This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).
In order to overcome various stiff scales in computational simulations, this research team will develop a fully implicit solver based on a Jacobian-free Newton-Krylov (JFNK) approach with an appropriate preconditioner for FLASH, a multiphysics code framework used widely in the astrophysics community. FLASH provides a block-structured adaptive mesh refinement (AMR) parallel simulation grid and explicit solvers based on Eulerian hydrodynamics as well as N-body astrophysics. It was originally developed for simulations of Type Ia supernovae and related phenomena. Although FLASH has been successfully applied to a range of problems allowing different length scales using its AMR technology, its explicit scheme is of limited use for problems that are stiff in time. This weakness occurs because explicit compressible solvers are extremely inefficient for resolving disparate time scales in the stiff system. The main goal in this project is to develop a time accurate JFNK-based implicit solver and combine it with the current second-order accurate unsplit explicit MHD solver in FLASH. We will use the combined hybrid (explit/implicit) solver to perform cutting-edge simulations addressing the heating of the solar corona problem whose nonlinear physics is dominated by a stiff system of equations.
This development will fully leverage the existing professional management of the FLASH code, and its wide user base, to reach the computational physics community. FLASH is an astrophysics community code available online without cost.
