Mathematical Modeling and Analysis
Mathematical Modeling and Analysis
Robust and efficient multilevel iterative solvers are vital for the predictive simulation of complex multiscale and multicomponent nonlinear applications. Specifically, diffusive phenomena play a significant role in a wide range of applications, including radiation transport, flow in porous media, and composite materials. In fact, the solution of the diffusive component (elliptic component) of these systems frequently dominates the simulation cost because it is characterized by a discontinuous diffusion coefficient with fine-scale spatial structure. To address this problem we have implemented an MPI-based parallel version of Dendy's robust Black Box Multigrid Code for structured grids (BoxMG) in both two and three dimensions. We have demonstrated that this code scales well over one sector (3584 processors) of the LANL supercomputer Q, and performs well on Beowulf class Linux clusters. We anticipate that the BoxMG solvers will have a positive impact on several code projects at LANL. Thus far, the code has been integrated into the ZATHRAS gray thermal radiation transport equation solver. In the very near future we plan to make a public release of the BoxMG solver package.
