FOReCAST: A New Parallel Code for Self-Gravitating Magnetized Flows

This work is devoted to the description of our newly developed multidimensional parallel astrophysical magnetohydrodynamics (MHD) code FOReCAST (short for FORtran Coarray Astrophysical Simulation Tool). It is written in the modern Fortran language, utilizing its native parallelism extension via Coarray Fortran (CAF) technology. The code can be used for the simulation of self-gravitating magnetized flows in 1-, 2-, and 3-dimensional settings, with a focus on magnetorotational astrophysical processes, while also being capable of solving other MHD-related problems. The core algorithm is based on Godunov-type methods with a Galilean-invariant magnetic-field divergence-cleaning technique. Various high-order spatial reconstruction (PLM, PPM, and Monotonicity Preserving) methods and TVD Runge–Kutta time-stepping are implemented in FOReCAST to increase the order of accuracy of the numerical solution. One of the code’s virtues is a mesh-coarsening procedure, which is applied in curvilinear geometries near the coordinate singularities. It allows to avoid a restrictive Courant–Friedrichs–Lewy time-step condition near geometric singularities and simulate the long-term evolution of collapsed stellar (or protostellar) cores on a spherical grid. Several test problems are presented to demonstrate the code’s robustness and accuracy.