20 23 31 P. 13 APRIL 2023 | VOL 181 | NO 3 Invention of Ultrasonic AM for Unique Applications Review of AM Processes for Ceramics and Composites iTSSe Newsletter Included in This Issue MACHINE LEARNING AIDS MECHANICAL PROPERTY ALLOWABLES DEVELOPMENT ADDITIVE MANUFACTURING

SAVE THE DATE SEPTEMBER 12–13, 2023 | UNIVERSITY WEST, SWEDEN At TS4E 2023, you will learn about: • Innovative solutions to improve coating performance in the aerospace, energy generation, and transportation industries • How S&STS technologies can replace more expensive coating processes • Latest developments and trends in EBCs and CMCs for next-generation gas turbines • Top-notch speakers from world-renowned institutions (industry, academia, and government) • Enhanced learning & networking via 30-min academic/industrial presentations, table-tops with coˆee breaks, networking dinners, roundtable discussions, and 3-min elevator pitches Thermal Spray of Suspensions & Solutions Symposium + EBCs Registration opens spring 2023. Visit ts4eevent.org to learn more. Special Reduced Registration Fees! Attendee space is LIMITED to 150 people, be sure to register early! Due to the generous support of our host, co-organizers, and sponsors, the registration fees for TS4E are extremely a‹ordable. ASM Thermal Spray Society Member registration starts at $250 and will include a two-day symposium, two lunches, refreshment breaks, and two networking dinners. Back by Popular Demand! The ASM Thermal Spray Society will once again oˆer a symposium focused on suspension and solution thermal spray (S&STS) technology. This symposium is a chance for scientists and engineers interested in emerging S&STS technologies to address both research challenges and the development of industrial applications. In addition, the topics of environmental barrier coatings (EBCs) and ceramic matrix composites (CMCs) will be covered. Organized by:

20 23 31 P. 13 APRIL 2023 | VOL 181 | NO 3 Invention of Ultrasonic AM for Unique Applications Review of AM Processes for Ceramics and Composites iTSSe Newsletter Included in This Issue MACHINE LEARNING AIDS MECHANICAL PROPERTY ALLOWABLES DEVELOPMENT ADDITIVE MANUFACTURING

2023 INTERNATIONAL MATERIALS, APPLICATIONS & TECHNOLOGIES OCTOBER 16–19, 2023 | HUNTINGTON PLACE | DETROIT, MICHIGAN ADVANCED MATERIALS AND MANUFACTURING TECHNOLOGIES Here’s what you can expect: • 5000+ attendees • Over 700 technical presentations, keynotes, and panel discussions • More than 500 exhibits • 4 days of technical programming • 2.5 days of exposition • Education courses and workshops • Networking events • Sessions organized by the IDEA Committee • Programming and activities for emerging professionals REGISTRATION OPENS MAY 2023 ORGANIZED BY: PARTNERED WITH: CO-LOCATED WITH: Visit imatevent.org to learn more. IMAT 2023, co-located with Heat Treat and Motion + Power Technology Expo, unites diŠerent market segments that span the entire materials community and connects industry, academia, and government to solve global materials challenges. Core programming from all six of ASM International’s aŠiliate societies will serve as the backbone of IMAT’s technical sessions. IMAT will bring together all the global expertise you need to tackle modern materials challenges, address environmental issues, and incorporate the latest digital technologies.

47 ASM NEWS Highlights of the ASM Strategic Plan as well as the latest news about ASM members, chapters, events, awards, conferences, affiliates, and other Society activities. MACHINE LEARNING: PROGRESS TOWARD ADDITIVE MANUFACTURING MATERIALS PROPERTY ALLOWABLES DEVELOPMENT Annie Wang, Zach Simkin, and William E. Frazier Results from two research projects show machine learning to be a cost effective and flexible way to accelerate the process of mechanical property allowables development. 13 A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 2 Senvol recently led an additive manufacturing program for the U.S. Army. Image of one of the builds, done by EWI, using 17-4 PH stainless steel on an EOS M290 machine. Courtesy of Annie Wang, Olga Elsieeva, and William E. Frazier, FASM. On the Cover: 56 3D PRINTSHOP Unique materials designed for 3D printing are described in this issue. 10 MACHINE LEARNING Scientists are using machine learning to automate x-ray diffraction analysis and to extract data from videos of their experiments.

