The Standing Committee on Defense Materials, Manufacturing and Infrastructure (the DMMI standing committee) convened a workshop on December 5 and 6, 2012, to discuss new and novel processes that are on the verge of industrial modernization. The standing committee is organized under the auspices of the National Materials and Manufacturing Board of the National Research Council (NRC) and is sponsored by Reliance 21, a Department of Defense (DOD) group of professionals that was established in the DOD science and technology (S&T) community to increase awareness of DOD S&T activities and increase coordination among the DOD services, components, and agencies.
The workshop was conducted as a convening activity. In accordance with NRC procedures, all participants provided individual opinions at the meeting, and no consensus findings, conclusions, or recommendations were developed at the workshop or as an outcome of the workshop. This report is a record of the workshop event prepared by the workshop rapporteur, and any statements or views summarized in the report must be considered an opinion expressed by an individual participant at the workshop, not a consensus view.
To organize its workshop on new and novel processes, the DMMI standing committee first organized a workshop planning group to identify workshop topics and agenda items, speakers, and invited guests. The workshop planning group consulted with Reliance 21 and members of the community to develop and organize the workshop. The workshop was held at the Keck Center of the National Academies, in Washington, D.C. Approximately 55 participants, including speakers, members of the DMMI standing committee, Reliance 21, invited guests, and mem-
bers of the public, took part in the 2-day workshop. The workshop was organized into three sessions, focusing on the following topics:
• Additive manufacturing,
• Electromagnetic field manipulation of materials, and
Additive manufacturing is defined as the process of making three-dimensional (3D) objects from a digital description or file; it is considered an additive process because the materials are deposited in successive layers. Electromagnetic field manipulation of materials is the use of electric and/or magnetic fields to change the mechanical or functional properties of a material or for the purposes of sintering. “Design of materials” refers to the application of computational and analytic methods to materials to obtain a desired material characteristic.
To assist the reader, recurring themes from the workshop discussion are summarized below, organized by session topic. These recurring themes represent discussion items that were addressed by multiple speakers or participants during the course of the workshop; they were identified for this report by the rapporteur, not by the workshop participants. The recurring themes are as follows:
• Additive manufacturing
—Qualification and certification
—New materials for additive manufacturing
—Access to manufacturing capabilities and resources
—Unlawful uses of additive manufacturing
• Electromagnetic field manipulation of materials
—Understanding the physical phenomena
—Permanence of the material changes made
After briefly describing each recurring theme in this overview, the report goes on to summarize the workshop presentations and discussions. Appendix A contains the statement of task for the workshop, Appendix B lists the workshop participants, and Appendix C contains the workshop agenda.
ADDITIVE MANUFACTURING THEMES
Qualification and Certification
It was noted by multiple participants and speakers that qualification and certification is a significant challenge in additive manufacturing. The qualification and certification process in its current form is not well suited to additive manufacturing. With traditional bulk manufacturing techniques, by way of contrast, all of the initially produced parts manufactured are certified; then, a sampling (perhaps 2 percent) of subsequent manufactured parts is certified. In additive manufacturing, where parts are individualized, the small sample size will not allow for statistical analysis. It was noted that perhaps the process should be certified, not the part. It was important to connect the additive manufacturing community with certification experts.
Each of the three speakers was asked about the impact of certification on additive manufacturing during the question-and-answer period following his presentation. Prabhjot Singh was asked to describe the certification process in general. He spoke of a very involved process, where the specification is written for each materials processing step. He pointed out that the current certification process takes 10-12 years, whereas he would like it to take 1-2 years. In response to a question from the audience, David Bourell noted that while from a research perspective it may make sense to develop new materials that would be better suited to the additive manufacturing process, DOD may not be interested in a new material because of the lengthy and tedious certification process associated with its use. Richard Martukanitz, in response to a question following his presentation, pointed out that the welding industry might be a good qualification model for the additive manufacturing industry; a welded structure has a design code to provide guidelines, qualified materials, qualified operators, processes, and repeatability within a range of process parameters.
New Materials for Additive Manufacturing
Several participants questioned whether it would be either needed or desirable to develop new materials that would be optimized for the additive manufacturing process. On the one hand, it could be useful to have materials optimized to this new manufacturing technique. On the other hand, however, new materials might add complications of their own that outweigh their potential benefits. Dr. Bourell pointed out that while most of the materials currently used in additive manufacturing are polymers and plastics, there is much interest in the development of new metals and new ceramics for additive manufacturing. He went on to say that there is a need to develop new metal alloys with characteristics appropriate for additive
manufacturing; however, because of the certification concerns noted above, new metal alloys might not be considered desirable by either DOD or industry.
