming a mainstream manufacturing alternative for series production. It eliminates the need for costly tooling, such as molds or dies, and can produce highly complex parts. The production of final parts using AM facilitates small batch sizes up to now, custom parts, prototyping, lightweight structures, complex internal/external features, and the consolidation of many different components into one. As AM improves, it will continue to establish its presence in an increasing number of markets and industries. According to [13], it is too early to know for sure that AM will bring the next industrial revolution; however, there are

4.1 Introduction to AM Applications

97

Fig. 4.3 Maturity stage of AM in different industrial sectors. Adapted from [7]

Fig. 4.4 Main applications trends of AM [11]

encouraging signs that might this is the discussion. AM removes barriers to entering the product development and manufacturing businesses by significantly reducing upfront costs and simplifying or decentralizing supply chains. It is important to highlight that more and more companies continue to increase their use of AM for parts that go into final use and operation. Indicatively, the amount of money spent annually on final part production worldwide is shown in the following graph (Fig. 4.6).

98

4 AM Applications

Metal AM is making important steps toward its adoption as a valid production alternative because it often offers similar or better material properties than parts made with traditional processes, such as casting. Moreover, AM can reduce the amount of material, waste, and weight of parts. It is ideal for high-value, low-volume production of complex parts in several sectors including aerospace, medical/dental, power/energy, tooling, and motorsports. In general, AM is a cost-effective alternative for end-use part production when it adds to the product value in comparison to conventional manufacturing processes. Manufacturers are willing to evaluate a new production process when it is significantly less expensive and/or improves the product’s performance. AM can accomplish the above by influencing several key aspects of product development and manufacturing. Among the rest, AM can eliminate tooling, provide on-demand manufacturing, reduce lead times, inventory, and labor, design lightweight parts, optimized structures, and consolidated parts thus leading to lower part numbers and sustainability (less waste). Moreover, custom and limited-edition components can be produced by AM, incorporate design changes even after the production has started, and facilitate prototyping using the same process and materials. As per the previous, even though AM offers unique advantages and capabilities and its technology maturity level is high enough to allow for extensive industrial application, its wider adoption is hindered by various reasons–barriers. According to the survey conducted in [9], those barriers can be classified into four main categories, namely AM machine cost, part quality issues, limited expertise, and limited material choices (Fig. 4.7).

Fig. 4.5 AM expectations over time [12]

4.1 Introduction to AM Applications

99

Fig. 4.6 Production of AM part from independent AM service providers (in millions of dollars) [13]

Fig. 4.7 Barriers to the wider adoption of AM. Adapted from [9]

The research community, AM machine developers, and larger enterprises are already contributing to overcoming those barriers. More specifically, machine costs are rapidly decreasing, due to increased demand and technology improvement [14]. Additionally, more AM machine options become available, including desktop metal AM machines, which are much more economical and also suitable for Small–Medium Enterprises (SMEs) [15]. Moreover, it has to be considered that initial machine costs account for 45–74% of the total cost to manufacture a part with AM [16], while

100

4 AM Applications

the rest of the costs are significantly lower than conventional manufacturing; for example, tooling costs associated with AM are about 30% that of tooling for injection molding [17]. Regarding product quality improvement, it is the key focus of AM research studies, including simulation [14, 18], and experimental approaches [19], featuring continuous progress, tangible results, and critical improvements [20]. Finally, material choices have been expanded dramatically, including a wide range of plastics, resins, composites, and bio-compatible and cementitious materials. Moreover, a great variety of metal alloys is now available in AM, including Inconel 718 and 625 [21], the machining of which is highly demanding with conventional means [22], as well as the utilization of functionally graded materials [23]. However, the same cannot be stated for the barrier of limited expertise. AM has been characterized as a disruptive technology [24], since it is changing core aspects of manufacturing at multiple levels, including design, product development stages, workforce requirements, supply chain, as well as interaction with clients [2]. This results in rapidly evolving expertise demands for the lucrative industrial utilization of AM, ranging on all levels, from technical to managerial. The same conclusion is drawn from [12] since most of the depicted factors hindering the industrial uptake of AM are related to a lack of expertise (Fig. 4.8). The combination of the previous requirements with the continuous evolution of AM [25] leads the industrial sector to a perpetual lack of the knowledge to assess and understand in depth the AM capabilities which leads to a reluctance for a potential AM venture uptake. This is further corroborated by the lack of qualification and

Fig. 4.8 Factors hindering industrial uptake of AM. In red can be seen the factors connected with lack of expertise. Adapted from [19]

4.2 Production of End-Use AM Parts

101

certification methodologies [26], and its effect is being more prominent in SMEs [27] rather than larger enterprises. However, AM constitutes, in fact, an ideal market opportunity, especially for SMEs [28], since it is not labor intensive, it requires highly trained individuals, and it is particularly flexible, focusing on small volume-lot size productions of high-value parts, the limits of which are set only by the ingenuity and skills of the designer and engineer. The above unique characteristics of AM also render it an important business opportunity for the European manufacturing sector in general, even for countries that lack conventional manufacturing tradition, assets, and know-how, since it plays on the strengths of European economies and societies. Considering that SMEs represent 99% of all businesses in the EU [29], further amplifies the need for consistent and organized training courses on AM.

4.2 Production of End-Use AM Parts As it is reported in [30], sales of AM systems for metal parts declined in 2020. An estimated 2.169 metal AM machines were sold in 2020, a decline of 7% from the 2.333 metal systems sold in 2019. The COVID-19 pandemic contributed to a temporary saturation of the market. The interest in end-use products made with AM remains high, but new workflows and supply chains could slow the adoption. However, significant effort is being put into the development of new metal AM systems that will increase productivity and automation and will contribute to the industrial uptake of the technology. From the literature can be noted that despite the high level of industrialization of Powder Bed Fusion (PBF) machinery, they can be used only for small/medium-sized components with a maximum size of 800 mm in height. Instead, the Directed Energy Deposition (DED) machines can process larger components that can arrive at up to 4000 mm [31]. Another advantage of the DED process with respect to the other metal AM processes is that the substrate could coincide with the surface of an existing component, rendering DED highly suitable for repair operations [13]. Moreover, it is possible to change the material during the deposition process, thus obtaining components that are characterized by different properties in different areas [32]. The graph of Fig. 4.9 illustrates the distribution of metal AM systems sold worldwide and it can be observed that the metal AM market is dominated by the PBF processes with a percentage of 82% [32]. DED is the second technology in terms of sales, with a percentage of only 8% (ten times lower than that of the PBF processes). However, it should be noted that the industrial interest regarding DED processes is growing exponentially, and its application will be studies in the cases presented in this chapter, along with the PBF technology, highlighting their capabilities and future trends [32]. The benefits of AM presented in this book via various mature application cases can justify the use of AM for small batches, as well as for series production; the key to their success and cost-effectiveness is increasing product value. However, challenges related to AM can often contribute negatively if not managed properly. The primary expense of AM is the initial investment cost (machine) and the secondary is the raw

102

4 AM Applications

Fig. 4.9 Distribution of metal additive manufacturing systems in the market in 2019 [32]

material costs. The high investment cost can be justified mainly by the relatively small number of machine sales and the need for machine builders to recuperate development costs. This cost would be much lower if the volume of machines produced and sold was higher. However, the machine depreciation usually spans several years and is divided among all parts it builds over that period. On the other hand, some AM materials are expensive because they are costly to produce and the price of the base elements such as Nickel is influenced by international social and political conditions. The material costs will be reduced when competitive market conditions and economies of bulk production are realized. A key to success in an AM-based production scenario relies on a comprehensive and realistic cost justification taking into account all parameters. A business case based on a simplistic cost comparison between AM and conventional processes is destinated to fail since the range of products for which AM is suited is limited and its strengths lie in the fact that high part complexity does not have an important impact on the product cost. The broader life cycle, the total manufacturing cost, the improvement in product performance, the processability of new materials, and the automation should be considered in a proper AM cost model. In that direction, an elevated production cost for an aerospace component could be compensated and justified if this is designed in a way to be lighter by 25%, resulting in significant savings over the years of operation. Similar cases are possible for improvements in product performance, greater customer satisfaction, reduced product maintenance, and reduction in total manufacturing costs. Another practical application of AM is the production of spare parts toward the reduction of equipment down-time. In several instances, this application alone can justify the slightly higher cost of the initial AM investment. Taking advantage of this capability offered by AM, enterprises that utilize costly machinery equipment in their production plants have started to develop part databases and production workflows to

4.3 Business Perspective and AM Case Studies

103

enable on-demand spare part production through AM. Even though such databases are difficult to create and maintain, as they require significant upfront cost and time to be thoroughly vetted to ensure safety and quality [31], they offer the important advantage of eliminating the possibility of machinery getting out of order due to spare parts’ lead times and ensure that the production schedule will be met.