4 Editorial 5 Research Tracks 5 Feedback 10 Machine Learning 6 Metals/Polymers/Ceramics 8 Testing/Characterization 11 Process Technology 12 Emerging Technology 55 Editorial Preview 55 Special Advertising Section 55 Advertisers Index 56 3D PrintShop TRENDS INDUSTRY NEWS DEPARTMENTS Check out the Digital Edition online at asminternational.org/news/magazines/am-p ASM International serves materials professionals, nontechnical personnel, and managers wordwide by providing high-quality materials information, education and training, networking opportunities, and professional development resources in cost-effective and user-friendly formats. ASM is where materials users, producers, and manufacturers converge to do business. Advanced Materials & Processes (ISSN 0882-7958, USPS 762080) publishes eight issues per year: January/February, March, April, May/June, July/August, September, October, and November/December, by ASM International, 9639 Kinsman Road, Materials Park, OH 44073-0002; tel: 440.338.5151; fax: 440.338.4634. Periodicals postage paid at Novelty, Ohio, and additional mailing offices. Vol. 181, No. 3, APRIL 2023. Copyright © 2023 by ASM International®. All rights reserved. Distributed at no charge to ASM members in the United States, Canada, and Mexico. International members can pay a $30 per year surcharge to receive printed issues. Subscriptions: $499. Single copies: $54. POSTMASTER: Send 3579 forms to ASM International, Materials Park, OH 44073-0002. Change of address: Request for change should include old address of the subscriber. Missing numbers due to “change of address” cannot be replaced. Claims for nondelivery must be made within 60 days of issue. Canada Post Publications Mail Agreement No. 40732105. Return undeliverable Canadian addresses to: 700 Dowd Ave., Elizabeth, NJ 07201. Printed by LSC Communications, Lebanon Junction, Ky. 20 TECHNICAL SPOTLIGHT ULTRASONIC ADDITIVE MANUFACTURING: DEVELOPMENT AND COMMERCIALIZATION A unique method of additive manufacturing using ultrasonic energy excels at mixed metal applications, earning it the 2022 ASM Engineering Materials Achievement Award. 23 A REVIEW OF ADDITIVE MANUFACTURING PROCESSES FOR FABRICATING CERAMICS AND COMPOSITES Surojit Gupta, Daniel Trieff, Mackenzie Short, Maharshi Dey, Samuel J.A. Hocker, and Valerie Wiesner Knowing which additive manufacturing techniques are advantageous for specific manufacturing applications leads to better results. 28 PROFILES IN MATERIALS SCIENCE OSAKA UNIVERSITY’S ANISOTROPIC DESIGN & AM RESEARCH CENTER Learn about a unique research hub in this fascinating interview with center director Nakano Takayoshi and Hideyuki Kanematsu. FEATURES APRIL 2023 | VOL 181 | NO 3 ADVANCED MATERIALS & PROCESSES | APRIL 2023 3 20 28 31 23 31 iTSSe: INCLUDES ITSC SHOW PREVIEW The official newsletter of the ASM Thermal Spray Society (TSS). This timely supplement focuses on thermal spray and related surface engineering technologies along with TSS news and initiatives.

4 A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 ASM International 9639 Kinsman Road, Materials Park, OH 44073 Tel: 440.338.5151 • Fax: 440.338.4634 Joanne Miller, Editor joanne.miller@asminternational.org Victoria Burt, Managing Editor vicki.burt@asminternational.org Frances Richards and Corinne Richards Contributing Editors Anne Vidmar, Layout and Design Allison Freeman, Production Manager allie.freeman@asminternational.org Press Release Editor magazines@asminternational.org EDITORIAL COMMITTEE Adam Farrow, Chair, Los Alamos National Lab John Shingledecker, Vice Chair, EPRI Somuri Prasad, Past Chair, Sandia National Lab Beth Armstrong, Oak Ridge National Lab Margaret Flury, Medtronic Surojit Gupta, University of North Dakota Nia Harrison, Ford Motor Company Michael Hoerner, KnightHawk Engineering Hideyuki Kanematsu, Suzuka National College of Technology Ibrahim Karaman, Texas A&M University Ricardo Komai, Tesla Bhargavi Mummareddy, Dimensional Energy Scott Olig, U.S. Naval Research Lab Christian Paglia, SUPSI Institute of Materials and Construction Amit Pandey, Lockheed Martin Space Satyam Sahay, John Deere Technology Center India Kumar Sridharan, University of Wisconsin Jean-Paul Vega, Siemens Energy Vasisht Venkatesh, Pratt & Whitney ASMBOARDOF TRUSTEES David B. Williams, President and Chair Pradeep Goyal, Senior Vice President Navin Manjooran, Vice President Judith A. Todd, Immediate Past President John C. Kuli, Treasurer Burak Akyuz Amber Black Ann Bolcavage Pierpaolo Carlone Elizabeth Homan Toni Marechaux André McDonald U. Kamachi Mudali James E. Saal Sandra W. Robert, Executive Director STUDENT BOARDMEMBERS Jaime Berez, Ashlie Hamilton, Nicole Hudak Individual readers of Advanced Materials & Processes may, without charge, make single copies of pages therefrom for personal or archival use, or may freely make such copies in such numbers as are deemed useful for educational or research purposes and are not for sale or resale. Permission is granted to cite or quote fromarticles herein, provided customary acknowledgment of the authors and source is made. The acceptance and publication of manuscripts in Advanced Materials & Processes does not imply that the reviewers, editors, or publisher accept, approve, or endorse the data, opinions, and conclusions of the authors. ACCELERATING INNOVATION How do you accelerate innovation? This question was posed directly and indirectly throughout the AeroMat conference and the co-located SMST Entrepreneurial Workshop in Fort Worth, Texas, in March. For many, the answer was additive manufacturing (AM). With four AeroMat sessions on AM, it is clearly a vital and sustainable path forward for the air and space industries. In one session, Kevin Stonaker from the Federal Aviation Administration presented the perfect complement to the machine learning (ML) article in this issue. The article illustrates the use of Metallic Materials Properties Development and Standardization (MMPDS) as an important resource for determining materials allowables in AM processes. At AeroMat, Stonaker announced the upcoming release of MMPDS Volume 2, slated for July. This early release will include general information and guidelines for gathering datasets. Subsequent releases will include materials data tables for alloys of steel, aluminum, magnesium, and titaniumusing the submitted datasets. Timothy Bunning of the Air Force Research Laboratory (AFRL) was an AeroMat keynote presenter. He mentioned innovations that AFRL was involved with from the beginning, e.g., rare earth magnets, composites, and specialty coatings. His lab refers to these types of developments as having “AFRL Inside.” Going forward, the Materials & Manufacturing Directorate he leads aims to discover materials of the future. While materials are critical, many speakers emphasized processes as pivotal to engineering advances. Jeffrey “Scotty” Sparks, retired from NASA Marshall Space Flight Center, explained how the NASA team regrouped and reprioritized after the Columbia disaster. They didn’t changematerials; they changed processes. Processes also came up in a panel session with space experts from Blue Origin, Northrop Grumman, Constellium, Boeing, and ArianeGroup. The panel was asked if space solutions can be adapted for aviation. For example, are there lessons learned about hydrogen storage that can be applied to air flight? Yes. A key takeaway was the need for new nondestructive testing techniques because aircraft energy storage, defueling, and restorage must endure considerably more cycles than what is required for one space launch. It was also noted that some of this research and testing necessitates an appetite for development and risk tolerance. A whole room full of risk tolerant, inventive attendees were at the SMST Entrepreneurial Workshop hosted by the International Organization on Shape Memory and Superelastic Technologies. The meeting was energetic with frequent interaction between the speakers and participants. Tom Duerig, FASM (pictured left), an SMST founder and early co-inventor of Nitinol usage for stents and eyeglass frames, gave a talk on the “Six Ways to Fail as an Entrepreneur.” One easy way is to run out of cash! Some factors that led to success were having a great culture, a beloved technology, and a cohesive team. At another session, one astute attendee commented that, by design, all engineering teams innovate. That’s what they do. Wherever technological innovations are taking off—whether at a Fortune 500 company or a startup— there are engineers inside. joanne.miller@asminternational.org A few SMST panelists during a session break.

