A Multi Level Multi Domain Method for Particle In Cell plasma simulations

  • Authors:

  • M. E. Innocenti;G. Lapenta;S. Markidis;A. Beck;A. Vapirev

  • Affiliations:

  • Center for Plasma Astrophysics, Department of Mathematics, K.U. Leuven, Celestijnenlaan 200B, B-3001 Leuven, Belgium;Center for Plasma Astrophysics, Department of Mathematics, K.U. Leuven, Celestijnenlaan 200B, B-3001 Leuven, Belgium and ExaScience Intel Lab Europe, Kapeldreef 75, B-3001 Leuven, Belgium and Spac ...;Center for Plasma Astrophysics, Department of Mathematics, K.U. Leuven, Celestijnenlaan 200B, B-3001 Leuven, Belgium and ExaScience Intel Lab Europe, Kapeldreef 75, B-3001 Leuven, Belgium and PDC ...;Center for Plasma Astrophysics, Department of Mathematics, K.U. Leuven, Celestijnenlaan 200B, B-3001 Leuven, Belgium and ExaScience Intel Lab Europe, Kapeldreef 75, B-3001 Leuven, Belgium;Center for Plasma Astrophysics, Department of Mathematics, K.U. Leuven, Celestijnenlaan 200B, B-3001 Leuven, Belgium and ExaScience Intel Lab Europe, Kapeldreef 75, B-3001 Leuven, Belgium

  • Venue:
  • Journal of Computational Physics
  • Year:
  • 2013

Quantified Score

Hi-index 31.45

Visualization

Abstract

A novel adaptive technique for electromagnetic Particle In Cell (PIC) plasma simulations is presented here. Two main issues are identified as regards the development of the algorithm. First, the choice of the size of the particle shape function in progressively refined grids, with the decision to avoid both time-dependent shape functions and cumbersome particle-to-grid interpolation techniques, and, second, the necessity to comply with the strict stability constraints of the explicit PIC algorithm. The adaptive implementation presented responds to these demands with the introduction of a Multi Level Multi Domain (MLMD) system, where a cloud of self-similar domains is fully simulated with both fields and particles, and the use of an Implicit Moment PIC method as baseline algorithm for the adaptive evolution. Information is exchanged between the levels with the projection of the field information from the refined to the coarser levels and the interpolation of the boundary conditions for the refined levels from the coarser level fields. Particles are bound to their level of origin and are prevented from transitioning to coarser levels, but are repopulated at the refined grid boundaries with a splitting technique. The presented algorithm is tested against a series of simulation challenges.