4.3 Business Perspective and AM Case Studies In this section, three different AM applications will be presented from a business perspective, showcasing the steps required for the development of an effective cost model suitable for the study of whether AM is a cost-effective production alternative using realistic industrial data.

4.3.1 Aerospace Case The aerospace industry was an early adopter of AM. The major players in the sector like NASA, European Space Agency, Space X, and others are already using AM to produce igniters, injectors, and combustion chambers for rocket engines. Aerospace applications can highly benefit from the advantages offered by AM since the lowcost complexity can be utilized for the manufacturing of low-weight products, which leads to a significant increase in performance and efficiency. Additionally, the fact that small lot sizes are required in aerospace mitigates the weakness of AM for low production rates. Moreover, the materials used for the combustion chambers like Inconel alloys and copper alloys are difficult to be processed with conventional technologies, whereas this is not the case for AM technologies that utilize material addition and thermal-based process mechanisms. Another advantage of AM that is relevant for the aerospace sector is the capability for the consolidation of multipart assemblies in one part, thus significantly lowering weight, assembly time and assembly line cost, and time to market [33]. As per the previous, the maturity of AM in the aerospace sector is high, and therefore, a case for this sector has been selected. The scope of this case is the feasibility evaluation of a nozzle part which is the final section of a rocket engine, which is 1000 mm in diameter and 900 mm in height, based on a real case scenario. The first step is the verification of the design and the material (Inconel 718), followed by the selection of the most suitable AM technology. In this case productivity and deposition rate had been set as priorities, followed by the minimization of the powder handling. Therefore, DED has been selected, due to its higher production rate and because it requires at about three times less powder in terms of handling and loading than PBF [33]. However, the minimum thickness of the part’s features was not compatible with the laser beam diameter of the machine’s standard configuration. The machine available

104

4 AM Applications

was equipped with an optical chain that can provide a 2 mm laser spot diameter and a powder nozzle with four convergent powder streams, which created a powder spot of 4 mm [33]. Therefore, a study for the modification of the optical chain system (collimator and laser fiber) in order to obtain a laser spot of 0.6 mm was conducted. This was combined with the acquisition and integration of a new coaxial powder nozzle leading to the reduction of the dimensions of the deposited geometry. This concluded the process selection, which is the second step of the analysis [33]. The next step is the estimation of the printing time and the production steps needed to organize and realize a printing job. Additionally, the definition of the process parameters with the new optical and powder nozzle configuration is a critical aspect to be addressed, to which simulation and design of experiments can significantly contribute. Finally, the last but not least step is the definition of the post-processing activities. • Cleaning of the part for the removal of the remaining powder in channels and cavities [33]. • Measurement of the dimensions using a caliper to confirm that dimensional requirements are met (outer/inner diameter at the base and at the top—thickness of internal ribs and inner/outer walls at the top). • Perform Electrical Discharge Machining (EDM) cutting for the separation of the part from the base plate. • Measurement of the roughness in various surfaces that have functional significance. • Three-dimensional scanning before and after the separation of the part from the baseplate in order to verify possible deformations due to internal thermal stresses. The results demonstrated satisfactory mechanical properties and surface roughness, while the dimensional accuracy both before and after the separation of the part from the baseplate was within the specified requirements [33]. The latter indicates that heat treatment is not necessary since no deformations were observed. This can be attributed to the large mass of the part, its specific geometry, and stiffness which contributed to a relatively uniform thermal field distribution and minimization of thermal stresses, resulting in acceptable dimensional precision. However, it was deemed necessary to add material in some zones due to the lower deposition that occurred locally [33]. In conclusion, the application developed and presented in this section aims to demonstrate the maturity level of DED process in producing components of large dimensions, with particular and thin features that are difficult to manufacture in one step using conventional manufacturing technologies [33]. Moreover, the flexibility offered by AM, from a machinery point of view, for this complex component could not be achieved through conventional manufacturing processes [33].

4.3 Business Perspective and AM Case Studies Table 4.1 Estimated printing time of a rocket’s nozzle [33]

105

Deposition/production time

AM–DED

Time (h)

1800

Profitability Analysis Following the above technical description of the case, the profitability analysis follows. As stated previously, the component has an external diameter of 1000 mm and a height of 900 mm. The volume of the nozzle is approximately 20,850 cm3 , and using the same print setup programming approach as the one described in Sect. 4.2, the total printing time was calculated and is given in Table 4.1. In order to calculate the final production cost of this aerospace nozzle, the following data have to be obtained and considered. • • • • • •

AM machinery depreciation cost. Gas costs. Energy costs. Raw material costs. Machinery setup costs. Post-processing costs.

Regarding the depreciation cost (below mentioned as hourly cost), it is at the manufacturer’s discretion to decide whether to include this cost within the process cost analysis or in a different logistic manner. The reasoning behind this decision depends on a company’s available cash flow and the desired time frame in which the manufacturer wants to have a return on the investment [33]. For the case presented in this book, the first option has been selected, since it is the most common approach and the relevant data can be seen in Table 4.2. Moreover, the hours available were estimated based on the production objective of the lot size of the nozzles within a year. Normally depreciation period is considered to 5 years. Thus, the final hourly cost is calculated by the following equation and its value for this case is provided below [33]. hourly cost =

Table 4.2 Machine data for calculation of machine hour cost [33]

machinery cost . hours available × depriciation duration

Production data

AM–DED

Machinery cost (e)

1,000,000

Hours available (h/year)

4928

Depreciation duration (years)

5

Hourly cost (e/h)

40.5

(4.1)

106

4 AM Applications

Table 4.3 Total gas cost for the nozzle application [33]

Production data

AM–DED

Gas flow rate (L/min)

40

Volume rate (m3 /h) ( / ) Gas cost e m3

2.4

Total gas cost (e)

12,960

3

Regarding the costs related to the shielding and the powder carrier gas, it has been chosen to consider Nitrogen for technical reasons. This cost can be simply derived using the price per cubic meter of gas and its volume flow rate, as described in the equation below. The results are summarized in Table 4.3. total gas cost = volume rate × process duration × gas cost

(4.2)

For the calculation of energy costs, an important clarification has to be made. It has to be highlighted that energy cost is a very labile value since it depends on socio-political and geographic conditions. The energy costs used in this case study are derived from Eurostat sources for the year 2022, using the Italian national average [34]. The second important consideration is that even though laser power was maintained constant in the production plan of the specific application, the laser source and the machinery could be affected of power peaks. To maintain a conservative approach in the energy cost calculations, the power consumption of the machinery has been considered at its maximum peak for the whole processing time [33]. As per the above, the total energy consumption was calculated, using the below equation. E con = Ppeak Pt

(4.3)

where E con is the energy consumption, Ppeak is the maximum achievable power value of the machine, and Pt is the processing time. The next step is the calculation of the total energy cost related to the process utilizing the following equation [33]. ECtot = E con EC,