A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 5 SMART SUPERCAPACITOR SETS RECORDS A team of scientists from Clemson University, South Carolina, and the Indian Institute of Science designed a smart supercapacitor using an innovative stack of metal oxides — vanadium pentoxide and zinc oxide — that can efficiently harvest energy from sunlight and simultaneously store it. The research could eventually lead to self-charging consumer electronics such as health monitoring devices. Zinc oxide has been widely used in light-dependent charge transfer appli- cations such as photovoltaics due to its low cost, high carrier mobility, long carrier diffusion, ease of synthesis on substrates, and nontoxicity. It features electron transporting and hole- blocking characteristics and readily generates photoexcited electrons upon light irradiation. However, zinc oxide has a wide band gap that restricts its utility to a narrow light spectrum, requiring a creative strategy to improve its performance. The researchers stacked vanadium pentoxide and zinc oxide to create a unique heterostructure, one that improved on the ability of traditional materials to convert light into electrical energy. In a test to measure its ability to store harvested light, the new device beat the previous record by a factor of RESEARCH TRACKS / FEEDBACK A new supercapacitor made of vanadium pentoxide and zinc oxide harvests and stores energy from sunlight. four. “We’ve come up with a two-in-one device that not only harvests light more efficiently but also stores it as electrical energy, unlike the other systems that exist right now,” says Clemson post-doctoral fellow Mihir Parekh. “The materials we chose allowed us to engineer the band gap so that the light-to-electricity conversion was very efficient.” The team found that the device showed excellent electrochemical performance and stability for over 5000 cycles. clemson.edu. MIXED MOLECULES BUILD BETTER BATTERIES Researchers from Florida State University (FSU) and Lawrence Berkeley National Laboratory developed a new strategy to build solid-state batteries that are less reliant on specific elements, especially expensivemetals with supply chain issues. The team demonstrated that a mix of various solid-state molecules could result in a more conductive battery that was less dependent on a large quantity of one individual element. For example, instead of creating a battery made with germanium, which rarely appears naturally in high CELEBRATING WOMEN AVIATORS The editorial about women in air and space (AM&P, March 2023) reminded me of an audiobook I enjoyed several years ago, “Fly Girls: How Five Daring Women Defied All Odds and Made Aviation History” by Keith O’Brien. It covered many interesting topics, including the fact that Amelia Earhart was once an aeronautics advisor at Purdue University. The Museum of Flight in Seattle also includes some interesting history about women pilots and flight attendants and is well worth a visit. Patrick Mizik, P.E. Haldex Brake Products Corp. FEEDBACK We welcome all comments and suggestions. Send letters to joanne.miller@asminternational.org. Scanning transmission electron microscope images reveal the elemental distribution in a “disordered” solid electrolyte. Top row: titanium (Ti), zirconium (Zr), and tin (Sn); bottom row: hafnium (Hf), phosphorus (P), and oxygen (O); scale bar 50 nanometers. Courtesy of Yan Zeng and Gerd Ceder/ Berkeley Lab. concentrations, the researchers created a mixture of titanium, zirconium, tin, and hafnium. “There’s no hero element here,” explains FSU scientist Bin Ouyang. “It’s a collective of diverse elements that make things work. What we found is that we can get this highly conductive material as long as different elements can assemble in a way that atoms can move around quickly. And there are many situations that can lead to these so-called atom diffusion highways, regardless of which elements it may contain.” fsu.edu.

