Next Generation Blast Simulation

With the current state of world events, the threat of explosives used against high-value targets is more pronounced than ever before. As a result, simulation of blast events and their effects on structures has become an increasingly vital capability. Current blast simulations on Central Processing Units (CPU) resources require a significant amount of computational time, limiting their overall value in planning and decision making. In this project, we propose to develop a next-generation, 2D Computational Fluid Dynamics (CFD) code, which is targeted to run on commodity Graphics Processing Units (GPUs). The 2D code will be used to evaluate the potential speedup available using the GPU approach when compared to a CPU implementation. The final Phase I product will implement the Discontinuous Galerkin (DG) method with an explicit time-stepping Runge-Kutta type algorithm. The DG method is algorithmically compatible with GPUs and accurately models highly complex fluid flow with high velocities and discontinuities, which are common to blast conditions. Future phases of the project would extend the capabilities of the CFD code to 3D and investigate integration with a structural model.