In the session on the design of materials, Gregory Olson spoke of research interest in investigating metals for additive manufacturing. He described research in the use of design tools to create new alloys optimized for the capabilities of the additive manufacturing process. Researchers are particularly interested in new alloys with relatively fast solidification rates for additive manufacturing. Dr. Olson specifically mentioned the Open Manufacturing project of the Defense Advanced Research Projects Agency (DARPA) in collaboration with Honeywell, in which laser-based additive manufacturing is being considered.
During the open discussion following all three presentations, participants agreed that materials in production today are not specifically tailored to additive manufacturing and thus might be somewhat deficient. Further, participants noted that designers are not very involved in the materials process; they are the ones who should be giving materials scientists the specifications for the new materials.
Access to Manufacturing Capabilities and Resources
The lack of access to processing equipment, raw materials, and appropriate modeling tools was consistently noted as a potential handicap for the additive manufacturing community. Several participants pointed out that because both the electron beam and the laser system used in additive manufacturing are primarily made by companies that are not based in the United States, access to these manufacturing processes can be difficult. Both systems use a closed architecture, where the processes and materials used to service the equipment are known only to the manufacturer. Several participants discussed the need to have machines supplied by more companies and to have a more open, collaborative environment. Dr. Singh, who discussed this in particular during his presentation, was asked about access to machines in his question-and-answer session. He responded that no U.S. companies make electron beam machines and only one Swedish company does, which makes the supply quite limited. While six companies manufacture laser machinery, electron beam processing is 4-5 times faster than laser processing. Dr. Martukanitz also discussed the access to processing equipment in his talk: He pointed to the need for an open architecture for the equipment system used to allow everyone to have full access to the processes. The powder bed electron beam system currently used is a closed system—that is, the parameters and the materials used to service the equipment are held closely by the equipment manufacturer.
During the open discussion period it was noted that, in general, there are more manufacturing capabilities in Europe than in the United States. This led to a discussion on how the United States could stimulate its own manufacturing capabilities. Some participants noted that research in Europe tends to be more application-
specific, or at least more project-specific, and that European governments award more grants for additive manufacturing processes. The European Union also builds on a more traditional approach to technology transfer, which tends to lead to more manufacturing. Other participants responded that research in the United States is more fundamental, and when it proves useful, the hope is that it will work its way into industry by market forces.
During the open discussion period, participants noted that while there is a U.S. source for most materials feedstocks, some manufacturers for niche applications may have more difficulty sourcing their material.
Another resource discussed was the modeling tools needed for additive manufacturing. Dr. Martukanitz’s presentation focused heavily on the modeling and simulation tools for virtual experimentation that are being developed at Penn State University. During the open discussion period, participants discussed the lack of a standard modeling tool set for additive manufacturing. They noted that the current tools are better suited for the mechanistic models than for the process or materials structure and composition models.
While not discussed in detail, surface finish was brought up as a research area that would benefit from additional research and development. Dr. Bourell noted that research into surface finish has tended to lag behind research into other areas and was in need of technical advancements. During the general discussion participants agreed that surface finish is consistently recognized as needing more technical research and development.
Unlawful Uses of Additive Manufacturing
The group discussed the potential dangers associated with putting additive manufacturing technology in the wrong hands. Participants discussed how additive manufacturing could be used to make unlicensed weapons, unlawful versions of copyrighted material, and substandard copies of goods at a lower cost. In each of these cases, however, there are circumstances that may make the application of additive manufacturing less attractive to a potential criminal. In the case of unlicensed weapons, for example, the weapons could only be made of plastics using current technology, and thus the weapons would not be practical. In the case of copyrighted material, the issues that arise for 3D designs are not substantially different from those for two-dimensional (2D) designs and can be addressed in the same way. In the case of substandard copies, the costs associated with the small production runs in additive manufacturing are high enough to deter criminals.
These topics were discussed by Dr. Bourell in his presentation and by the group as a whole in the discussion that ensued.