A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 6 METALS | POLYMERS | CERAMICS Scientists at the University of Bristol, U.K., discovered a chromium-cobalt-nickel alloy that exhibits increased strength at temperatures as cold as -250°C, making it the toughest material on record. The alloy’s behavior is due to a phase transformation that combines with other nanoscale mechanisms to prevent crack formation and propagation. www.bristol.ac.uk. Set to debut this fall, Colorado School of Mines will offer a bachelor of science in ceramic engineering. In the U.S., only two other such programs exist, one at Alfred University in New York and another at Missouri University of Science & Technology. Students will receive hands-on training in ceramic processing, sintering, glass science, and thermal, mechanical, and electrical properties. mines.edu. BRIEFS on single crystals over the last three decades, but growing the crystals with melt processing and controlling their orientations has been quite challenging,” explains Karaman. “The method Hande discovered now saves us a lot of time and provides more flexibility.” Controlling the size, shape, and crystallographic orientation of single crystals is vital to exploit the desired properties, the researchers say. Single crystals are essential to microelectronics, optical crystals, magnetic devices, solar cells, piezoelectric components, and multifunctional alloys. Another advantage to their new technique, according to the researchers, is that it does not require complex and expensive equipment. The newmethod is called the solid-state crystal growth (SSCG) technique, where large bulk crystals with different crystallographic orientations could be made with simple heat treatments. The Texas A&M research team demonstrated the SSCG method in two alloy systems—FeMnAlNi CARBON-REDUCING SUPERALLOY In a collaborative effort between Sandia National Laboratory, Iowa State University, Ames National Laboratory, and Bruker Corp., Billerica, Mass., scientists used a 3D printer to create a superalloy that could help power plants generate more electricity while producing less carbon The superalloy has an unusual composition that makes it stronger and lighter than state-of-the-art materials currently used in gas turbine machinery. The findings could have broad impacts across the energy sector as well as the aerospace and automotive industries, and hints at a new class of similar alloys waiting to be discovered. Sandia’s experiments showed that the new superalloy—42% aluminum, 25% titanium, 13% niobium, 8% zirconium, 8% molybdenum, and 4% tantalum— was stronger at 800°C than many other high-performance alloys and still stronger when it was brought back down to room temperature. The team at Sandia used a 3D printer to quickly melt together powdered metals and then immediately print a sample of it. Sandia’s creation also represents a fundamental shift in alloy development because no single metal makes up more than half the material. By comparison, steel is about 98% iron combined with carbon, among other elements. Moving forward, the team is interested in exploring whether advanced computer modeling techniques could help researchers discover more members of what could be a new class of high-performance, additive manufacturing-forward superalloys. sandia.gov, iastate.edu, bruker.com. CRYSTAL GROWTH TRANSFORMS MATERIALS Researchers at Texas A&M University, College Station, discovered a new crystal growth and orientation control method in solid-state—without melt processing. Materials scientists Hande Ozcan and IbrahimKaraman, FASM, led the research, which focused on growing large single crystals and their ability to change crystallographic orientation. “We have been working Sandia technologist Levi Van Bastian works to print material on the laser engineered net shaping machine, used for 3D printing superalloys. Courtesy of Craig Fritz. Hande Ozcan working with an x-ray machine. Courtesy of Texas A&M Engineering.

A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 7 and CuMnAl—and achieved repeated, massive orientation changes in the solid state. These findings offer a new strategy for manipulating the orientation of large single crystals on demand to take advantage of their superior and highly anisotropic properties. The team believes the discovery will open up extensive research areas and that this is just the beginning of this exciting new path for finding new materials. tamu.edu. METALLIC POLYMER BREAKTHROUGH A team of scientists from Fudan University and East China University of Science and Technology, both in Shanghai, successfully synthesized a stable polymer with a backbone made of nickel atoms. Because of the different electronic structures of metal and nonmetal atoms, it’s challenging to confer the properties of metals—such as high thermal and electrical conductivity—into polymers. Polymers with a metal backbone could combine the advantages of both types of material and open routes to materials with novel functionality. The team used a chalice-shaped molecule, calixarene, with four binding sites as the scaffold for the metal polymer. Next, they attached four polyaminopyridine chains to the calixarenes, bundling the four chains and aligning them in parallel form. Synthesis of the chains can be carried out stepwise from individual building blocks, or several larger blocks can be linked together. The team demonstrated their polymer synthesis method and produced a nickel backbone with precisely controlled length. They created versions containing three to 21 nickel atoms. Notably, the distance between nickel atoms decreases as the chain length increases, strengthening the Ni–Ni bonds. The new materials may conduct electricity, are thermally stable, and can be processed in solution. They demonstrate strong length-dependent light absorption with narrow band gaps, which is promising for opto- electronic devices and semiconductors. The new synthetic strategy could also be expanded to other transition metals such as copper and cobalt. www.fudan. edu.cn/en, www.ecust.edu.cn/en. Researchers discovered a polymer with a metallic backbone that is conductive. Courtesy of Angewandte Chemie International Edition (2023). DOI: 10.1002/anie.202216060.

8 A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 load shuffling, that could enable the design of better-performing lightweight materials for vehicles. Researchers monitored a version of ORNL’s ACMZ— aluminum, copper, manganese, and zirconium—alloy for deformation that occurs when the material is under persistent mechanical stress at high temperatures. Using neutron diffraction, researchers studied the material’s atomic structure and observed that the overall stress was absorbed by one part of the alloy but transferred to another part during deformation. This backand-forth shuffling prevents strengthening in some areas. “Neutrons offer opportunities to study metallurgical phenomena in multiphase structural materials,” ORNL’s Amit Shyam says. “We’ve gained unprecedented insight into elevated-temperature material behavior that will allow us to design improved aluminum alloys for extreme conditions.” ornl.gov. TESTING | CHARACTERIZATION X-RAY INNOVATION A team of researchers at Arizona State University, Tempe, launched a new compact x-ray light source (CXLS), which will enable scientists to see deeper into matter and living things. The device is the first stage of a larger compact x-ray free electron laser project that aims to build two instruments including a coherent x-ray laser. The CXLS generates a high-flux beam of hard x-rays, with wavelengths short enough to resolve the atomic structure of complex molecules, and pulsed at extremely short durations of a few hundred femtoseconds. Such capabilities have so far only been available at large x-ray free-electron laser facilities, whose construction costs run into the billions. The ASU device provides a uniquely compact facility for ultrashort x-rays that fits into the size of a basement, making leading-edge x-ray technology accessible to a university campus. The CXLS will be available to scientists from all over the U.S. and serve as a training ground for ASU students. “I’m most intri- gued by what lies at the edge of our knowledge, pursuing phenomena that have never been observed before,” says lab director Robert Kaindl. “With the conclusion of the compact x-ray light source commissioning, our focus will shift to early experiments with its ultrashort x-rays and the transition to a user facility.” The first set of experiments will begin later in 2023. According to lead scientist William Graves, “The CXLS will be a boon to awide range of imaginative scientists working to unlock the secrets of biology, chemistry, and physics.” asu.edu. NEUTRONS REVEAL ALLOY BEHAVIOR Scientists at the DOE’s Oak Ridge NationalLaboratory,Tenn.,discovereda mechanism in a 3D-printed alloy, called Triangular holes make this material more likely to crack from left to right. Courtesy of N.R. Brodnik et al./Phys. Rev. Lett. Scientists at the University of California, Irvine and the DOE’s Brookhaven National Laboratory are exploring high-nickel-content layered cathodes, which hold promise for future batteries. The team used a transmission electron microscope and atomistic simulations to learn how oxidation phase transitions impact lithium-ion battery (LIB) materials. Knowing how these batteries operate at the atomic scale could enable development of cobalt-free LIBs with vastly improved power and life cycles. uci.edu. William Graves stands next to one of the magnets used in the CXLS instrument. Courtesy of ASU. BRIEF Researchers used neutron diffraction experiments to study the 3D-printed ACMZ alloy and observed a phenomenon called “load shuffling.” Courtesy of ORNL, U.S. Dept. of Energy.

