National Academies Press: OpenBook

Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop (2019)

Chapter: 6 Summary of Challenges from Subgroup Discussions and Participant Comments

« Previous: 5 Product and Process Qualification and Certification
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×

6

Summary of Challenges from Subgroup Discussions and Participant Comments

During the third day of the workshop, participants met in subgroups to discuss some of the challenges in additive manufacturing (AM). These groups aligned with the four sessions of the workshop:

  1. Measurements and modeling for process monitoring and control;
  2. Developing models to represent microstructure evolution, alloy design, and part suitability;
  3. Modeling aspects of process and machine design; and
  4. Accelerating product and process qualification and certification.

Breakout groups were asked to discuss two or three principal topics and consider the following overarching questions:

  • What are the greatest technological challenges?
  • What are the most important areas for research?
  • What are “nontechnical” challenges to commercialization of AM?
  • How can industry and academia better interact and collaborate to address technical and nontechnical challenges?
  • Are there concrete actions that could help address the challenges identified?
  • What topics could be addressed in a follow-on workshop?

Workshop participants were also asked to provide individual responses to similar questions about top priority research needs for advancing AM, top “nontechnical” challenges to commercialization of AM, and actions

Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×

that could help address these nontechnical challenges. Descriptions of the subgroup discussions and individual responses are provided in the following subsections.

MEASUREMENTS AND MODELING FOR PROCESS MONITORING AND CONTROL

Subgroup Members

Jarred Heigel (National Institute of Standards and Technology), Carolin Körner (Friedrich-Alexander Universität Erlangen-Nürnberg), Amit Surana (United Technologies Research Center), R. Allen Roach (Sandia National Laboratories), Kilian Wasmer (Empa), Shoufeng Yang (KU Leuven), and Celia Merzbacher (SRI International)

Breakout Discussion

This breakout group discussion was led by Heigel, and conversations focused on sensor technology, algorithm development and use, knowledge transfer, challenges, and priorities moving forward. The following three questions were proposed to start the discussion:

  1. What is good enough? How much information is needed from the process to meet the desired goals? Some subgroup members noted that clearly identifying what process information is needed will enable the development of useful sensors.
  2. What information must be exchanged between real-time monitoring sensors and process models? Several subgroup members commented that this is specific for model-based control and is an immediate need.
  3. How can decisions and guidelines be made for processing and saving measurements? Many subgroup members commented that this question also addresses issues of data management.

Several subgroup members highlighted the following technical challenges:

  • Correlating process phenomena with structures and defects and incorporating real data into process models. This could help improve the understanding of the overall AM process and the underlying physics (e.g., understanding what may increase the chance of failures or unsatisfactory parts), which in turn could help improve the sensor design and the data analysis.
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
  • Clarifying what needs to be measured to control the outcome. This may include defining the industrial needs for real-time monitoring.
  • Understanding the material, structure, and defect specifications. This is particularly important in terms of understanding areas of concern within the part regarding defects and what defect density is acceptable. It is important to be able to define what is and is not acceptable for specific parts and design criteria.
  • Providing better input and output definitions for models and sensors. This could help to improve communication throughout the system. While each sensor represents a different aspect of the process, they can provide a more complete picture of the process when they are combined.
  • Assessing whether sensor systems are capable of measuring critical parameters and providing real-time analysis. It is important to question what hardware and analysis are needed if current systems are not fast enough to enable sufficient process control.
  • Enabling the long-term goal of a feed-forward loop based on reliable models. This is a significant challenge that is also dependent on the previously mentioned challenges.

Subgroup members discussed the challenge of machine interoperability and how to encourage machine manufacturers to be more transparent with their systems and processes. Currently, the high cost of developing these systems and the associated intellectual property deters manufacturers from making their systems more transparent. However, many manufacturers are small organizations that may lack the resources and expertise required to develop real-time monitoring and process control strategies required by the end user. On the other end of the spectrum, organizations with monitoring and control expertise often are not as familiar with the intricacies of the process and lack the ability to communicate directly with the machines. Some members of the subgroup speculated that case studies and cost analysis could help to convince manufacturers that increasing the transparency of their machines and enhancing collaboration will serve the greater good of the AM community and ultimately increase manufacturers’ customer base. The semiconductor industry—which has benefited from collaborations and partnerships—could be an exemplar of how to encourage transparency and collaboration among small companies. Finally, many subgroup members suggested that a follow-on workshop could focus on data collection and improved decision making.

Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×

DEVELOPING MODELS TO REPRESENT MICROSTRUCTURE EVOLUTION, ALLOY DESIGN, AND PART SUITABILITY

Subgroup Members

Annett Seide (MTU Aero Engines), Lyle Levine (National Institute of Standards and Technology), John Turner (Oak Ridge National Laboratory), Ade Makinde (General Electric Global Research Center), Kyle Johnson (Sandia National Laboratories), Eric Jägle (Max Planck Institute), Deniece Korzekwa (Los Alamos National Laboratory), and Christian Leinenbach (Empa)

Breakout Discussion

This breakout group discussion was led by Seide, who noted that many of the challenges in representing microstructure evolution, alloy design, and part suitability are encompassed by the larger material science research effort. However, there are unique areas of ongoing research that are specific to AM materials and conditions. The subgroup first discussed the lack of thermophysical data under AM conditions, and several members suggested the following short-, intermediate-, and long-term goals:

  • Short-term goals: Identify the data that are needed for process measurement and for modeling input.
  • Intermediate-term goals: Obtain data for a limited set of AM materials.
  • Long-term goals: Take a deep look at the quality of data and explore first principles and machine learning approaches.

These members emphasized that the most important areas of research for the lack of thermophysical data are first principles and machine learning approaches.

The subgroup next discussed microstructure evolution and the challenge of developing and validating models. In particular, several members described high-fidelity models, coupled multiphysics models (e.g., to get location-specific microstructure evolution and to look through the solidification and intrinsic heat treatment processes as well as post-build processing), and reduced-order models as particularly challenging. Many subgroup members noted several promising short-, intermediate-, and long-term research areas:

  • Short-term research areas: Conducting sensitivity analysis of current model parameters, coupling physical phenomena in models,
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
  • temporal and spatial scale bridging, and three-dimensional microstructure characterization information.

  • Intermediate-term research areas: Modeling phenomena of interest under nonequilibrium conditions and calibration of model parameters.
  • Long-term research areas: Predicting metastable phases, predicting models for nucleation, and predicting interfacial energies.

The third topic discussed was coupled multiphysics and multiscale capabilities for AM, including the laser-material interaction, the time-dependent thermal profile (including fluid flow), microstructure evolution, micromechanics, and macroscale thermomechanics. Many subgroup members noted several promising short-, intermediate-, and long-term research areas:

  • Short-term research areas
    • Analysis of coupling between relevant physics. This is challenging because it requires a fully coupled model.
    • Exploration of approaches for modeling laser-scan strategies. These approaches may be done through parallel in-time approaches.
    • Prediction of site-specific properties. This includes properties throughout the parts and the ability to use site-specific properties in macroscale models.
  • Intermediate-term research areas
    • Development of reduced-order models informed by both high-fidelity models and experimental data.
    • Development of advanced design optimization tools and approaches.
    • Advancement of site-specific control of microstructure through process parameters for real parts, including complex shapes and complex alloys.
  • Long-term research areas: Integration of site-specific microstructure control into design optimization.

Several subgroup members also discussed the following nontechnical challenges across AM:

  • The lack of stable, long-term research funding;
  • A lack of willingness to fund testing and measurement;
  • The use of proprietary alloys;
  • The lack of a community standard file and standardized formats for experimental and simulation data;
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
  • The need for increased collaboration between domain scientists and computational scientists; and
  • The lack of students and staff with necessary expertise such as computational material science, manufacturing, model-based engineering, computer-aided-design-based topology optimized design, and software development for modern computer architecture.

To increase collaboration and better address technical and nontechnical challenges, several subgroup members suggested that industry and academia support efforts that provide foundations for collaboration (e.g., AM-Bench). Industry might consider funding defined challenges in which academia and laboratory teams could compete. Programs could be created for targeted collaborative industry–academia–laboratory research to tackle specific application challenges. These subgroup members emphasized the importance of having adequate, stable funding available over extended time periods and suggested that the U.S. Department of Energy Hubs1 concept could be applicable for AM. Many subgroup members suggested specific actions that could help address these challenges, including a call for proposals in the industry–academia–laboratory research areas and the expansion of educational programs that are domain specific and multidisciplinary. Some members of this breakout group suggested that a follow-on workshop could address topics such as challenges and opportunities in topology and shape optimization with site-specific microstructure control as well as multidisciplinary educational programs for AM processes.

