The South Dakota Board of Regents has awarded a $90,000 Competitive Research Grant for Modeling and Design of Magnetic Soft Materials Accelerated by Neural Operators, running from June 2025 through April 2027. It is the second award announced this week, alongside the NSF ERI award.
The work has three parts. The first is to compare the available models of effective material properties in magnetic soft composites, and establish which of them can be relied on for design rather than description. The second is to build topology optimization solvers that treat structural density and magnetization as one coupled problem. The third is to use neural operator surrogates to make the repeated forward solves that optimization requires affordable. This connects the material-design problem with the group’s work on residual-based error estimation and correction for neural operators.
Progress is on the design of magnetic soft materials and neural operators project pages.
The group is recruiting undergraduate and graduate researchers interested in computational mechanics, and in applying numerical and machine-learning methods to fracture, functional soft materials and granular media. Write to prashant.jha@sdsmt.edu.