A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 9 TAKING NANOMATERIALS TO THE LIMIT Researchers at Florida State University’s High-Performance Materials Institute in Tallahassee are conducting research on the thermal limits of advanced nanomaterials. Their work has the potential to enhance medicine delivery systems, electronics, space travel, and other applications. The team completed the first-ever study on how purified boron nitride nanotubes (BNNTs) remain stable in extreme temperatures in inert environments. BNNTs are measured by the nanometer and are stronger and more resistant to high temperatures than carbon nanotubes. However, manufacturing these materials is challenging, and current methods for BNNTs do not yet yield the same quantities as those used to produce carbon nanotubes. The researchers found that BNNTs are fully stable at up to 1800°C in an inert environment and can withstand temperatures of 2200°C for short periods without losing their mechanical properties. Potential applications for these light, strong composite materials are plentiful. BNNTs show particular promise for use in space exploration due to their ability to conduct heat, insulate electrical current, and block radiation. They could be used in space rovers or spacecraft during reentry into Earth’s atmosphere. Additionally, their properties make them useful for high- performance electronics. “Understanding the behavior of these nanotubes at high temperatures is crucial for creating materials that can withstand extreme conditions, both in manufacturing and in their final use,” says lead researcher Mehul Tank. “As we understand better how they function in these conditions, we’ll be able to develop better manufacturing of composites that employ high-temperature processing matrices, like ceramics and metals.” fsu.edu. Boron nitride nanotube material in a crucible for heating at Florida State University’s High-Performance Materials Institute. Courtesy of Mark Wallheiser/ FAMU-FSU Engineering. LE-237i oct- 2023.ps T:\MISC\ADS\LE-237\LE-237i oct- 2023.cdr Wednesday, March 8, 2023 3:59:40 PM Color profile: Disabled

A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 1 0 MACHINE LEARNING | AI MACHINE LEARNING AUTOMATES X-RAY DIFFRACTION ANALYSIS Researchers at the National Institute forMaterials Science (NIMS), Japan, have automated the process of ana- lyzing results from x-ray diffraction studies on crystalline materials. The team developed a robotic process automation (RPA) system that uses machine learning to perform Rietveld analysis automatically, reducing human costs and speeding up data analysis. The new RPA system is run on a personal computer and can be combined with numerous graphical user interface applications used to calculate a material’s properties, control experimental equipment, or analyze material data, creating a closed-loop system to design and analyze materials with minimal human intervention. The scientists verified the accuracy of their procedure by analyzing samples of powdered compounds with known crystal structures. They explain that the ability to determine the structures from powdered samples is a strength of Rietveld analysis, avoiding the need to grow large single crystals, which can be difficult for some materials. The team now plans to refine the procedure for more complex crystal structures and explore the use of machine learning RPA strategies for general applications in materials science. Some of the possibilities include various simulation methods used to calculate material properties, as well as applications for controlling experimental equipment. www.nims.go.jp. AI TRANSFORMS NUCLEAR MATERIALS RESEARCH At the DOE’s Argonne National Laboratory, Lemont, Ill., scientist WeiYing Chen in the nuclear materials group is using artificial intelligence (AI) to revolutionize the way scientists analyze videos of their experiments. Chen is using a deep learning-based multi-object tracking (MOT) algorithm to extract data from videos of experiments in an effort to help the U.S. improve advanced nuclear reactor designs. The experiments are being conducted at the Intermediate Voltage Electron Microscope (IVEM) facility, a transmission electron microscope with ion beam accelerator capabilities. Re- sults from the MOT algorithm are helping scientists study the effect of ion irradiation on materials and defects within a picosecond timeframe, Graphical abstract. Courtesy of Sci. Technol. Adv. Mat., 2022. Wei-Ying Chen uses computer vision to collect data about material defects and structural voids in the same way facial recognition software looks for unique faces. which is too fast for manual tracking. With computer vision tools, Chen can extract data from every video frame, which will enable scientists to discover materials that are resistant to higher temperatures and irradiation doses. Traditional methods of capturing data during experiments can only provide a limited number of data points, but the MOT algorithm can track every defect and structural void. This is enabling Chen to build a reliable collection of information about material properties such as temperature resistance, irradiation resilience, microstructural defects, and material lifetimes. As a result, this enhanced data can lead to better models and experiments, allowing scientists to design advanced reactors that can run at higher tem- peratures and generate more clean electricity. Using AI and computer vision also provides a more efficient use of time, as scientists can make on-the-spot adjustments to their experiments and capture important information. “Computer vision can provide information that, from a practical standpoint, was unavailable before,” says Chen. “It’s exciting that we now haveaccess to somuchmore rawdataof unprecedented statistical significance and consistency.” anl.gov.