MODELING ASPECTS OF PROCESS AND MACHINE DESIGN

Subgroup Members

Mustafa Megahed (ESI Group), Wing Kam Liu (Northwestern University), Jian Cao (Northwestern University), Tahany El-Wardany (United Technologies Research Center), and Winfried Keiper (European Technology Platform for Advanced Engineering Materials and Technologies)

Breakout Discussion

Megahed led this breakout group, which discussed modeling aspects of process and machine design. Megahed, Liu, Cao, El-Wardany, and

___________________

1 For more information on the U.S. Department of Energy’s Hubs, see https://www.energy.gov/science-innovation/innovation/hubs, accessed March 11, 2019.

Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×

Keiper proposed the following challenges and research needs for this topic:

  • Identifying the source of process variability, which can be done by determining the sensitivity of the process to certain parameters, uncertainty quantification, and process control.
  • Calibrating and validating models, even in the absence of experimental data.
  • Designing the experiments needed to deliver necessary data.
  • Developing a community database for relevant data in standardized forms.
  • Advancing models to capture details such as environmental effects, alloying elements, and doping.2 These may utilize artificial intelligence and machine learning methods.
  • Improving the use of data reduction and reduced-order modeling to increase efficiency.

These subgroup members also highlighted three major nontechnical challenges:

  • Data sharing. The research community would benefit from increased access to data. Several subgroup members speculated that the reticence to share data might be a cultural problem since most researchers are not used to sharing their data. They highlighted nuclear physics databases as a possible example to emulate, particularly the use of a centralized body to help transform raw data into evaluated data. Shared databases also need to be sustainable as well as continually maintained and updated.
  • Interpretable machines. Manufacturers have historically been reluctant to share the inner workings of their machines for a variety of business reasons. However, these subgroup members noted that having more transparent machine processes would enable research advancements.
  • Interdisciplinary education. These subgroup members explained that there needs to be a more efficient way of learning about a wide variety of topics relating to AM, including hardware, underlying physics, metrology, algorithm development, optimization, numerical simulation, thermodynamics, statistics, and data analytics.

___________________

2 Alloying elements are defined as metallic or nonmetallic elements that are added in specified or standard amounts to a base metal to make an alloy (Business Dictionary, 2019), and doping is the mixing of a small amount of an impurity into a silicon crystal (Brain, 2001).

Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×

The subgroup members also discussed partnerships. Several members noted successful models such as America Makes,3 Horizon 2020,4 CleanSky,5 and other data sharing efforts that encourage community databases. Other options could be for industries to enable more internships and fellowships for students and researchers. A number of subgroup members also suggested more partnerships among researchers in the European Union and the United States and among small- and medium-size enterprises; this could encourage more collaboration, data exchange, and international research funding.

For a possible follow-on workshop, some of the subgroup members proposed themes including the definition of joint standards and tolerances, digital twin and threads for AM, interdisciplinary education, and the various intermediate-term challenges and goals that were discussed throughout the workshop.

ACCELERATING PRODUCT AND PROCESS QUALIFICATION AND CERTIFICATION

Subgroup Members

David Teter (Los Alamos National Laboratory), Jens Telgkamp (Airbus Operations GmbH), Vincent Paquit (Oak Ridge National Laboratory), Paolo Gennaro (GF Precicast Additive SA), Johannes Henrich Schleifenbaum (Fraunhofer Institute for Laser Technology), Richard Ricker (National Institute of Standards and Technology), Josh Sugar (Sandia National Laboratories), and Ben Dutton (Manufacturing Technology Centre)

Breakout Discussion

Teter and Telgkamp led the discussion for this subgroup, which focused on accelerating product and process qualification and certification. This discussion was divided into short-term (less than 5 years) and intermediate-term (5 to 10 years) goals that could enable a long-term vision for AM.

Teter explained that the long-term vision is the ability to design, print, and qualify a product correctly the first time. This includes as-built quality, in which people have very limited destructive evaluation for parts

___________________

3 For more information on America Makes, see https://www.americamakes.us, accessed March 11, 2019.

4 For more information on Horizon 2020, see https://ec.europa.eu/programmes/horizon2020/, accessed March 11, 2019.

5 For more information on CleanSky, see https://www.cleansky.eu, accessed March 11, 2019.

Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×

being generated, and built-in quality assurance, in which data are collected as a part is being printed. Modeling and simulation play an important role—a multiphysics process–structure–property–performance prediction is needed. Cybersecurity is another concern, particularly in terms of building resiliency to the threat of fraudulent components over the next 10 years. Several subgroup members noted that the ability to track each part is needed, including attaching the license to build and proof of quality to each part. Lastly, some subgroup members commented on the need for government-to-government agreements on AM with shared objectives, data, and frameworks. They suggested that long-term efforts should focus on the need for AM to be operational and fully accepted by certification groups.

