Control of the number of particles in fluid and MHD particle in cell methods

https://doi.org/10.1016/0010-4655(94)00180-AGet rights and content

Abstract

We describe an algorithm to control the number of particles in fluid, particle-in-cell calculations. In problems with large variations in mass density or grid spacing, the ability to increase or decrease the number of particles in each cell of the mesh becomes essential. Here, a cell-by-cell replacement algorithm which preserves grid data and positivity of the particle data, where appropriate, is described. The algorithm preserves contact discontinuities, introduces little diffusion, and adds little or nothing to the cost of a calculation.

Reference (16)

  • NordmarkH.O.

    J. Comput. Phys.

    (1991)
  • BurgessD. et al.

    J. Comput. Phys.

    (1992)
  • EastwoodJ.W.

    Comput. Phys. Commun.

    (1987)
  • BrackbillJ.U. et al.

    J. Comput. Phys.

    (1986)
  • BrackbillJ.U.

    J. Comput. Phys.

    (1991)
  • LapentaG. et al.

    J. Comput. Phys.

    (1994)
  • RobertA.

    J. Meteor. Soc. Japan

    (1982)
  • RobertA. et al.

    Mon. Wea. Rev.

    (1985)
There are more references available in the full text version of this article.

Cited by (32)

  • Strategies for particle resampling in PIC simulations

    2021, Computer Physics Communications
    Citation Excerpt :

    A way to perform thinning with conservation of several arbitrary particle and grid quantities is proposed in Ref. [15]. Within the approach of complete resampling, Lapenta and Brackbill proposed a method of replacing the macroparticles in a given cell with a new subset, preserving the contributions to the grid quantities and also maximizing the uniformity of the distribution of these quantities over the new subset [16]. A way of doing such resampling with the conservation of grid values for charge and current density, as well as of the total energy of the resampled macroparticles, has been proposed by Assous et al. [17] and further developed in Ref. [18].

  • Moment preserving constrained resampling with applications to particle-in-cell methods

    2020, Journal of Computational Physics
    Citation Excerpt :

    However, because here we are concerned with a sample-based representation of a distribution function (f), the motivation is quite different. By minimizing the variance of the particle weights, we ensure that the particles carry near-maximum information about the distribution; whereas, in [23], the objective is to ensure a smooth representation of a continuous field. The motivation here is more closely related to the use of uniformly weighted samples in Monte Carlo methods.

View all citing articles on Scopus
View full text