A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 1 1 PROCESS TECHNOLOGY SUPERFAST MANUFACTURING OF THERMOELECTRIC DEVICES University of Notre Dame, Indiana, researchers developed a new superfast way to create high-performance, energy-saving thermoelectric devices. The novel process uses intense pulsed light to sinter thermoelectric material in less than a second. The team sped up this method of turning nanoparticle inks into flexible devices by using machine learning to determine the optimum conditions for the ultrafast but complex sintering process. According to the researchers, flexible thermoelectric devices offer great opportunities for direct conversion of waste heat into electricity as well as solid-state refrigeration. These products have additional benefits as power sources and cooling devices—they don’t emit greenhouse gases, and they are both durable and quiet since they don’t have moving parts. storage, microelectronics, and nano- lithography. Much more versatility is possible by gaining precise control over the sequence in which different monomer molecules combine to form blocks, which then link together further on their own. Gaining fine control over the formation of these materials, known as block polymers, generally requires complicated cycles of chemical reactions. The achievement is an example of a one-pot, one-step reaction, because it involves simply adding the required monomers to a single reaction vessel and using chemistry to control the assembly of the monomers into blocks, and then into a block polymer. A crucial key to controlling how the monomers react is the use of an alkali metal carboxylate to switch the polymer-building processes between different forms of reaction. www.global.hokudai.ac.jp. Despite their potentially broad impact in energy and environmental sustainability, thermoelectric devices have not achieved large-scale application due to cost-prohibitive manufacturing processes. Now, machine-learning-assisted ultrafast flash sintering will make it possible to produce high-performance, eco-friendly devices much faster and at a far lower cost. “The results can be applied to powering everything from wearable personal devices to sensors and electronics, to industry Internet of Things,” notes researcher Yanliang Zhang. nd.edu. SIMPLIFIED POLYMER SYNTHESIS Researchers at Hokkaido University, Japan, are simplifying the production of complex polymers with precisely controlled structures. Using a new one-pot, one-step synthesis procedure, the new process brings a sought-after level of control to making long and geometrically interlinked polymer molecules from several alternating molecular units joined in a controlled sequence. The method could open new avenues for producing a wide range of advanced materials, with applications in many fields, including drug delivery, data BRIEF Wall Colmonoy, Madison Heights, Mich., will invest $2.5 million to modernize its alloy products plant in Los Lunas, N.M. The plan includes upgrades to plant infrastructure, alloy furnace equipment, and the R&D laboratory. wallcolmonoy.com. Artistic rendering of high-performance thermoelectric devices for energy harvesting and cooling. Courtesy of University of Notre Dame. A wide variety of synthetic polymers with diverse structures can be constructed using newmethods. Courtesy of Xiaochao Xia, et al., Journal of the American Chemical Society, 2022.

A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 1 2 MAKING FREESTANDING MEMBRANES FOR SMART MATERIALS Anewmethod formaking thin films of perovskite oxide semiconductors was created by a team of scientists and engineers from the University of Minnesota Twin Cities, Minn. The discovery will allow researchers to harness the properties of a class of smart materials and even combine them with other emerging nanoscale materials to make better devices such as sensors, smart textiles, and flexible electronics. Producing materials in thin-film form makes them easier to integrate into smaller components for electronic devices. Most thin films created via the conventional epitaxy method are stuck on their host substrate, limiting their uses. If the thin film is detached from the substrate to become a freestanding membrane, it becomes much more functional. And now, researchers have found a new way to successfully create a membrane of a particular metal oxide—strontium titanate—circumventing several issues that have plagued the synthesis of freestanding metal oxide films in the past. “We have created a process where we can make a freestanding membrane of virtually any oxide material, exfoliate it, and then transfer it onto any subject of interest we want,” says Professor Bharat Jalan. “Now, we can benefit from the functionality of these materials by combining them with other nano- scale materials, which would enable a wide range of highly functional, highly efficient devices.” umn.edu. NEW MATERIALS DISCOVERY METHOD Researchers at the DOE’s Argonne National Laboratory, Northwestern University, and the University of Chicago, all based in Illinois, developed a new method for discovering and making new crystalline materials with two or more elements. Their process yielded 30 previously unknown compounds, and ten of them have structures never seen before. The team’s invention method starts with a solution made of two components—one is a highly effective solvent that dissolves and reacts with any solids added to the solution; and the other, a less-effective solvent, tunes the reaction to produce a new solid upon addition of different elements. This tuning involves changing the ratio of the two components and the temperature, in this case ranging between 750-1300°F. EMERGING TECHNOLOGY Cornell University, Ithaca, N.Y., is leading a new $34 million research center that aims to accelerate the creation of energy-efficient semiconductor materials and technologies, and develop new approaches for microelectronics systems. The SUPeRior Energy-efficient Materials and dEvices (SUPREME) Center will bring together researchers from 14 higher education institutions, in collaboration with the center’s sponsor, Semiconductor Research Corporation. cornell.edu. BRIEF “We are not concerned with making known materials better but with discovering materials no one knew about, or theorists imagined even existed,” says Northwestern Professor Mercouri Kanatzidis. “We expect that our work will prove extremely valuable to the chemistry, materials, and condensed matter communities for synthesizing new and currently unpredictable materials with exotic properties.” This is only the beginning of what is possible since the method can be applied toalmost any crystalline solid. It can also be applied to producing many different crystal structures, including multiple stacked layers, a single layer an atom thick, and chains of molecules that are not linked. Such unusual structures have varying properties and are key to developing next-generation materials applicable to not only superconductors, but also microelectronics, batteries, magnets, and more. anl.gov. Reaction pathway from simple precursor to complex structure. The result is a layered structure with five elements— sodium, barium, oxygen, copper, and sulfur. Courtesy of Argonne National Laboratory. Professor Bharat Jalan’s team developed a newmethod for making nano-membranes of “smart” materials. Courtesy of Olivia Hultgren/UMN.

