Abstract:We extend to multi-dimensions the work of [1], where new fully explicit kinetic methods were built for the approximation of linear and non-linear convection-diffusion problems. The fundamental principles from the earlier work are retained: (1) rather than aiming for the desired equations in the strict limit of a vanishing relaxation parameter, as is commonly done in the diffusion limit of kinetic methods, diffusion terms are sought as a first-order correction of this limit in a Chapman-Enskog expansion, (2) introducing a coupling between the conserved variables within the relaxation process by a specifically designed collision matrix makes it possible to systematically match a desired diffusion. Extending this strategy to multi-dimensions cannot, however, be achieved through simple directional splitting, as diffusion is likely to couple space directions with each other, such as with shear viscosity in the Navier-Stokes equations. In this work, we show how rewriting the collision matrix in terms of moments can address this issue, regardless of the number of kinetic waves, while ensuring conservation systematically. This rewriting allows for introducing a new class of kinetic models called \emph{regularized} models, simplifying the numerical methods and establishing connections with Jin-Xin models. Subsequently, new explicit arbitrary high-order kinetic schemes are formulated and validated on standard two-dimensional cases from the literature. Excellent results are obtained in the simulation of a shock-boundary layer interaction, validating their ability to approximate the Navier-Stokes equations with kinetic speeds obeying nothing but a subcharacteristic condition along with a hyperbolic constraint on the time step.

Contraction and Convergence Rates for Discretized Kinetic Langevin Dynamics

Contraction and Convergence Rates for Discretized Kinetic Langevin Dynamics

Contraction Rate Estimates of Stochastic Gradient Kinetic Langevin Integrators

Appropriate State-Dependent Friction Coefficient Accelerates Kinetic Langevin Dynamics

Convergence of kinetic Langevin samplers for non-convex potentials

Explicit convergence rates of underdamped Langevin dynamics under weighted and weak Poincaré--Lions inequalities

Discrete-to-continuum linearization in atomistic dynamics

On explicit $L^2$-convergence rate estimate for underdamped Langevin dynamics

Numerical integrators for confined Langevin dynamics

Quantitative hydrodynamic limits of the Langevin dynamics for gradient interface models

On the Definition of Velocity in Discrete-Time, Stochastic Langevin Simulations

Accelerated convergence to equilibrium and reduced asymptotic variance for Langevin dynamics using Stratonovich perturbations

Global Convergence of Langevin Dynamics Based Algorithms for Nonconvex Optimization

Ergodicity of Langevin Dynamics and its Discretizations for Non-smooth Potentials

The velocity jump Langevin process and its splitting scheme: long time convergence and numerical accuracy

A class of asymptotic-preserving schemes for kinetic equations and related problems with stiff sources

Convergence of Kinetic Langevin Monte Carlo on Lie groups

Exponential Stability and Hypoelliptic Regularization for the Kinetic Fokker–Planck Equation with Confining Potential

A new local and explicit kinetic method for linear and non-linear convection-diffusion problems with finite kinetic speeds: II. Multi-dimensional case

Kinetic Interacting Particle Langevin Monte Carlo