Nathan M. Soltes
CNS Y-12
Thanks to AIST’s support, I had the unique opportunity to research and implement advanced manufacturing capabilities while working as a metallurgical development intern for Consolidated Nuclear Security at the Y-2 National Security Complex. Throughout the summer, I was able to learn and apply materials data science techniques to better inform and guide production decisions in foundry and steel technology applications.
A key contribution I made was mapping the variable performance of a critical manufacturing process with a novel 3D power density map, complete with cubic interpolation, spatial clustering and a refinement algorithm. Leveraging that map, I optimized an inverse solving function that systemically calculates the ideal input conditions that generate any number of targeted outputs. Furthermore, I combined a dense grid sweep algorithm and user-input performance floors to estimate the real-world operator error margin down to 1/100 of a milliamp, providing adaptive error margins that reduce uncertainty and increase production throughput.
At the end of this internship, I am proud to have supported Y-12’s national security mission by drastically reducing the time and cost associated with experimental-heavy “guess and check” testing procedures. I am incredibly grateful to AIST for this opportunity to learn about the cutting edge of manufacturing technology and am excited to apply the skills and techniques I’ve learned to overcome the steel industry’s current and future challenges.