Vector implementation of nonlinear Monte Carlo Coulomb collisions (Q1815914): Difference between revisions

The authors present a nonlinear Monte Carlo collision model based on Coulomb binary collisions. The model uses the nonlinear Monte Carlo collision operator with emphasis on its efficient implementation. This operator is simple in its form and easy to implement. The basic features of the operator are that it quasilocally fulfills basic conservation laws, namely, momentum, energy and particle number conservations, which are characteristics of the exact Fokker-Planck operator. As a consequence, the basic collisional transport properties can be correctly described. The efficiency of the scalar implementation is on the same level as that of the earlier model of \textit{T. Takizuka} and \textit{H. Abe} [J. Comput. Phys. 25, 205-219 (1977; Zbl 0403.76091)]. For the practical application, two vectorizable algorithms have been designed. In order to be applied, the Monte Carlo operator is tested by simulation of various problems in velocity space concerning relaxation processes, deviation from Maxwellian distribution and electrical conductivity. The test results show that the model is practically applicable.