(4.4)

where ECtot t is the total energy costs, E con , and EC is the energy cost is the energy consumption. In Table 4.4, all the data and results related to the energy costs are consolidated. As mentioned above, the worst-case scenario for the energy cost was considered, and therefore, the overall production cost could be further reduced if a less conservative approach is followed [33]. For the material cost calculations, the market price of Inconel 718 was considered. However, noting the price trend of Nickel, many fluctuations can be observed, which are attributed to social-economic reasons. It is important to highlight that in such cases, agreements between machine builders and/or companies and AM powder suppliers should be made to ensure price stability, given the high demand for raw material required for such a production plan. For this case, the raw material price

4.3 Business Perspective and AM Case Studies Table 4.4 Total energy costs for the nozzle production [33]

Table 4.5 Total material costs for the production of the nozzle [33]

107

Production data

AM–DED

Energy cost (e/kWh)

0.25

Machine power peak (kW)

35

Process duration (h)

1800

Total energy consumption (kWh)

63,000

Total energy cost (e)

15,750

Production data

AM–DED

Material cost (powder) (e/kg)

76

Deposition or processed mass (kg)

0.171

Material use eff. (%)

90

Total material cost (e)

14,296

is considered fixed at 76 e/kg [33]. Moreover, an efficiency ratio should be used to take into account material waste, as per the machine manufacturer guidelines and the specific geometry. In Table 4.5, the outcome of the calculations for the material usage and cost for one nozzle is presented. Having calculated all the above, the total cost per part can now be obtained by summing up the individual costs. In this estimation, a post-process cost was also added since the full-scale nozzle has to be removed from the plate using EDM. Few EDM machines can be used due to the big dimensions of the particular component [33]. Therefore, the use of a saw machine is studied as an alternative. However, for this case, extra material has to be added to the bottom of the part to compensate for the lower precision of such machines. Moreover, a short milling operation for the removal of the extra material and the finishing of the external surfaces was also considered [33]. Therefore, the final cost per part for the company would be the following one (Table 4.6). For the calculation of the annual income from the sales of a satellite nozzle component, it is necessary to consider two facts. The first is the annual production volume, which has already been estimated above and the second is the market price at which Table 4.6 Final production cost for the aerospace nozzle [33]

Production data

AM–DED

Total gas cost (e)

12,960

Total energy cost (e)

15,750

Total material cost (e)

14,296

Total process cost (e)

72,900

Total post-process cost (e)

5000

Final cost (e)/per part

120,906

108

4 AM Applications

the company will distribute its products. The evaluation of the profitability of the investment of one AM machine takes place, assuming that for the first years the production rate will not increase and the analysis takes into account the whole duration of the depreciation of the machine [33]. If it is necessary to increase the initial production plan, more machines have to be acquired and this analysis has to be updated. For this book, the use of one machine and fixed yearly production, except for the first year, are considered. As per the previous calculations, the cost per part for a new manufacturer is 120,906 e, which allows for a considerable operating profit margin for each part. For the definition of the market price of the product, it is very important to clarify the market status and the competitive advantages of the specific AM product over the conventionally manufactured ones. Taking advantage of the specific benefits of AM, the redesigned nozzle can be characterized as very high quality, as well as improved efficiency when compared to competitor products [33]. The part of this case was meant for the European Space Agency (ESA), and a market price estimation as per the specific standing agreements is around 350,000 e [35]. A detailed breakdown of the gross profit generated by the rocket nozzle can be seen in Table 4.7. The total revenue is calculated by multiplying the price per unit by the number of units sold. Due to the machine installation, the training of the personnel, and the characterization of the component that will take place during the first year, only one nozzle will be printed. This is the reason why the first year’s sales are only 8% of the sales of any consequent year. The total number of units sold multiplied by the cost per unit (as shown in Table 4.7) equals the cost of goods sold which can be defined as the costs directly linked to the production of the goods sold in a company. The gross profit of the company equals the total revenue minus the total costs of goods sold [33]. It has to be mentioned that further analysis assuming the operating expenses, depreciation, and amortization before interest and taxes would be more complex, as well as detailed data are not available. Therefore, using the previously presented data and considerations, at least 200,000 e have to be returned each year to cover the depreciation of the investment (machinery). There will probably be a loss during the Table 4.7 Gross profit predictions in a 5-year-long period [33] Profitability data

Year 1

Year 2–5

Total number of parts produced in the machine per year

1

12

Number of units sold per year

1

12

Price per unit (e)

350,000

350,000

Cost per unit (e)

−120,906

−120,906

Revenues Nozzle (e)

350,000

4,200,000

Costs of goods sold

−120,906

1,450,872

Gross profit

229,094

2,749,128

Gross profit (%)

65.4

65.4

4.3 Business Perspective and AM Case Studies

109

first year; however, starting from the second year, the investment can be gradually covered, including the rest of the expenses and taxes, considering a gross profit of 687,282 e per year [33]. As per the presented analysis, it can be concluded that an AM investment for an aerospace application can be very profitable, given that the required considerations and programming take place. Additionally, the utilization of AM instead of conventional processes led to improved product performance, offering technical advantages, and fewer production steps, eliminating various operating, maintenance, and consumables costs. Since AM technology becomes more mainstream, offering increasing production rates and lower investment costs, the profit margin will steadily increase, constituting AM a viable alternative for other sub-assemblies of the rocket and aerospace industry [33].

4.3.2 Dental Case The maturity level of a dental AM application is demonstrated in this case, including the main process, as well as production, and cost aspects. The main two materials used in this sector are Titanium alloys or CoCr alloys (Nickel free). For this particular application, a CoCr alloy was selected, in order to fulfill the end-user requirements regarding the dimensional accuracy of the components before and after postprocessing, as well as the surface roughness and the ductility of the material that will permit the hand finishing of the surfaces. Once the technical and economical requirements of the end-user have been defined, the approach of such an application study contains should follow the below steps [33]. (1) (2) (3) (4) (5) (6) (7)

Feasibility analysis—printing time and cost estimation. Evaluation of the number and the kind of parts to be printed. Material characterization and process parameters definition. Printing job realization. Post-processing activities. Quality evaluation. Further elaboration on the parts made by the end-user.

Once the technical feasibility has been demonstrated, as well as the coating process, it was concluded that AM is a valid manufacturing method alternative and can eliminate several manual steps of the current production process. To prove the economical sustainability and profitability of such a machinery investment for a service provider company, calculations similar to the previous case (Sect. 4.3.1) have to take place, utilizing the data provided in Table 4.8. To maintain simplicity, only one type of geometry was selected, the dentures. Based on the input of the end-user, the production volume is 1200 dentures per month or 14,400 parts per year [33]. In this case, the PBF process was selected due to the higher mechanical properties, better roughness, and accuracy it offers, since it can meet the aforementioned production requirements. Therefore, based on the job simulation performed above,

110

4 AM Applications

40 dentures could be printed on one plate. Considering the plate dimension (150 mm) and the process parameters extracted from the case analysis, the job duration is 4 h and 30 min (Table 4.9). The total volume of these 40 dentures is 24 cm3 and the total height with the supports is 12.5 mm [33]. As per the previous case, the following data have to be calculated for the cost evaluation. • • • • • •

Depreciation cost of the machinery. Gas costs. Energy costs. Material costs. Machinery setup costs. Post-processing costs.