ELECTROMAGNETIC FIELD MANIPULATION OF MATERIALS THEMES
Understanding the Physical Phenomena
Overall, participants noted that processing with electromagnetic fields offers a wealth of opportunities. However, participants also noted the need to better understand the underlying phenomena that explain the physical changes caused by processing with electromagnetic fields. Rosario Gerhardt began the open discussion period by remarking on how thoroughly the presenters were sold on the ability of high magnetic fields to improve mechanical and magnetic properties of the materials, as well as the ability of electric fields to enable sintering capabilities. However, researchers do not always understand the physics behind the technologies. For instance, Robert Dowding explained in his presentation that the mechanisms to describe electric field-assisted sintering technology (FAST) are not well understood and discounted the most popular explanations, particularly any relationship to plasmas, sparking, or even sintering, despite the fact that the most popular name for the technology is spark plasma sintering (SPS). Despite the poor understanding of the underlying physics, Mr. Dowding pointed out that the process is still a very useful one, with lowered sintering times and temperatures, increased production rates, and lower energy costs. There were also several examples of processing in a high magnetic field in which the underlying mechanisms are currently being explored.
A number of applications of electromagnetic field processing techniques and applications were described in this session. While some participants noted that high magnetic fields offer several intriguing possibilities for further exploration, others wondered what research idea(s) the community should investigate in detail next. The workshop goals did not include focusing on or prioritizing research ideas, and several participants remarked on the vast array of examples provided. Examples of magnetic field processing techniques and applications given in the session on electromagnetic field manipulation included these:
• High and thermomagnetic processing (H&TMP),
• Examining creep life in metal alloys,
• High-temperature superconductors,
• Compaction and sintering of commercial magnet materials,
• Next-generation composites,
• Point defects,
• Oxidation reactions,
• Polymeric and organic materials,
• Magnetoplasticity and reduction of residual stress,
• Functional material applications (nanomaterials for magnetic storage, nanostructures, superconducting materials, and magnetic control of hydrogen storage),
• Electromagnetic acoustic transducer (EMAT) solidification/casting technology,
• Single crystal growth of iron-nickel-cobalt alloys,
• Crystallizing biological macromolecules,
• Changing the structural properties of high-strength steels,
• Processing and aligning nanotubes,
• Improving rare earth magnets, and
During the open discussion period, Dr. Gerhardt said that there were many examples but not many details in the presentations in this session. She said that while the presenters clearly advocated the technology and the ability to change the material microstructure at will, she wanted to get a sense of priority among the many applications. Dr. Gerhardt asked the presenters which area(s) would have the highest priority if only one or two areas of research were possible. Gerard Ludtka responded by saying that industry would not be interested in high-temperature superconductors. Industry would be most interested in changing the mechanical properties of materials. DOD would want to push the envelope for performance, so they might be more interested in the high-temperature superconducting magnet technologies. Ke Han noted that industry would probably want to get the most value for the smallest investment. For instance, manufacturers would like to use a low-performance material but find a way to get a 20-30 percent improvement from it. A 1 T magnet is cheap enough and can be built large enough to contain an entire truck for processing. This type of practical application might be useful to industry.
Permanence of Material Changes Made
Participants asked the speakers whether the changes to a material brought about by an electromagnetic field are permanent, and what aspects of the material can or cannot be relied upon. During the open discussion, Dr. Gerhardt asked the group if, once a field is used to induce a particular change, the resulting material is in a metastable state. Would the treated material revert to its original state at different temperatures? Or could it change its properties in some other way? Dr. Han
replied that it depends on what property has been changed. If a material’s texture is changed and the material is in a solid state, the texture will not be lost. If the solubility is affected at a high temperature, that effect will not be lost. But if a material is processed at a certain temperature and then subjected to a higher temperature, the effect that had been achieved could be lost. The issue of metastability was also discussed during the question-and-answer period following the presentation of Gregory Boebinger and Dr. Han, who were asked if the material property changes were permanent. Dr. Han responded that at ambient temperatures, the changes are permanent in the examples presented. However, the materials would reanneal at different temperatures, which could further change material properties. Another participant said that high-temperature solution treatments tend to be stable, but it is important to understand what has changed and what has been manipulated in the microstructure.
DESIGN OF MATERIALS THEMES
A recurring theme of several presentations in this session was the need for a national fundamental materials database. Dr. Olson opened his presentation by restating one of the goals of the Materials Genome Initiative, which is to expand the database of materials’ properties.1 He also detailed a recommendation from the 2004 NRC study Accelerating Technology Transition: Bridging the Valley of Death for Materials and Processes, advocating the construction of a fundamental database for materials and suggesting that the White House Office of Science and Technology Policy (OSTP) lead the effort. The idea would be to promote the science-based engineering of materials instead of empirically based engineering, which was common then (and still largely is). While implementation of this recommendation began with formation of the Materials Genome Initiative, Dr. Olson reiterated his belief that focused support is necessary to make significant progress. Other participants asked what was needed to make significant progress in federal support for fundamental materials databases, as six NRC reports have discussed this need in the last 15 years. Dr. Olson responded that the database project is taking place but that additional force and energy must be brought to bear.