To advance this long-term vision, the subgroup highlighted the following short- and intermediate-term research goals:

  • Short term: Several subgroup members suggested a short-term focus on AM technology and materials development, such as making the process less sensitive to variability and defects. Below are some specific open challenges that these members highlighted.
    • Improving the understanding of the influence of feedstock parameters, taking into consideration the key material properties and process parameters. These subgroup members emphasized this as a high priority.
    • Developing guidance on sensor technology.
    • Improving the openness of control systems.
    • Refining the definition of “good” data as well as a common test part/object for qualification and microstructure. ASTM F426 may be able to help determine goals, objectives of test part/object, and number of object definitions needed.
    • Collecting defect catalogues for critical flaw size and type, frequency, distributions, and criticality of locations. Telgkamp noted that this is particularly important for highlighting research and development needs.
    • Strengthening the understanding of current sensor technology, limits, capabilities, stability, and reliability.
  • Intermediate term: Several subgroup members suggested an intermediate-term focus on continuous standardization activities, such as development and maturation. Below are some specific open challenges that they identified.

___________________

6 For more information on ASTM F42, see https://www.astm.org/COMMITTEE/F42.htm, accessed March 11, 2019.

Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
  • Using process monitoring in decision making, such as when and how to repair a part or when to discard it.
  • Developing reduced-order models for decision making.
  • Maturing tools for sensor data fusion and reduction.
  • Exploring machine learning methods to improve microstructure and property predictions.
  • Increasing data sharing and establishing a common or global database. These subgroup members noted that this was mentioned throughout the workshop.
  • Improving machine-to-machine knowledge transfer.
  • Developing high-throughput characterization and development for new and mature sensors, based on the sensing needs to be identified.

INDIVIDUAL RESPONSE RESULTS

Participants at the workshop were also asked to provide their thoughts on the top priority research needs for advancing AM, top “nontechnical” challenges to commercialization of AM, and actions that could help address these nontechnical challenges. The individual responses were analyzed by a workshop subgroup and summarized by Celia Merzbacher (SRI International). She explained that the technical challenges suggested by the workshop participants centered on needing more AM materials, improving the understanding of microstructure prediction, developing standards and benchmark measurements, and improving in-situ monitoring capability. For nontechnical challenges, she explained that the responses centered on encouraging data sharing, increasing funding, improving training and education, enabling machine transparency, and increasing trust in AM parts. Many participants suggested that these challenges could be approached by increasing coordination and communication among stakeholders, perhaps through more convening activities, collaborations, standards, and funding.

REFERENCES

Brain, M. 2001. “How Semiconductors Work.” April 25. HowStuffWorks.com. https://electronics.howstuffworks.com/diode1.htm, accessed March 11, 2019.

Business Dictionary. 2019. “Alloying Element.” http://www.businessdictionary.com/definition/alloying-element.html, accessed March 11, 2019.

Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
Page 47
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
Page 48
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
Page 49
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
Page 50
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
Page 51
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
Page 52
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
Page 53
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
Page 54
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
Page 55
Suggested Citation:"6 Summary of Challenges from Subgroup Discussions and Participant Comments." National Academies of Sciences, Engineering, and Medicine. 2019. Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25481.
×
Page 56
Next: Appendixes »
Data-Driven Modeling for Additive Manufacturing of Metals: Proceedings of a Workshop Get This Book
×
Buy Paperback | $65.00 Buy Ebook | $54.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Additive manufacturing (AM) is the process in which a three-dimensional object is built by adding subsequent layers of materials. AM enables novel material compositions and shapes, often without the need for specialized tooling. This technology has the potential to revolutionize how mechanical parts are created, tested, and certified. However, successful real-time AM design requires the integration of complex systems and often necessitates expertise across domains. Simulation-based design approaches, such as those applied in engineering product design and material design, have the potential to improve AM predictive modeling capabilities, particularly when combined with existing knowledge of the underlying mechanics. These predictive models have the potential to reduce the cost of and time for concept-to-final-product development and can be used to supplement experimental tests.

The National Academies convened a workshop on October 24-26, 2018 to discuss the frontiers of mechanistic data-driven modeling for AM of metals. Topics of discussion included measuring and modeling process monitoring and control, developing models to represent microstructure evolution, alloy design, and part suitability, modeling phases of process and machine design, and accelerating product and process qualification and certification. These topics then led to the assessment of short-, immediate-, and long-term challenges in AM. This publication summarizes the presentations and discussions from the workshop.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!