A GPU-accelerated high-order multiphase computational tool for asteroid fragmentation and pulverization modeling

Abstract

<p>The impact threat of asteroids or comets, referred to as near-Earth objects (NEOs), is a growing concern to the global community. A NEO collision could have severe consequences, especially in highly populated regions. To combat this threat, several asteroid deflection strategies have been introduced, but computational modeling is needed to investigate the feasibility of such missions. To this end, an asteroid disruption software tool was built to handle the simulation of multiple disruption techniques, namely high-energy explosives and kinetic-energy impactors, and to investigate the feasibility and effectiveness of these approaches. In addition, the software is intended to use graphics processing units (GPUs) as the primary computational resource rather than central processing units (CPUs). While GPUs are quickly becoming an alternative computing platform for numerical simulations, it is not clear which numerical schemes provide the highest computational efficiency for different problem types.</p> <p>The numerical accuracies and computational work of several numerical methods are compared using GPU computing implementation. The Correction Procedure via Reconstruction (CPR), Discontinuous Galerkin (DG), Nodal Discontinuous Galerkin (NDG), Spectral Difference (SD), and Finite Volume (FV) methods are investigated for smooth and discontinuous problems to determine the most efficient method to apply towards asteroid disruption simulations. The computational time to reach a set error criteria and total time to compute solutions are compared across the methods. It is shown that while FV methods can produce solutions with the lowest computation time for discontinuous problems, they produce larger errors for smooth problems at the same order of accuracy. The SD method illustrates an excellent trade-off in terms of error and total work, computing both smooth and discontinuous problems faster than most other methods while providing low error norms.</p> <p>From the aforementioned study, the SD method is applied to multifluid modeling for asteroid disruption applications. In order to model the multiple material phases associated with the problem, a Diffused-Interface Method (DIM) approach is integrated into the SD method (SD-DIM). This allows high-order solution reconstructions for problems containing multi-material interactions, where different equations of states define each material. In addition, a damage model is developed to simulate asteroid fracturing based on material phase changes. This results in a novel GPU-based high-order SD-DIM computational tool to explore the complex problem of asteroid fragmentation and pulverization. Several asteroid disruption simulations are completed, including kinetic-energy impactors, multi-kinetic energy impactor systems, and nuclear options. Results illustrate the benefits of using a multi-kinetic energy system when compared to a single impactor system for non-nuclear options. The effectiveness of nuclear options is also observed, where complete target destruction is shown.</p>