The hourly cost was calculated in the same way that was described in the previous case of the aerospace industry case (Sect. 4.3.1). However, in this case, the PBF process was selected, and therefore, some of the data are different. The investment cost for such a piece of machinery is 260,000 e. Furthermore, the printing hours per year are set to 4800, considering continuous printing for 24 h, 5 days per week, one shift, and 10 months of operation. This was estimated by including annual leaves, maintenance stops, and load/unloading activities. The depreciation period, in this case, was considered as 5 years as well [33]. In Table 4.10, the data and the hourly costs can be found, utilizing the same formula presented in the previous case (Sect. 4.3.1). The gas that was selected for this application is Nitrogen. Initially, in order to fill the working chamber with inert gas and create a protected atmosphere, the machine consumes 20 L/min, but after forty minutes, the consumption is 7 L/min, which is sufficient to maintain the required gas level. The gas consumption calculation results, as per the previous data, can be seen in Table 4.11. The energy consumption and cost were also estimated as per the procedure followed in the previous case (Sect. 4.3.1). The total electrical power consumption of the selected machine is 8 kW, which is much less than the DED machine of the previous case. The worst-case scenario was considered for the calculations, Table 4.8 Technical details of the dental printing job [33] Part description

Dental printing job

Technology

Powder bed fusion

Material

CoCr

Number of parts per job

40 (conservative approach)

Table 4.9 Printing time of the dentures [33] Deposition/production time

AM–PBF

Time (h)

4.5

4.3 Business Perspective and AM Case Studies

111

Table 4.10 Machine data for calculation of the machine hour cost [33] Production data

AM–PBF

Machinery cost (e)

260,000

Hours available (h/year)

4800

Depreciation duration (year)

5

Hourly cost (e/h)

10.8

Table 4.11 Total gas cost for the printing of the dentures [33]

Production data

AM–PBF

Gas flow rate (L/min)

14

Volume rate (m3 /h) ( ) Gas cost e/m3

0.84

Total gas cost (e)

11.34

3

assuming that maximum power is consumed during the whole machine operation. The results are demonstrated in Table 4.12. The market price of the CoCr alloy of 65 e/kg was considered for the following calculations. Due to increased competition between the powder suppliers, the increased production capacity, and the unstable social-economic background, it is always preferred that specific agreements are made on the cost per kilo of the powder materials, to ensure a constant price. In the PBF technology, the majority of the powder used in a printing job can be sieved, recycled, and re-utilized. Therefore, in this case, the material use coefficient is even higher than that of the DED process (Table 4.13). Table 4.12 Total energy costs for the dentures printing job [33]

Table 4.13 Total material costs for the dentures printing job [33]

Production data

AM–PBF

Energy cost (e/kWh)

0.25

Machine power peak (kW)

8

Process duration (h)

4.5

Total energy consumption (kWh)

36

Total energy cost (e)

9

Production data

AM–PBF

Material cost (powder) (e/kg)

65

Deposition or processed mass (kg)

0.3

Material use eff. (%)

98

Total material cost (e)

19.89

112 Table 4.14 Final production cost of dentures [33]

4 AM Applications

Production data

AM–PBF

Total gas cost (e)

11.34

Total energy cost (e)

9

Total material cost (e)

19.89

Total process cost (e)

48.75

Total post-process cost (e)

50

Final cost (e) per job (40 dentures)

138.98

Final cost (e) per part

3.47

Having extracted all the necessary individual costs, the total cost per part can now be calculated. It has to be highlighted that supporting cones have been used for this printing job, which are placed under the dentures. This is necessary in order to support the exposure surfaces which are parallel to the plate. Additionally, it allows for easier removal of the dentures from the plate, avoiding the use of EDM and further reducing costs. However, in this case, a heat treatment is advised to eliminate thermal stresses and deformations when the dentures are removed from the plate, while simultaneously improving ductility. This will also facilitate the manual finishing of the dentures before the ceramization [33]. The cost of the oven required for the thermal post-process is included in the sale price of the PBF machine, and therefore, no extra costs, except for the operating costs, have to be considered. The total cost per job has to be calculated and reduced to one part (Table 4.14). It is necessary to consider two facts for the calculation of the annual income from the sales of a single denture component. The first is the annual production volume, which has already been estimated, and the second is the price per part, namely the market price at which the company will sell its products. For the evaluation of the investment profitability of one PBF machine, it is considered that for the first year, the production volume will be constant at 14,400 dentures, as per the indication of the end-user. For the next two years, an increase of 15% in annual production is foreseen. This leads to 16,560 dentures per year or 1380 per month. Dividing the last with the 20 working days per month leads to 69 dentures per day. Considering two printing jobs of 40 dentures, the increased production still is within the production capacity of the machine [33]. For the duration of the last two years, a further increase of 15% in production is considered. This results in almost 80 dentures per day, which still is within the production capabilities of the machine since the end-user can produce this demand with two printing jobs in one day [33]. As per the feedback of the end-user, the market price of the denture is 7 e each. Utilizing this value, the gross profit calculations over a 5-year-long depreciation period have to take place to indicate the level of investment absorption. Additionally, the alternatives of acquiring another more productive machine or further optimizing the process in the future have to be studied. By multiplying the total number of product units sold by the cost per unit, the total cost of goods is calculated, as shown in Table 4.15. The gross profit of the company is equal to the total revenue minus

4.3 Business Perspective and AM Case Studies

113

the total costs of goods sold. The total number of units sold has to be multiplied by the cost per unit equals the cost of goods sold [33]. The gross profit of the company equals the total revenue minus the total costs of goods sold. The results of those calculations can be seen in Table 4.15. In the calculations of the gross profit, the operating expenses such as staff salaries, consumables, maintenance, and assets are not deducted for the sake of simplicity since these data are not easily obtainable. Therefore, assuming that 50,000 e must be returned every year for the depreciation of the investment (machinery), a loss will probably be noticed for the first year. However, starting from the second year, both the investment, as well as the rest of the expenses and taxes, will be covered, considering a gross profit of around 70,000 e per year [33]. In a nutshell, after the profitability analysis of the dental sector application, it is evident that this is a field where AM can replace the current manufacturing method and provide advantages not only at an economic level but also technical, offering a higher degree of customization, flexibility, and the possibility to investigate new material alloys compatible with the human body [33]. As per the presented analysis, it can be concluded that an AM investment for a dental application can be very profitable, given that the required considerations and programming take place. Additionally, the utilization of AM instead of conventional processes led to improved product performance, offering technical advantages, a higher degree of customization, and flexibility. Finally, the presented setup also offers the capability for future investigation of new, innovative material alloys, compatible with the human body, which could not be used cost-effectively by conventional manufacturing technologies [33]. Considering the constant increase of demand in this sector, the investment in such a machine can be easily absorbed in a few years. In case of increased production rate demands, modifications and more advanced equipment can serve this need, like Table 4.15 Gross profit predictions for dental application in a 5-year-long period [33] Profitability data

Year 1

Year 2

Year 3

Year 4

Year 5

Total number of parts produced in the machine 14,400 per year

16,560

16,560

19,044

19,044

Number of units sold per year

14,400

16,560

16,560

19,044

19,044

Price per unit (e)

7

7

7

7

7

Cost per unit (e)

3.47

3.47

3.47

3.47

3.47

Revenues dentures (e)

100,800 115,920 115,920 133,308 133,308

Costs of goods sold

49,968

57,463

57,463

66,082

66,082

Gross profit

50,832

58,457

58,457

67,226

67,226

Gross profit%

50.4

50.4

50.4

50.4

50.4

114

4 AM Applications

the addition of additional laser sources that can operate in parallel, increased buildenvelope, and build-platform, which point out the flexibility of this AM solution. Therefore, it can be concluded that AM can fully replace the current production approach, creating fruitful market opportunities both for dental enterprises, as well as AM machine manufacturers and AM service providers [33].

