English

A direction preserving discretization for computing phase-space densities

Numerical Analysis 2021-06-29 v1 Numerical Analysis Computational Physics

Abstract

Ray flow methods are an efficient tool to estimate vibro-acoustic or electromagnetic energy transport in complex domains at high-frequencies. Here, a Petrov-Galerkin discretization of a phase-space boundary integral equation for transporting wave energy densities on two-dimensional surfaces is proposed. The directional dependence of the energy density is approximated at each point on the boundary in terms of a finite local set of directions propagating into the domain. The direction of propagation can be preserved for transport across multi-component domains when the directions within the local set are inherited from a global direction set. The range of applicability and computational cost of the method will be explored through a series of numerical experiments, including wave problems from both acoustics and elasticity in both single and multi-component domains. The domain geometries considered range from both regular and irregular polygons to curved surfaces, including a cast aluminium shock tower from a Range Rover car.

Keywords

Cite

@article{arxiv.2106.14506,
  title  = {A direction preserving discretization for computing phase-space densities},
  author = {David J. Chappell and Jonathan J. Crofts and Martin Richter and Gregor Tanner},
  journal= {arXiv preprint arXiv:2106.14506},
  year   = {2021}
}

Comments

23 pages, 10 figures

R2 v1 2026-06-24T03:39:32.195Z