1 3 A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 MACHINE LEARNING: PROGRESS TOWARD ADDITIVE MANUFACTURING MATERIALS PROPERTY ALLOWABLES DEVELOPMENT Annie Wang and Zach Simkin Senvol LLC, New York William E. Frazier, FASM* Pilgrim Consulting LLC, Lusby, Maryland Results from two research projects show machine learning to be a cost e ective and flexible way to accelerate the process of mechanical property allowables development. *Member of ASM International A D D I T I V E M A N U F A C T U R I N G 1

1 4 A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 Fig. 1 — Machine learning technology stack. The application of machine learning (ML) in the field of materials science and engineering has rapidly matured over the past decade. However, the full potential of this methodology has yet to be unleashed. This article starts with a succinct synopsis of ML and explores its many diverse characteristics. The results of two recently completed research projects investigating the potential use of ML to establish additive manufacturing (AM) materials property allowables are described. Although continued research and development (R&D) work is required, the results are very promising. The authors’ thoughts on the maturation of ML for this application are then delineated. BACKGROUND The technologies involved in ML are illustrated in Fig. 1 and may be conveniently divided into seven broad areas: (i) data, (ii) ML categories, (iii) environment/infrastructure, (iv) data science and ML libraries, (v) algorithms, (vi) quality, and (vii) models. For the qualification of AMmaterials, the goal of ML is the development of accurate predictive models. Models may be thought of as human artifacts, i.e., representations of reality within prescribed limits. The generation of ML models is based on the analysis of data. AM is a digitally intense process generating an abundance of data. Data quality, type, and quantity is important, and data must be scrubbed to ensure its pedigree and provenance. This is not trivial and can consume 85% of a data scientist’s time[1,2]. The type of data (continuous or categorical/discrete) must be established. Further, prior to deciding upon an ML approach, the quantity of data needed to derive a meaningful model and the required data environment and infrastructure should be considered carefully. Mathematical algorithms are then used to transform data into models. There are many algorithms used in ML, and each has their appropriate applications, strengths, and weaknesses. Table 1 provides a representative partial list of ML algorithms and their notional characteristics[3,4]. Common types of algorithms include regression, neural nets (NN), deep learning (DL), decision trees (DT), k nearest neighbors (KNN), and support vector machines (SVM)[5]. This article discusses regression, which has broad application, and polynomial regression algorithms, which were used in the research reported by the authors in this work. Regress i on analysis encompasses a large variety of statistical methods to estimate the relationship between input variables and their associated features. Typical regression methods include a) linear regression, b) multilinear regression, c) polynomial regression, and d) logistic regression. Regression analysis is not a new technique but its application to big data sets with large numbers of independent variables to establish (with statistical confidence) a set of dependent materials properties allowables is novel. ML can be applied to AM in a variety of ways in varying stages of maturation. Currently the lowest hanging fruit TABLE 1 — GENERAL CHARACTERISTICS OF MACHINE LEARNING ALGORITHM[3,4] Algorithm Algorithm type* Learning type** Data required Computational time to learn Linear regression R Low Low Polynomial regression R Low Low Logistic regression C Low Low Naïve Bayes C S Low Low Neural network C S, U, R High High Deep learning C S, U, R High High Decision tree C & R S Low Low k-Nearest neighbor I I Low Low Support vector machine C S Low High k-Means C U Low *Type of algorithm: Regression (R), Classification (C), Instance (I) **Learning types: Supervised (S), Unsupervised (U), Reinforced (R), Instance (I)