4.3.3 Power and Energy Sector Case The third case will demonstrate the maturity level for the industrial uptake of AM in the power and energy sector. The typical machinery used in the industrial sector includes setups for the generation of power from natural resources, including oil, gas, wind, solar, and other sources, as well as energy transfer. AM applications in this sector are rapidly increasing over the past years, both regarding PBF and DED technologies. Typical parts made using AM for the power and energy sector include turbine blades, motor parts, and stators [36]. It has to be pointed out that important technical and economical advantages come from the repair of these components when AM is used, the turbine blades in particular. The presented case was developed by an industrial machine manufacturer in collaboration with a company oriented toward the production and repair of turbine blades for the energy sector. The specifications provided by the end-user regarded a first-stage turbine blade like the one illustrated in Fig. 4.10. The representation of the exact geometry was not possible due to confidentiality reasons. The blades are worn on the tip, which has to be milled down in order to be reconstructed. The material that was selected is Inconel 625, which is a Nickel-based alloy that is very similar to the one used for the production of the blade [33]. Since this is a repair case, the DED technology is the standard and best-suited AM technology, due to the unique advantages it offers, in terms of reachability and exact powder deposition without the need to encompass the whole product in Fig. 4.10 Turbine blade design used in this case study

4.3 Business Perspective and AM Case Studies

115

Fig. 4.11 Mock-up for repair trials

powder, like in the PBF process. A mock-up of the repair trials can be seen in Fig. 4.11. Regarding the printing, the deposition should follow the geometry of the tip, using an over-deposited tolerance of +0.3/+0.5 mm both internally and externally [33]. This will allow the milling and finishing post-processing operations to take place, ensuring that the exact design requirements will be met (total height of 3 mm). The finishing and thermal treatment, as well as the evaluation of the internal defects, will be performed by the end-user in this case, as indicated in specific quality standards [33]. The first step of the evaluation procedure is the definition of the process parameters, while another critical aspect is the automation of the whole process, which was specifically requested by the end-user. More specifically, each blade under repair has a slightly different height, which leads to a slightly different section in each of them [33]. The standard procedure for the identification of the surface border was the use of a probe, which requires constant manual interference, and it is a relatively slow method. Therefore, in order to achieve high production rates and achieve full automation of the process, a system was developed capable of optically recognizing the border of each component under repair, utilizing a vision system, and adjusting the printing parameters accordingly [33]. This led to the reduction of the repair process to 15 min instead of the initial 50 min of the probe-supported process. Details regarding the productivity evaluation of this repair case are provided in Table 4.16. Profitability Analysis For the evaluation of the profitability of this repair case, the use of various sensors and automation was required. The end-user normally repairs more than 2000 blades Table 4.16 Performance evaluation between the current repair process and DED repair [33] Performance indicators

Current manufacturing method

DED repairing

Production time

50 min

15 min

Parts per day (1 shift)

8 parts

24 parts

Parts per year

1920

5760

116 Table 4.17 Total repair time for one turbine blade [33]

4 AM Applications

Deposition/production time

AM–DED

Time (min)

15

per year, and they aimed to increase the production capacity. Utilizing the aforementioned setup, the total production was reduced to 15 min from the initial 50 min, and therefore, the scenario analyzed herein demonstrates the profits from the full exploitation of the presented AM production alternative [33]. The volume that is required to be deposited for this type of blade is 6 cm3 and the production time can be seen in Table 4.17. As in the previous cases, the following data need to be calculated for the cost evaluation. • • • • • •

Depreciation cost of the machinery. Gas costs. Energy costs. Material costs. Machinery setup costs. Post-processing costs.

The hourly cost for this case was calculated using a similar procedure as in the previous ones. The particular DED machine, however, has a higher price (around 800,000 e), as it has a significantly larger printing envelope (1100 × 800 × 600 mm). In this case, only one shift per day has been considered since the machine cannot operate unsupervised during the night because the total process cycle is only 15 min. Therefore, the total available hours, in this case, are 1920 h per year. The depreciation period was considered as 5 years [33]. In Table 4.18, the data and the hourly cost are provided, utilizing the same formula presented in the previous cases. Argon was used for this application due to technical requirements. More specifically, due to the relatively low density of Argon, its capability to remove any possible fumes generated during the process is improved, in this way contributing to maintaining the melt-pool temperature more uniform over time [33]. The gas consumption, in this case, is 60 L/min and the price of Argon is 6 e/m3 . Therefore, considering a medium consumption, the following cost was calculated (Table 4.19). In a similar way to the previous cases, the energy consumption and cost were also estimated. The total electrical power of the DED machine is 35 kW. The worst-case Table 4.18 Machine data for the calculation of the machine hour cost [33]

Production data

AM–DED

Machinery cost (e)

800,000

Hours available (h/year)

1920

Depreciation duration (years)

5

Hourly cost (e/h)

83

4.3 Business Perspective and AM Case Studies Table 4.19 Total gas cost for the repair of the turbine blade [33]

Table 4.20 Total energy cost of the repair case [33]

117

Production data

AM–DED

Gas flow rate (L/min)

60

Volume rate (m3 /h) ( / ) Gas cost e m3

3.6

Total gas cost (e)

5.4

6

Production data

AM–DED

Energy cost (e/kWh)

0.25

Machine power peak (kW)

35

Process duration (h)

0.25

Total energy consumption (kWh)

8.75

Total energy cost (e)

2.2

scenario was considered in this case as well, assuming maximum energy consumption during the whole production time [33]. The results can be seen in Table 4.20. The market price of the Inconel 625 alloy was considered for the following calculations, which is 82 e/kg. As mentioned for the Inconel 718, the price of the alloy used has a lot of variations due to issues regarding Nickel production, which have a direct impact on the cost of this AM powder. Even though in the DED technology, a significant portion of the unused powder can be recycled and re-utilized, the material usage coefficient for the DED process is lower than that of PBF [33]. However, due to the small mass of the parts of this particular case, the material cost remains relatively low (Table 4.21). Following the above individual cost estimations, the total final cost per repair can now be calculated. Except for the standard costs of gas, electricity, material, and depreciation cost, an amount was added for post-process activities. This includes a short finishing of the external surface of the tip so as to obtain the final dimensions of the requirement. Furthermore, the blade has to be heat treated to increase its hardness and mechanical properties. Finally, it must also be prepared for shipment to the customer. These activities are performed internally; therefore, an estimated amount was added to provide a complete view of the costs (Table 4.22). For the calculation of the annual income of the sales of a single component, it is necessary to consider two facts. The first is the annual production volume, which Table 4.21 Total material cost for the repair [33]

Production data

AM–DED

Material cost (powder) (e/kg)

82

Deposition or processed mass (kg)

0.50

Material use eff. (%)

55

Total material cost (e)

4.5

118 Table 4.22 Final production cost of the turbine blade’s repair process [33]

4 AM Applications

Production data

AM–DED

Total gas cost (e)

5.4

Total energy cost (e)

2.2

Total material cost (e)

4.5

Total process cost (e)

20.75

Total post-process cost (e)

200

Final cost (e)/per part

232.85

is already estimated above, and the second is the price per part, the market price at which the company will distribute its products. Evaluating the profitability capacity of such an investment, it was considered that the maximum number of repairs per year is 5760, based on the 1920 h available [33]. It was stated by the end-user that their plan is to start with an initial number of 2000 repairs annually and increase this number yearly, to reach 5760 annual repairs in the fifth year. Regarding the selling price, there is a high-profit margin since the market price of a new blade is around 15,000 e. Assuming a conservative approach, as per the feedback of the end-user, the selling price for a repaired blade was set at 350 e. The cost of goods sold is calculated by multiplying the cost per unit by the total number of units sold. The gross profit of the company is equal to the total revenue minus the total costs of goods sold. The gross profit prediction for the repair of the turbine blades is shown in Table 4.23. In the calculations of the gross profit, the operating expenses such as staff salaries, consumables, maintenance, and assets are not deducted, mainly because these data are not easily obtainable. Therefore, assuming that 160,000 e must be returned every year for the depreciation of the investment (machinery), a loss will probably Table 4.23 Gross profit prediction for the repair of the turbine blades [33] Profitability data

