Development of AI-powered Application for Advanced Nuclear Reactor Design

PI Jun Fang, Argonne National Laboratory
Co-PI Aleksandr Obabko, Argonne National Laboratory
Tingzhou Fei, Argonne National Laboratory
Project Summary

This project is creating an easy-to-use AI app that taps advanced simulations to quickly test and improve designs for safer, more efficient molten-salt nuclear reactors—reducing time, cost, and expertise needed and helping speed progress toward cleaner energy.

Project Description

This project develops an artificial intelligence (AI)-powered application to accelerate and optimize the design of advanced nuclear reactors, with a focus on the Molten Salt Reactor (MSR), a promising technology known for its intrinsic safety and high thermal efficiency. By integrating high-fidelity Computational Fluid Dynamics (CFD) simulations with machine learning (ML), the project aims to automate the complex design process that traditionally requires expert input and time-consuming iterations. The resulting app will enable users—regardless of their background in CFD or ML—to explore optimized reactor configurations by simply inputting key design parameters. The system will feature two main components: a CFD module based on the NekRS code, and an AI/ML module that builds surrogate models from simulation data to guide design improvements. 

The project is expected to significantly lower the barriers to advanced reactor design by reducing time, cost, and expertise requirements. It supports the U.S. Department of Energy’s mission to advance clean, secure, and sustainable energy solutions by enabling digital engineering tools that can accelerate innovation in nuclear energy. By demonstrating a novel workflow that bridges nuclear engineering and AI, the project lays the groundwork for a transformative shift in how reactors are conceptualized and built, potentially drawing interest from a wide range of technology vendors and stakeholders across the energy sector.

Allocations