Radiative cooling is a ubiquitous process in several high energy density physics (HEDP) experiments, however numerical modelling of the final compressed state is often resolution-limited. This work presents the development of a static mesh refinement capability for use in the radiation-magnetohydrodynamics code, Chimera, in order to perform high resolution simulations of pulsed-power driven experiments.
A block-based refinement strategy is used, where the grid is recursively split into blocks of equal number of cells, but at different resolutions. Each processor is allocated a single block, therefore all processors perform an equal number of computations, and load balancing is inherently achieved. The exchange of grid variables across different resolutions is performed via MPI communications. A novel algorithm is described for exchanging MHD variables in a constrained-transport solver, to maintain the divergence-free properties of the solution.
Use of this capability is demonstrated on simulations of three systems: single wire Z-pinches, X-pinches (in wire and solid configurations), and magnetic reconnection using inverse wire arrays. A systematic study of single wire experiments on a 150 kA generator is performed in 1D, 2D, and 3D at equivalent high resolution. Simulations in 1D demonstrate the need to use EoS and transport models with condensed matter corrections to perform simulations with ‘cold-start’ initialisation. Simulations in 2D and 3D demonstrate the effect of MHD instabilities on hotspot formation. High resolution wire X-pinch simulations in 3D are presented that, for the first time, demonstrate the 3D asymmetries that arise at the crossing point due to wire offsets. 2D axisymmetric simulations of solid X pinch experiments are presented demonstrating extreme conditions reached at the core of the crossing point. Finally, 2D and 3D simulations of radiatively-cooled magnetic reconnection experiments performed on the Z facility are presented, that demonstrate the sensitivity of the reconnection dynamics to resolution and dimensionality.