Year 2

Year 3

Year 4

Year 5

Total number of parts produced in the 2000 machine per year

Year 1

2800

3500

4600

5760

Number of units sold per year

2000

2800

3500

4600

5760

Price per unit (e)

350

350

350

350

350

Cost per unit (e)

232.85

232.85

232.85

232.85

232.85

Revenues blade (e)

700,000 980,000 1,225,000 1,610,000 2,016,000

Costs of goods sold

465,700 651,980 814,975

1,071,110 134,126

Gross profit

234,300 328,020 410,025

538,890

674,784

Gross profit (%)

33.4

33.4

33.4

33.4

33.4

References

119

be noticed for the first year [33]. However, starting from the second year, both the investment, as well as the rest of the expenses and taxes, will be covered, considering a gross profit of around half a million Euros per year. As per the previous data and calculations, the competitiveness of AM in the energy industrial sector is validated, given that the right application for AM is identified, characterized, and developed. Additionally, it has to be pointed out that performing repairs using AM can offer economic advantages for any sector, since the initial machinery investment can be absorbed efficiently and fast, offering significant profit return [33]. Moreover, it was demonstrated that production time reduction can be achieved with AM in comparison to conventional processes, given that the proper sensors and automation are used. Concluding, the AM repair application from the power and energy industry has demonstrated the technical feasibility and advantages of DED technology. Furthermore, it has to be highlighted that the integration of tailor-made automation and sensors can be the turning point that enables the cost-effective and high productivity utilization of AM [33]. Consequently, important innovations from a machinery point of view that can facilitate other similar applications in other sectors as well can lead to the improvement of the requirements and specifications of the next generation of industrial AM machines, allowing for larger lot size productions, paving the way for AM as a cost-effective repairing manufacturing alternative for every factory [33].

References 1. P. Stavropoulos, P. Foteinopoulos, Modelling of additive manufacturing processes: a review and classification. Manuf. Rev. 5, 2 (2018). https://doi.org/10.1051/mfreview/2017014 2. N. Hopkinson, R. Hague, P. Dickens, Rapid Manufacturing: An Industrial Revolution for the Digital Age (John Wiley & Sons, 2006). ISBN-13: 978-0-470-01613-8 3. P. Foteinopoulos, Computational and empirical modeling of additive manufacturing processes. Dissertation, University of Patras, 2021. http://hdl.handle.net/10442/hedi/50503 4. G.E. Additive, How to Create a Comprehensive Business Case (2019), https://www.ge.com/ additive/roi-playbook-form-a. Accessed 09 Nov 2022 5. M. Pérez, D. Carou, E.M. Rubio, R. Teti, Current advances in additive manufacturing. Procedia CIRP 88, 439–444 (2020). https://doi.org/10.1016/j.procir.2020.05.076 6. AMFG, The Additive Manufacturing Landscape (2020), https://amfg.ai/whitepapers/the-add itive-manufacturing-landscape-2020-report/#. Accessed 09 Nov 2022 7. 3D Natives, How are different industries adopting additive manufacturing? (2019), Source: https://www.3dnatives.com/en/industries-additive-manufacturing-120820195/. Accessed 09 Nov 2022 8. A.K. Lianos, S. Koutsoukos, H. Bikas, P. Stavropoulos, Manufacturability assessment and design for AM. Procedia CIRP 91, 290–294 (2020). https://doi.org/10.1016/j.procir.2020. 02.178 9. HUBS, Additive manufacturing trend report (2021), https://f.hubspotusercontent10.net/hubfs/ 4075618/Additive%20manufacturing%20trend%20report%202021.pdf?utm_campaign= Gated%20Content%20Downloads&utm_medium=email&_hsmi=82605589&_hsenc=p2A Nqtz-9pM5X9Hu5jL_49jLKLdBARaGb5PVPhM3kYwzk2c36z2bmdpMZRLJiSd4n5ZK SZOJoCbPnSwFXfi64pLJvjvFjceDPf3w&utm_content=82605589&utm_source=hs_automa tion, Accessed 09 Nov 2022

120

4 AM Applications

10. Jabil, 3D printing technology trends (2021), https://www.jabil.com/dam/jcr:82f12c7a-747542a0-a64f-0f4a625587d8/jabil-2021-3d-printing-tech-trends-report.pdf. Accessed 09 Nov 2022 11. Prima Additive, Metal AM (2022), https://www.primaadditive.com/en. Accessed 09 Nov 2022 12. Gartner, Gartner hype cycle (2019), https://www.3dnatives.com/en/gartner-hype-cycle-3dprin tingpredictions-150120194/ 13. Wohlers Associates, Wohlers Report 2021: 3D Printing and Additive Manufacturing State of the Industry (Wohlers Associates Inc., Fort Collins, Colorado, 2021) 14. P. Foteinopoulos, A. Papacharalampopoulos, K. Angelopoulos, P. Stavropoulos, Development of a simulation approach for laser powder bed fusion based on scanning strategy selection. Int. J. Adv. Manuf. Technol. 108(9), 3085–3100 (2020). https://doi.org/10.1007/s00170-02005603-4Accessed09Nov2022 15. AMFG, The additive manufacturing landscape (2020), https://amfg.ai/whitepapers/the-add itive-manufacturing-landscape-2020-report/#. Accessed 09 Nov 2022 16. D.S. Thomas, S.W. Gilbert, Costs and cost effectiveness of additive manufacturing. NIST Spec. Publ. 1176, 12 (2014). https://doi.org/10.6028/NIST.SP.1176 17. Deloitte Review, 3D opportunity for production (2014), https://www2.deloitte.com/content/ dam/insights/us/articles/additive-manufacturing-business-case/DR15_3D_Opportunity_For_ Production.pdf, Accessed 09 Nov 2022 18. P. Foteinopoulos, A. Papacharalampopoulos, P. Stavropoulos, On thermal modeling of additive manufacturing processes. CIRP J. Manuf. Sci. Technol. 20, 66–83 (2018). https://doi.org/10. 1016/j.cirpj.2017.09.007 19. A.K. Lianos, H. Bikas, P. Stavropoulos, A shape optimization method for part design derived from the buildability restrictions of the directed energy deposition additive manufacturing process. Designs 4(3) (2020). https://doi.org/10.3390/designs4030019 20. D.R. Eyers, A.T. Potter, Industrial additive manufacturing: a manufacturing systems perspective. Comput. Ind. 92, 208–218 (2017). https://doi.org/10.1016/j.compind.2017.08.002 21. Prima Additive, Nickel alloys for AM (2022), https://www.primaadditive.com/en/materials/dir ect-energy-deposition/nickel-alloys-ded. Accessed 09 Nov 2022 22. D. Bourell, J.P. Kruth, M. Leu, G. Levy, D. Rosen, A.M. Beese, A. Clare, Materials for additive manufacturing. CIRP Ann. 66(2), 659–681 (2017). https://doi.org/10.1016/j.cirp.2017.05.009 23. N. Li, S. Huang, G. Zhang, R. Qin, W. Liu, H. Xiong, J. Blackburn, Progress in additive manufacturing on new materials: a review. J. Mater. Sci. Technol. 35(2), 242–269 (2019). https://doi.org/10.1016/j.jmst.2018.09.002 24. S. Shah, S. Mattiuzza, Adoption of additive manufacturing approaches: the case of manufacturing SMEs, in IEEE International Conference on Engineering, Technology and Innovation (ICE/ITMC) (2018), pp. 1–8. https://doi.org/10.1109/ICE.2018.8436257 25. P. Schmitt, S. Zorn, K. Gericke, Additive manufacturing research landscape: a literature review. Proc. Des. Soc. 1, 333–344 (2021). https://doi.org/10.1017/pds.2021.34 26. L. Vendra, A.J.S.F.F. Achanta, Metal additive manufacturing in the oil and gas industry, in 2018 International Solid Freeform Fabrication Symposium (University of Texas at Austin, 2018). https://doi.org/10.26153/tsw/17035 27. N. Chiadamrong, C. O’Brien, Decision support tool for justifying alternative manufacturing and production control systems. Int. J. Prod. Econ. 60, 177–186 (1999). https://doi.org/10. 1016/S0925-5273(98)00182-0 28. P. Kulkarni, A. Kumar, G. Chate, P. Dandannavar, Elements of additive manufacturing technology adoption in small-and medium-sized sized companies. Innov. Manag. Rev. (2021). https://doi.org/10.1108/INMR-02-2020-0015 29. European Commission, Entrepreneurship and small and medium-sized enterprises (SMEs) (2022). https://ec.europa.eu/growth/smes_en, Accessed 09 Nov 2022 30. Wohlers Associates, Wohlers Report 2019: 3D Printing and Additive Manufacturing State of the Industry (Wohlers Associates Inc., Fort Collins, Colorado, 2019) 31. Wohlers Associates, Wohlers Report 2020: 3D Printing and Additive Manufacturing State of the Industry (Wohlers Associates Inc., Fort Collins, Colorado, 2020)

