学术文献库

We study the impact of an Upwind scheme on the numerical convergence of simulations of the Hall and Ohmic effect in neutron stars crusts. While simulations of these effects have explored a variety of geometries and wide ranges of physical parameters, they are limited to relatively low values of the Hall parameter, playing the role of the magnetic Reynolds number, which should be not exceed a few hundred for numerical convergence. We study the evolution of the magnetic field in a plane-parallel Cartesian geometry. We discretise the induction equation using a finite difference scheme and then integrate it via the Euler forward method. Two different approaches are used for the integration of the advective terms appearing in the equation: a Forward Time and Central in Space (FTCS) and an Upwind scheme. We compare them in terms of accuracy and performance. We explore the impact of the Upwind method on convergence according to the ratio of planar to vertical field and the Hall parameter. In the limit of a low strength planar field the use of an Upwind scheme provides a vast improvement leading to the convergence of simulations where the Hall parameter is 2 orders of magnitude higher than that of the FTCS. Upwind is still better if the planar field is stronger, yet, the difference of the maximum value of the Hall parameter reached is within a factor of 10 or a few. Moreover, we notice if the schemes diverge their behaviour is very different, with FTCS producing infinite energy, while the Upwind scheme only temporarily increasing the overall magnetic field energy. Overall, the Upwind scheme enhances the efficiency of the simulations allowing the exploration of environments with higher value of electric conductivity getting us closer than before to realistic environmental conditions of magnetars.