Texas A and M receives funding to train machine learning tools in discovery of new materials for fusion power plants
An integral part of a fusion power plant is its first wall, or the interior surface that faces the plasma during energy production. The strength and longevity of the first wall are key to the power plant's service life, creating the potential for lower operational expenses. Texas A and M researchers have been awarded $2.36 million to discover new materials suitable to fabricate this first wall. The Texas A and M-based team is led by Dr. Raymundo Arroyave, in collaboration with Dr. Ibrahim Karaman, Dr. George Pharr, Dr. Lin Shao, and Dr. Kelvin Xie, and includes collaborators Danny Perez from Los Alamos National Laboratory and Tim Graening from Oak Ridge National Laboratory.
The research will leverage synthesis, irradiation, and testing facilities, along with advanced computer simulations enhanced by machine learning, to conduct a high-impact materials discovery campaign. Simulations will screen hundreds of materials daily, while experimental facilities will synthesize and test the most promising candidates. Guided by Texas A and M's BIRDSHOT AI framework - proven to accelerate materials discovery by 100-fold - the project aims to rapidly identify and fabricate optimized, stronger, and more durable first-wall materials.
"Finding the best materials for these transformative applications is like finding a needle in a multi-dimensional haystack, only much harder. Finding the right material in the least number of trials requires approaches beyond current methods. BIRDSHOT is such a tool," said Arroyave, a Segers Family Dean's Excellence Professor in the Department of Materials Science and Engineering. "The first version of BIRDSHOT was developed under ARPA-E's ULTIMATE program but now we propose to significantly enhance its performance based on all the experience we have gained over the past few years."
This research is one of several projects managed through the Creating Hardened and Durable Fusion First Wall Incorporating Centralized Knowledge (CHADWICK) program, which aims to discover or develop a class of first wall materials to maintain design performance over the lifetime of a fusion power plant.
"This project brings together a great team of researchers from TAMU and national laboratories with a broad range of skills to work collectively on a problem of great national and international importance - developing entirely new types of materials for the future of energy production," said Dr. George Pharr, a University Distinguished Professor in the Department of Materials Science and Engineering.
The longevity and durability of the first wall - which bears the power plant's mechanical load and protects the plant from the extreme environment in the core - is a major obstacle in the commercialization of fusion power. Identifying an optimal first wall material is essential for accelerating commercial fusion.
"The ARPA-E CHADWICK program will build on the success of our recent work in the ARPA-E GAMOW program, bringing us a step closer to making commercial fusion energy a reality. In the GAMOW program, we showed that it's possible to create functionally graded structures, starting with tungsten for the plasma-facing surface and transitioning to ferritic-martensitic steels for the structural components of fusion reactors," said Materials Science and Engineering Department Head Dr. Ibrahim Karaman.
"However, challenges remain in ensuring these materials can last at high temperatures over long periods. To address this, we need to design new materials that can withstand the extreme conditions inside fusion reactors. Our goal is to use the advanced BIRDSHOT platform we recently developed, which accelerates the discovery of new materials, to solve this challenge."