Optimization of multiphysics at extreme conditions with differentiable codes and agent-driven simulations

Inertial fusion energy promises nearly unlimited, clean power if it can be achieved. However, the design and engineering of fusion systems requires controlling and manipulating matter at extreme energies and timescales; the shock physics and radiation transport governing the physical behavior under these conditions are complex requiring the development, calibration, and use of predictive multiphysics codes to navigate the highly nonlinear and multi-faceted design landscape. In this talk, we explore two novel technical approaches to optimization in hydrodynamics and multi-physics: differentiable codes and artificial intelligence reasoning models.

In part one, we develop a gradient based optimization approach for the equations of compressible, Lagrangian hydrodynamics and demonstrate how it can be employed to automatically uncover strategies to control hydrodynamic instabilities arising from shock acceleration of density interfaces. These instabilities lead to mixing which, in the case of laser driven ICF, quenches the runaway fusion process ruining the potential for positive energy return. We demonstrate that control of Richtmyer-Meshkov instability (RMI) can be achieved by optimization of initial conditions with (> 100) parameters.

In part two, we hypothesize that artificial intelligence reasoning models can be combined with physics codes and emulators to autonomously design NIF fusion fuel capsules. We construct a multi-agent system where natural language is utilized to explore the complex physics regimes around fusion energy. The agentic system is capable of executing a high-order multiphysics inertial fusion computational code. We demonstrate the capacity of the multi-agent design assistant to both collaboratively and autonomously manipulate, navigate, and optimize capsule geometry while accounting for high fidelity physics that ultimately achieve simulated ignition via inverse design.