ADVANCED MATERIALS & PROCESSES | MAY/JUNE 2025 10 MACHINE LEARNING | AI A machine learning tool developed at Pacific Northwest National Laboratory, Richland, Wash., analyzes thin film growth data and flags changes as they emerge. The program lays the groundwork for systems capable of adjusting growth conditions without human input. pnl.gov. BRIEF AI SUGGESTS BEST AEROSPACE ALLOYS Researchers at Skolkovo Institute of Science and Technology (Skoltech) and the Moscow Institute of Physics and Technology, both in Russia, developed a machine learning-driven approach to quickly select promising metal alloy compositions for aerospace applications. “The current approaches rely on a fundamental physical description of the process in terms of direct quantum mechanical calculations,” says researcher Viktoriia Zinkovich. “We, on the other hand, use machine- learned potentials, which are characterized by rapid computations and make it possible to sort through all possible combinations up to a certain cutoff limit, 20 atoms per supercell, for example. That means we won’t miss the good candidates.” The new approach was validated on two systems: five metals with high melting points (vanadium, molybdenum, niobium, tantalum, and tungsten) and five noble metals (gold, platinum, palladium, copper, and silver). In each system, the team considered three elemental compositions. For example, copper and platinum or all five noble metals at once. Notably, the five elements making up each list tend to adopt the same crystal structure. This simplifies calculations because the alloy is assumed to have that structure as well. The scientists then applied their search algorithm to each of the six elemental compositions: three for the noble and three for the high-melting-point metals. The algorithm enabled the team to discover 268 new alloys stable at zero temperature that are not listed in a popular materials database called AFLOW. For example, in the niobium-molybdenum-tungsten system, the approach using machine- learned potentials produced 12 alloy candidates, whereas AFLOW contained no three-component alloys of these elements. Properties of the new alloys must be verified in greater detail with experiments to determine which materials hold promise for practical applications. https:// new.skoltech.ru. STRENGTHENING TITANIUM THROUGH AI Scientists at Johns Hopkins Applied Physics Laboratory (APL), Laurel, Maryland, and the Johns Hopkins Whiting School of Engineering are using AI to make titanium alloy parts more quickly, stronger, and with nearly perfect Brendan Croom and his team are using AI to optimize titanium alloy production. Courtesy of APL. precision. The research focused on Ti-6Al-4V and employed AI-driven models to map out previously unexplored manufacturing conditions for metal 3D printing using laser powder bed fusion. The results challenge assumptions about process limits, revealing a broader processing window for producing dense, high-quality titanium with customizable mechanical properties. “By using AI to explore the full range of possibilities, we discovered new processing regions that allow for faster printing while maintaining or even improving material strength and ductility. Now, engineers can select the optimal processing settings based on their specific needs,” says APL materials scientist Brendan Croom. Building on earlier work, the team used a machine learning approach to reveal a high- density processing regime previously dismissed due to concerns about material instability. With targeted adjustments, the scientists unlocked new ways to process Ti-6Al-4V. “This isn’t just about manufacturing parts more quickly,” says Croom. “It’s about striking the right balance among strength, flexibility, and efficiency. AI is helping us explore processing regions we wouldn’t have considered on our own.” jhuapl.edu. Image generated by DDG DaVinci2 model with prompt from Nicolas Posunko/Skoltech.
RkJQdWJsaXNoZXIy MTYyMzk3NQ==