References

121

32. G. Piscopo, L. Iuliano, Current research and industrial application of laser powder directed energy deposition. Int. J. Adv. Manuf. Technol. 1–25 (2022) 33. J. Stavridis, Industrialization of Additive Manufacturing – A Holistic approach: From design to production (Unpublished doctoral dissertation). (Politecnico di Torino, Italy, 2022) 34. Eurostat, Electricity price statistics (2022), https://ec.europa.eu/eurostat/statistics-explained/ index.php?title=Electricity_price_statistics&oldid=575810. Accessed 09 Nov 2022 35. Arienespace, Vega user’s manual, issue 4 revision 0 April 2014 (2022), https://www.ariane space.com/ 36. Source: https://ec.europa.eu/growth/smes_en. Last accessed online on 15/06/2022 Accessed 09 Nov 2022

Chapter 5

Conclusions

Additive Manufacturing (AM) is one of the key technologies of Industry 4.0 offering unique advantages and capabilities. The interest in AM has been steadily increasing, leading to its rapid recent growth and improvement of all its aspects. However, its uptake is hindered by the lack of experience in the utilization of its advantages, as well as the minimization of its drawbacks in the various stages of production. In this book, a modular guide has been presented integrating the stages of AM product development in a practical, reader-friendly approach, aiming for the wider adoption of AM forming an essential aid for researchers, designers, engineers, simulation specialists, and post-graduate students. More specifically. • A holistic framework for AM has been followed, consisting of three pillars: design, processes, and applications. • A modular AM-driven design strategy toward exploiting the full design freedom potential of AM has been presented. • The AM process-centric approach of this book addresses the issues of product quality, energy efficiency, and build-time minimization in a step-by-step KPI optimization framework through all the stages for the stages of AM, namely design, process simulation, monitoring, and control, including digital twin and hybrid AM, post-processing, and industrial application-related issues, assisting in the evaluation of the entire supply chain of AM. Design for AM is a very complex subject, where numerous factors and their interactions should be taken into account in an attempt to maximize the benefit occurring by utilizing AM for a specific part/use case. This is because the design of every component to be additively manufactured can be affected and in parallel have effects both in the AM process and the materials selected, as well as on the subsequent post-processing steps required to obtain a ready-to-use component. In order to be able to successfully implement design for AM and deliver value through its use of it, the designer should not only consider the strong points of each AM process but also have in mind their inherent drawbacks. These factors can be © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 P. Stavropoulos, Additive Manufacturing: Design, Processes and Applications, SpringerBriefs in Applied Sciences and Technology, https://doi.org/10.1007/978-3-031-33793-2_5

123

124

5 Conclusions

“global” (i.e., applicable across all existing AM processes and orientating mainly to the layer-wise nature of AM) or process-specific, owing mainly to limitations related to the physical process mechanism. As such factors are highly dependent on the process and process mechanism, materials used, and (to a lesser extent) machine, only broad guidelines can be given, and fine-tuning of the design limitations for a specific machine and/or material requires targeted and structured experimentation based on the limiting factors expected by the process mechanism. Such experimentation can be facilitated by the use of standard test artifacts. These limitations should be taken into account during the early stages of a part’s design, as non-compliance is causing bottlenecks to the AM process. Both design aspects and design considerations should be considered. A design aspect is defined as any particular feature which can be quantified during the design phase. That includes geometric features of the part (overhangs, bores, channels, walls, etc.), as well as relevant build parameters that need to be set in order to manufacture the part (layer thickness, build orientation, etc.). The term “design consideration” can be used to describe the results of design aspects and the process itself on the finished product. That includes geometric characteristics and mechanical properties of the part, as well as KPIs of the AM process. To fully exploit the benefits of AM processes, transitioning from conventional feature-based design to function-based design using optimization algorithms is encouraged. Two main approaches topology optimization (TO) and generative design (GD) are currently used to morph the part geometry based on certain optimization criteria. One important limitation of both TO and GD methods is that they generate highly complex parts. Design modifications are then required in a second stage to address problematic aspects of the design and increase manufacturability; this fact has led to novel optimization methods that can take into account manufacturability aspects. Finally, parts designed for AM often need to be post-processed. The designer should be aware of the post-processing steps and the additional limitation their presence requires and design the part accordingly. Such limitations include the removal of unconsolidated material and support structures, as well as ensuring fixturing and referencing interfaces as well as sufficient strength and stiffness for further processing. Regarding AM processes, the selection of the material, the AM process family, and the specific machine that will be used is a multi-criterion decision-making problem that requires careful evaluation. It has to be pointed out that the intended use of a product is of critical importance because prototyping and end-use products have different requirements.

5 Conclusions

125

One of the most important issues hindering AM is that of part quality. The most important quality issues can be summarized as follows. • • • •

Surface roughness and layer-by-layer appearance. Porosity/void formation. Anisotropic microstructure and mechanical properties. Thermal residual stresses and deformations.

To achieve optimum part quality for a given combination of the process–AM machine–feedstock type, the optimization of the process parameters is crucial, especially for metal-based AM. The most important process parameters affecting roughness are layer thickness, laser power, scanning speed, power density, and overlap. Regarding pore formation, the most common reason mainly involves incorrect heat input in the melt pool and insufficient material overlap (hatch spacing). If a low energy density is used, the melting and fusion of the new layer with the previous one are not performed correctly, thus resulting in the formation of pores. However, the use of very high energy densities can lead to the same effect due to the creation of a keyhole in the melt pool, which leads to air inclusions during solidification. Due to the layer-by-layer nature of the AM processes, orientation dependence of mechanical properties is observed. To decrease the effects of anisotropy, careful selection of energy density and hatch spacing has to take place to allow for a sufficient fusion with the previously deposited layer and the adjacent deposited path. Additionally, the orientation of the part in the build chamber has to be optimized to ensure that the optimal mechanical properties are obtained depending on the load requirements for a specific part. The source of the anisotropy in AM lies in the grain evolution during manufacturing. The micro-structure can be controlled in situ utilizing re-melting, optimization of build-envelope temperature, as well as cooling rates and heating profiles, thus improving layer-by-layer grain uniformity. The role of the cooling rate is crucial: slower cooling rates lead to coarser grains, whereas faster rates lead to thinner ones. Moreover, the design of the part plays an important role because thicker cross-sections cool slower than thinner ones, hence leading to coarser microstructures. Additionally, the uneven heating and cooling that take place during the manufacturing of parts in AM lead to the development of thermal stresses and deformations, both in the build direction (Z-axis) and in the horizontal plane (XY). The cause for this phenomenon is the non-uniformity of the thermal field both in the build direction (Z-axis) and in-plane (XY), and it can be classified into three categories: (i) induced stresses from upper layers to the below solidified layers, (ii) thermal contraction of the current layer, and (iii) thermal gradients in the XY-plane. A more uniform distribution of temperature along the Z-axis is required for the reduction of the intensity of both the first and the second categories. This can be achieved by heating the base plate to lower the thermal gradient in the Z-direction. Additionally, the heating of the machine chamber helps in the mitigation of this effect. Both of the previously mentioned actions decrease the intensity of the in-plane thermal gradients as well. However, the latter category can be more effectively tackled by evaluating the different scanning strategy alternatives in terms of the resulting XY-plane thermal