In his presentation Dr. Olson also described a vision for design and databases. One area of opportunity, he said, is a large-scale public database system, which could reside at the National Institute of Standards and Technology (NIST). He recommended that such a database consist of protodata below the assessed genome level: raw information on experimentally measured phase relations, thermochemi-
cal measurements, and DFT calculations.2 The data could be pooled and periodically assessed to generate usable databases for public and private development channels. The website http://www.tms.org, which is maintained by The Minerals, Metals, and Materials Society, for example, stores 3D microstructural information. The standard database for materials selection is the set of discrete properties of materials searchable by the Ashby materials selection system. The opportunity exists to use the same architecture to search all databases at all levels, even to select atoms and components the same way as higher-level systems. Another opportunity is to hand off the material plus the associated information system for adapting to future manufacturing/field experience; this would allow moving beyond discrete properties into a system based on microstructural state variables.
Krishna Rajan also discussed materials databases in his presentation and said that his goal is to retrieve such data from their repositories and transmit them to laboratories where users can make new and unexpected discoveries.
Balancing Experiment-Based and Modeling-Based Design Approaches
A second recurring theme during this session was the appropriate balance between experimentation and modeling to develop new materials. During the open discussion period, Haydn Wadley, a committee member, framed this balance by asking the group to consider how they would design a new material, first, purely through experiment, and second, purely through modeling, and then to conduct this design process 30 years ago, in the present, and in the future. Respondents generally agreed that experiments would always be necessary for a material to be accepted for use, even a material designed purely by modeling. However, the future is likely to see materials design driven by models, followed by experiments to validate the new product. Participants also noted that experiments do not quantify uncertainty well; for that, computational tools are necessary.
Dr. Olson was asked during his presentation what types of models should be used—data-driven or physics-based. He responded that the most successful projects use a mechanistic, modeling-based approach rather than a data-driven one. Data-driven models tend to be superficial and to have too much empirical data to provide useful solutions. Modeling-based approaches can also point the way to experiments and lead to new discoveries.
Dr. Rajan also discussed the balance between experimentation and modeling, though he framed it as part of the strategy of informatics. Informatics integrates information of all types (including experiments, modeling, data, and theory) in an environment where there is no preexisting model. Informatics in this context is the computational strategy to integrate information associated with a mate-
2 Density functional theory (DFT) is a method for modeling the chemical bonding between atoms.
rial’s structure, chemistry, and performance to extract patterns for its behavior at multiple scales.
During the open discussion period, Dr. Wadley asked participants for challenge problems—important capabilities that are currently out of reach—in this field. Several people spoke of instances in which fabrication produces a material with predictive approaches, and they recommended that data should be kept on such failures. Others pointed out that the Materials Genome Initiative may result in new experimental protocols that will generate new data, rapidly expanding fundamental materials databases.
There was some discussion of the DFT as the basis for calculation of phase diagrams (CALPHAD) modeling. In the open discussion period, some workshop participants questioned the resolution of the DFT calculation, which is about 0.1 eV at an ambient temperature of about 0.05 eV.3 Others defended the use of the DFT, pointing out that it was not inhibiting the design process and that the relative uncertainties were already well known. During the question-and-answer period following Dr. Olson’s presentation, a participant asked about the accuracy of the DFT as well, and Dr. Olson replied that in surface thermodynamics, techniques are almost always DFT-based, and the accuracy is very good for grain boundary cohesion.
One of the goals of the design of materials approach is to reduce the amount of time needed to qualify materials. In his presentation, Dr. Olson described the qualification process for S53 and M54, the two landing gear steels QuesTek has developed. It took 8 years from material design through qualification for S53, and QuesTek is on track to get M54 qualified within 5 years. This computation-based qualification is much faster than a standard 10- to 20-year development cycle. Dr. Olson noted that the accelerated qualification of the stainless steel allowed QuesTek to share necessary inventory costs with the supplier because there was shared confidence in model accuracy.
3 One electron volt is a unit of energy equal to approximately 1.6 × 10–19 joule. In this case, it is customary to express the temperature in electron volts.