Construction and Performance of Kinetic Schemes for Linear Systems of Conservation Laws
Emmanuel Audusse, Sébastien Boyaval, Virgile Dubos, Minh-Hoang Le
TL;DR
The paper develops vectorial kinetic schemes for linear symmetric-hyperbolic conservation laws, using linear acoustics/elastodynamics as a benchmark. It constructs Maxwellians that enforce a discrete convex entropy extension and leverages isotropy-preserving lattices (D2Q5/D2Q9) to obtain stable, high-fidelity discretizations. Through a parameter-focused design (minimizing diffusion and optimizing CFL), it demonstrates that large-CFL kinetic schemes can outperform standard FD/FV methods in accuracy for smooth solutions. Numerical experiments on Cartesian grids show first-order convergence for ω=1 and second-order convergence for ω=2, with stability and error closely tied to the discrete entropy conditions and diffusion minimization criteria. The results motivate further rigorous analysis and extension of kinetic schemes to broader symmetric-hyperbolic systems.
Abstract
We describe a methodology to build vectorial kinetic schemes, targetting the numerical solution of linear symmetric-hyperbolic systems of conservation laws -a minimal application case for those schemes. Precisely, we fully detail the construction of kinetic schemes that satisfy a discrete equivalent to a convex extension (an additional non-trivial conservation law) of the target system -the (linear) acoustic and elastodynamics systems, specifically -. Then, we evaluate numerically the convergence of various possible kinetic schemes toward smooth solutions, in comparison with standard finite-difference and finite-volume discretizations on Cartesian meshes. Our numerical results confirm the interest of ensuring a discrete equivalent to a convex extension, and show the influence of remaining parameter variations in terms of error magnitude, both for ''first-order'' and ''second-order'' kinetic schemes\,: the parameter choice with largest CFL number (equiv., smallest spurious diffusion in the equivalent equation analysis) has the smallest discretization error.