126

5 Conclusions

gradients and by selecting the one that leads to the lowest. The scanning strategy selection criterion can decrease the in-plane thermal non-uniformity by up to 20% for a given part. The most important process parameters of thermal-based metal AM processes and their impact on crucial KPIs have been analyzed since the quality of parts in AM is highly dependent on their optimization of process parameters. The manner in which the heating for the fusion of a new layer with the previous one is performed has a very significant impact on part quality because it affects almost all the KPIs. The combination of heating intensity (laser power) and scanning speed form the energy density, which is the determining factor for the maximum temperature, met-pool dimensions, and the fluid-dynamics phenomena that take place. These phenomena determine the surface roughness, part density, mechanical properties, and microstructure of the final parts; therefore, their optimization is of crucial importance. Energy density and the different combinations of laser power and speed have to be defined taking into account the material used, the required part density, build time, and cost of the part, as well as the up-skin and down-skin considerations. Additionally, the scanning strategy has a direct impact on residual stresses, deformations, mechanical properties, anisotropy, and build time. The phenomena that take place in AM are implicit, multi-scale, and highly dynamic in time and space. It may be observed that for the creation of an allencompassing holistic simulation, a coupled multi-disciplinary and multi-level simulation is required. However, such an approach would have prohibitive computational requirements. Instead, the study of the different scales and phenomena in separate simulations is the most common approach, using the optimized process parameters from one simulation level to the next. To further understand how to utilize simulation for AM efficiently, the connection of the phenomena that take place and their impact on quality has been established. It should be noted that different phenomena require the simulation of different scales based on their physical mechanism. Regarding process control, the most common process parameters are the laser power and speed because they determine the energy density and melt-pool geometry, both of which are crucial for part quality. Through IoT and Industry 4.0 technologies, the development of cloud-based control applications has been made possible, allowing for easier and more user-friendly process-control applications. Post-processing for AM can be divided into two categories: essential and nonessential. The cleaning of the part, removing the excess material from the build chamber (e.g., powder-based processes), and removal of the part from the base plate and of the supports are considered essential. Once those actions are completed, there is a plethora of post-processes for AM serving different purposes. However, particular emphasis must be placed on the thermal-based post-processes because they allow the minimization—or even complete elimination—of the anisotropy and the residual stresses (for thermal AM processes). Moreover, thermal post-processes allow the calibration of the mechanical properties of a product (ductility, tensile strength, yield strength, elongation at break), as well as of its grain size, to the exact needs of a particular application within the limits of the material used for its production. It should be noted that post-processing plays an important role in the

5 Conclusions

127

final part cost, and as such, it needs to be taken into account. To minimize the time and, consequently, the cost of AM, the integration of common post-processes into the AM machine itself led to hybrid AM machines. There is a selection of commercially available hybrid AM machines; however, a custom solution can be developed to be tailored to specific needs. The utilization of digital twins in AM has various advantages, namely the minimization of unplanned downtime, accidents, as well as maintenance costs through preventive maintenance; all of the aforementioned constitute factors that contribute to improved production time. Finally, they are powerful process optimization tools because they are based on optimized simulations that are continuously updated and trained through real-time data. Regarding AM applications, the current state of end-use metal AM products has been presented, and the most mature industrial sectors regarding metal AM (MAM) have been highlighted. Three applications were developed in different sectors, namely aerospace, dental, and energy. The cases regarded both PBF and DED technologies. The requirements of the end-users were analyzed, and all the steps from the feasibility and cost evaluation, the material characterization, the process parameters definition, the actual printing, the post-process activities, and the performance evaluation were executed and presented. To achieve the demanding process and production requirements, the cost for several machinery modifications and different configurations was also taken into account. Additionally, possible innovations and upgrades were studied, which should be considered for future industrial MAM applications that will contribute to the increase of production numbers and automation. Calculating the operating costs and the market aspects provided by the end-users, a business evaluation in a 5-year long period has been presented demonstrating the real economical advantages that MAM can offer in production. Those indicative cases that were studied also provide insight for future changes that have to be made in industrial AM equipment, to facilitate the needs of more industrial sectors in a cost-efficient way, retaining the characteristic high product value of AM. An AM investment for an aerospace application can be very profitable, given that the required considerations and programming take place. Additionally, the utilization of AM instead of conventional processes led to improved product performance, offering technical advantages, and fewer production steps, eliminating various operating, maintenance, and consumables costs. Since AM technology becomes more mainstream, offering increasing production rates and lower investment costs, the profit margin will steadily increase, constituting AM a viable alternative for other sub-assemblies of the rocket and aerospace industry. The dental sector is a field in which AM can replace the current manufacturing method and provide advantages not only at an economic level but also technical, offering a higher degree of customization, flexibility, and the possibility to investigate new material alloys compatible with the human body. The presented AM setup also offers the capability for future investigation of new, innovative material alloys, compatible with the human body, which could not be used cost-effectively by conventional manufacturing technologies. The AM repair application from the power and energy industry has demonstrated the technical feasibility and advantages of DED technology. Furthermore, it has to be

128

5 Conclusions

highlighted that the integration of tailor-made automation and sensors can be the turning point that enables the cost-effective and high productivity utilization of AM. Consequently, important innovations from a machinery point of view that can facilitate other similar applications in other sectors as well can lead to the improvement of the requirements and specifications of the next generation of industrial AM machines, allowing for larger lot-size productions, paving the way for AM as a cost-effective repairing manufacturing alternative for every factory. The biggest cost for AM is the initial investment. Even though this cost can fast be mitigated through the increased revenue of AM end-use parts, there is another alternative, which nullifies the initial investment cost, namely the cooperation with research centers, service provider companies, or other similar entities, which are also supported by research funds. This approach allows a company to test the AM technology, realize new products, and use AM machinery without the need for an initial machinery investment, paving the road for the integration of AM as an in-house production method of said company. AM will gain a higher share in many industrial sections in the next years due to advancements in machinery and process level, such as increased productivity and working volume, the introduction of new materials, new laser wavelengths, automation, and custom configurations. AM also provides alternative business models for manufacturing on demand, as well as decentralized production through a network of valid and high-competence partners and collaborators. Subsequently, the increase in demand will bring more incentives for AM investments, leading to lower initial and operational costs, in this way constituting AM investments viable even from Small–Medium Enterprises. Based on the results and requirements obtained from the application cases to date, the possible innovations and developments needed to be carried out at the machinery level to promote competitiveness and integrate AM into the manufacturing reality of today mainly lie in facilitating the industrialization of AM aiming for higher production rates. This will allow the depreciation cost to be divided over a larger number of parts, offering smaller depreciation periods and in this way rendering AM a cost-effective alternative for a wider range of applications and sectors. This can be achieved using faster-operating speeds, larger build volumes, optimized part nesting, increased automation, reduced trial and error approach, simulations, and increased processing capacity of new materials. AM is leading to profound changes in the entire manufacturing product chain; therefore, the integration of an AM machine into the existing manufacturing chain is not sufficient for its fruitful application. The successful incorporation of AM into the manufacturing reality of the future requires the holistic training of the new generation of professionals, as well as having access to helpful and practical guidebooks regarding all the aspects of AM.