For collocated SBP finite difference operators, discrete Helmholtz Hodge decompositions have an unavoidable remainder caused by grid oscillations, and iterative least-squares projections overcome this in practice.
Numerical Methods for the Magnetic Induction Equation with Hall Effect and Projections onto Divergence-Free Vector Fields
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
The nonlinear magnetic induction equation with Hall effect can be used to model magnetic fields, e.g. in astrophysical plasma environments. In order to give reliable results, numerical simulations should be carried out using effective and efficient schemes. Thus, high-order stable schemes are investigated here. Following the approach provided recently by Nordstr\"om (J Sci Comput 71.1, pp. 365--385, 2017), an energy analysis for both the linear and the nonlinear induction equation including boundary conditions is performed at first. Novel outflow boundary conditions for the Hall induction equation are proposed, resulting in an energy estimate. Based on an energy analysis of the initial boundary value problem at the continuous level, semidiscretisations using summation by parts (SBP) operators and simultaneous approximation terms are created. Mimicking estimates at the continuous level, several energy stable schemes are obtained in this way and compared in numerical experiments. Moreover, stabilisation techniques correcting errors in the numerical divergence of the magnetic field via projection methods are studied from an energetic point of view in the SBP framework. In particular, the treatment of boundaries is investigated and a new approach with some improved properties is proposed.
fields
math.NA 1years
2019 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Discrete Vector Calculus and Helmholtz Hodge Decomposition for Classical Finite Difference Summation by Parts Operators
For collocated SBP finite difference operators, discrete Helmholtz Hodge decompositions have an unavoidable remainder caused by grid oscillations, and iterative least-squares projections overcome this in practice.