Description
High-resolution fluid simulations for plasma physics and astrophysics rely on Particle-in-cell (PIC) and hydrodynamic solvers (e.g., FLASH) to resolve shock-dominated, multi-scale phenomena, but their high computational cost severely limits scalability. This motivates the development of learning-based surrogate models, which offer a promising route to accelerate these simulations while preserving physical fidelity. We propose a data-driven graph neural network with a weak physics constrain to approximate shock phenomena from grid-based simulation data, increasing learning speed by leveraging multiple GPUs. Once trained, the surrogate model can then be applied to replicate the results of traditional simulations at a fraction of the computational expense, reducing their development time and cost. In this work, we study the Sedov–Taylor shock propagation problem using a physics-informed graph-based surrogate model, Physics-Informed MeshGraphNet (Phy-MGN), designed for grid-based hydrodynamics. By incorporating physics constraints derived from the Euler equations using finite difference method, the model captures the self-similar shock evolution and associated flow structures without explicitly solving the full hydrodynamic equations at each timestep. Comparing to the baseline MeshGraphNet model, Phy-MGN is able to generalize beyond the training regime with a higher accuracy and preserves differentiability in parameter space while achieving a substantial reduction in computational cost relative to conventional numerical solvers.