Artificial Viscosity Models for Vortex and Particle Methods

✍ Scribed by G.-H. Cottet

• Publisher: Elsevier Science
• Year: 1996
• Tongue: English
• Weight: 1000 KB
• Volume: 127
• Category: Article
• ISSN: 0021-9991
• DOI: 10.1006/jcph.1996.0176

✦ Synopsis

It turns out that the precise form of the error produced by vortex methods, and not only their order of conver-

The analysis of the truncation error produced by particle methods leads to artificial viscosity schemes. For vortex methods, they can be gence, can be precisely analyzed. This feature, which defiseen as eddy viscosity models, with anisotropic non-linear diffusion nitely distinguishes vortex methods from grid-based methtensors. Numerical experiments on decaying incompressible 2D ods, stems from the fact that vortex methods and, more turbulence illustrate the efficiency of the method, and in particular generally, particle methods are based on exact weak soluthe fact that the diffusion stops acting in large coherent eddies.

tions of advection equations. The truncation error solely

For compressible flows, this approach allows us to understand the oscillations produced by particle methods and to derive new artifi-results from the mollification used in practice in the compucial viscosity schemes.