1 5 A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 (applications that are most mature today) are AM process optimization and certain aspects of process management and control. Once an ML model is trained, AM users can interrogate the model with various questions, such as “what process parameter set should be used if we require ultimate tensile strength of Y?” or “what is the tradeoff between the different process parameter inputs?” or “which process parameter inputs have a large influence on part density?” This allows the AM user to rapidly develop optimized process parameters. Additionally, by monitoring the AM manufacturing process through testing coupon samples at regular intervals, the model can be further trained to detect process drift as a function of time or other factors (e.g., room temperature, humidity, personnel). More sophisticated but far less mature applications for process management and control could come from applying ML to real-time sensory inputs or measurements, such as using computer vision to monitor in-situ sensors, feature engineering of time-series based measurements or training a model to identify features in visual input (e.g., CT scans, microstructure images). In the medium to long term, ML could be used to develop methods for allowable calculations and assist in qualification, requalification, or delta qualification. While the concept of ML allowables is novel, it could be useful today as a “gate check” to help AM users decide whether or not a certain process is stable enough to warrant the time and resources needed to develop conventional allowables using the Metallic Materials Properties Development and Standardization (MMPDS) or Composite Materials Handbook-17 (CMH-17) methods. In other words, ML allowables today could be an estimate of conventional allowables yet to be developed. The ML approach could also be used to demonstrate equivalency, which would greatly aid with requalification or delta qualification. The ML approach allows the issue of equivalency to be easily inverted. Instead of fixing the process parameters and expecting future AM machines to achieve the same requirements with a frozen set of process parameters, the AM user can fix the requirements and ask the ML model to determine what process parameter window on the new AMmachine would allow the AM user to achieve the desired requirements. ML MATERIALS PROPERTY ALLOWABLES DEVELOPMENT Introduction. Senvol recently completed two programs that focused on demonstrating an ML-enabled approach to support materials allowables development. The first project was an Army program [funded via the Advanced Manufacturing, Materials, and Processes (AMMP) consortium] focused on stainless steel 17-4PH. Project members included Senvol, Lockheed Martin Missiles & Fire Control, EWI, Pilgrim Consulting, and Battelle. The second program was an America Makes program focused on a flame retardant polymer, where the project members were Senvol, WSU-NIAR, Northrop Grumman, Stratasys Direct Manufacturing, and Pilgrim Consulting. Both projects completed a sideby-side comparison that evaluated an ML-enabled approach to allowables development. Results showed that an ML-based approach can be more flexible, cost-effective, time-effective, and equivalent to the conventional (e.g., MMPDS in the case of metals, and CMH17 in the case of polymers) approach to materials allowables calculation. Despite the potential that AM offers, the rate of AM adoption is very slow due in part to the high cost and time associated with material allowables development. Furthermore, AM is an advanced manufacturing technique that is process-intensive by definition; the creation of thematerials and the part occurs in the same process. As such: • Conventional materials allowables development binds the user to a limited set of machines and build parameters. • The current allowables paradigm freezes the technology and user in time. • Deviations or multiple allowables require generation of large amounts of additional data. This results in an AM process that is not only costly and time-consuming to implement the first time, but equally costly and time-consuming to maintain in the long run when there are inevitably changes to the AM process. There were two primary objectives of these two projects. 1. Develop and demonstrate a new approach to calculate materials allowables that is not a fixed-point solution. • The projects developed an approach to AM allowables that leverages the digital nature of AM and leverages machine learning (ML). 2. Demonstrate an ML-enabled approach to statistically substantiating materials property predictions across an entire parameter range. • An ML approach is extremely flexible and is able to handle any change to the AM process, thus providing materials property predictions even when deviating from the point at which an allowable was developed. These two projects demonstrated that: 1. An ML approach enables a user to do parameter development and materials allowables development in parallel using the exact same empirical dataset. 2. An ML approach enables a user to make statistically substantiated predictions about performance and scatter everywhere in a given parameter range. This is particularly useful if a user needs to make parts using different parameters (e.g., one parameter set for performance reasons, and a different parameter set for efficiency/cost reasons). 3. ML allowables predicted materials behavior consistently with the

1 6 A D V A N C E D M A T E R I A L S & P R O C E S S E S | A P R I L 2 0 2 3 conventionally developed allowables (i.e., ML allowables were just as accurate as conventionally developed ones). These projects also included a validation portion, which included a performance assessment of the ML allowables against the conventional allowables, as well as a cost-benefit assessment. While the word “allowable” is used in this article, the authors wish to highlight an important caveat, which is that no true allowables were generated in either of the two projects discussed. First, the term “ML allowable” is used for convenience, however it should be noted that an ML-based approach is not an approved methodology for allowable development. Second, due to cost and programmatic constraints, several simplifying decisions needed to be made in generating the conventional allowables based on MMPDS or CMH-17 guidelines (e.g., only one lot of powder was used in each project). Project Steps. The project steps for the two projects are summarized in Table 2. Results and Discussion. The ML approach is vastly different from the conventional MMPDS or CMH-17 approaches to materials allowables. To illustrate, imagine a parameter space, such as the three-dimensional parameter space in Fig. 2, consisting of parameters A, B, and C. The ML approach can be applied to n-dimensions and is particularly suited for high- TABLE 2 — ADDITIVE MANUFACTURING ML ALLOWABLES PROJECT SUMMARIES U.S. Army funded (AMMP Consortium) project developing ML-allowables on stainless steel 17-4 PH America Makes project developing ML-allowables on fire retardant polymer Machine and material Machine: EOS M290 Material: Stainless steel 17-4 PH Machine: 3D Systems sPro60 Material: Nylon 11 flame retardant (FR-106) Step 1: Build and collect training data to develop ML model ML software was used for the DOE (design of experiments). 293 vertical coupons over 3 builds. All coupons were built on a single AM machine. Each coupon was a different parameter set. ML software was used for the DOE (design of experiments). 6 builds of 50 coupons each (i.e., 300 coupons total). Half of the builds were on machine 1, half were on machine 2. Each coupon was a different parameter set. Step 2: Select two optimized parameters based on two different engineering requirements ML model was the basis from which two optimized parameters were selected. Parameter set A is optimal to achieve requirement A. Parameter set B is optimal to achieve requirement B. ML model was the basis from which two optimized parameters were selected. Parameter set A is optimal to achieve requirement A. Parameter set B is optimal to achieve requirement B. Step 3: Calculate ML allowables ML model was used to calculate ML allowable A at parameter set A and ML allowable B at parameter set B. ML model was used to calculate ML allowable A at parameter set A and ML allowable B at parameter set B. Step 4: Develop conventional allowables Followed MMPDS S-basis guidelines: Parameter set A: Three builds with 10 coupons per build (30 coupons total) Parameter set B: Three builds with 10 coupons per build (30 coupons total). Followed CMH-17 B-basis robust sampling guidelines: Parameter set A: 10 builds of five coupons each (50 coupons total) Parameter set B: 10 builds of three coupons each (30 coupons total). Step 5: Calculate conventional allowables Based on MMPDs guidelines, calculate S-basis allowable A and S-basis allowable B. Based on CMH-17 guidelines, calculate B-basis allowable A and B-basis allowable B. Step 6: Build validation build (i.e., previously unseen data) Four representative parts total. 60 witness coupons: 30 built with ML-selected parameter set A, 30 built with ML-selected parameter set B. Eight representative parts total. 38 witness coupons total: 15 built with ML-selected parameter set A, 15 built with ML-selected parameter set B. Step 7: Analysis and comparison Accuracy and usability of ML allowables A and B was compared against those of MMPDS S-basis allowables A and B using data from the validation build. Accuracy and usability of ML allowables A and B was compared against those of CMH-17 B-basis allowables A and B using data from the validation build.