Section: New Software and Platforms

Platforms

NS2DDV-M  [42] is a Matlab code, developed by C. Calgaro, E. Creusé, and A. Mouton (CNRS research engineer at Université de Lille), for the simulation of homogeneous and inhomogeneous fluid flows by a combined Finite Volume-Finite Element method. The code embeds parallel computation facilities, and is freely distributed, to allow for easy comparisons with concurrent codes on benchmark test-cases, and to promote new collaborations in the domain.

KinDiff consists in a Python library and a Jupyter Notebook. It has been developed by M. Bessemoulin-Chatard (CNRS and Université de Nantes), M. Herda, and T. Rey. The KinDiff code approximates the solutions to the kinetic Fokker–Planck and linear BGK equations in a unidimensional periodic domain in space and a symmetric bounded unidimensional domain in velocity. Its outputs are the discretization of the distribution function in the phase space as well as the moments, along with some visualizations of these quantities. Kinetic equations display asymptotic behaviors such as return to equilibrium in large time, or convergence towards macroscopic equations in the diffusive limit. The finite volume method implemented within KinDiff reproduces these asymptotics at the discrete level. The presentation and numerical analysis of the scheme is available in the article [17]. The code is available at gitlab.com/thoma.rey/FV_HipoDiff.