This article was accepted into the corpus but its outbound wikilinks were never NER-processed — typical at the deepest BFS hop or when the run's entity cap was reached. No expansion funnel to show.
| Roe solver | |
|---|---|
| Name | Roe solver |
| Type | Numerical flux solver |
| Introduced | 1981 |
| Inventor | P. L. Roe |
| Field | Computational fluid dynamics |
Roe solver
The Roe solver is a widely used approximate Riemann solver for hyperbolic conservation laws, notable for linearizing nonlinear fluxes to produce upwind numerical fluxes. Developed to address discontinuities such as shocks and contact discontinuities, it has influenced finite-volume schemes in aerodynamics, astrophysics, and weather modeling. Its formulation balances accuracy and computational cost and has inspired numerous extensions and fixes.
The Roe solver was introduced by P. L. Roe in 1981 to handle the Euler equations of compressible flow and related systems arising in aerospace engineering and astrophysical fluid dynamics. It provides a linearized Jacobian that preserves conserved quantities across discontinuities and is often compared with the exact Riemann solver, Godunov methods, and Harten–Lax–van Leer (HLL) approaches. In practice it appears in codes for supersonic flows, turbomachinery, and magnetohydrodynamics where robustness and low dissipation are required.
The core idea is to construct a Roe-averaged Jacobian A~ that approximates the nonlinear flux derivative ∂F/∂U between left and right states. For a system of conservation laws ∂U/∂t + ∂F(U)/∂x = 0, the Roe flux is F_Roe = 1/2(F(U_L)+F(U_R)) - 1/2|A~|(U_R−U_L), where |A~| is built from the absolute values of eigenvalues of A~ and its eigenvectors. For the compressible Euler equations the Roe averages produce Riemann invariants that diagonalize A~ and yield wave strengths proportional to jumps in conserved variables. The method relies on properties similar to linearized acoustics and benefits from characteristic decomposition familiar from studies of the Navier–Stokes equations, Burgers' equation, and shallow water equations.
Implementing the Roe solver typically involves state reconstruction, computation of Roe-averaged states, eigen-decomposition of A~, and application of an entropy fix when necessary. Reconstruction choices range from piecewise-constant Godunov to MUSCL with slope limiters influenced by van Leer, Osher, and Sweby to prevent spurious oscillations near discontinuities. Eigenvalue evaluation and flux splitting are computational kernels in codes such as those used at research centers and national laboratories; optimization often targets implementations on heterogeneous hardware from clusters to GPUs. Popular software frameworks that incorporate Roe-like solvers include legacy CFD packages used by aerospace companies, university research codes, and open-source platforms for computational astrophysics.
Numerous variants address weaknesses and generalize the Roe approach. Entropy fixes by Harten, Hyman, and others modify small eigenvalues to enforce the Lax entropy condition. Extensions include Roe–Pike average formulations for multi-species flows, HLLC and HLLEM hybrids that restore missing contact or shear waves, and adaptations for magnetohydrodynamics developed after concerns raised in early MHD simulations. Positivity-preserving corrections and well-balanced adaptations improve behavior for low-density plasmas and geophysical flows governed by the Saint-Venant equations. Recent research couples Roe-based fluxes with discontinuous Galerkin and weighted essentially non-oscillatory schemes originated by Shu and Jiang.
The Roe solver has been applied across aerospace, astrophysics, and geoscience problems where hyperbolic systems dominate. In aerospace it supports transonic and supersonic wing analysis, nozzle flow design, and shock–boundary-layer interaction studies performed by companies and agencies in the aviation sector. In astrophysics it appears in supernova modeling, accretion disk simulations, and studies of interstellar medium turbulence used by observatories and university groups. In geophysical applications Roe-like fluxes are used for tsunami modeling, dam-break problems, and river hydraulics, where the Saint-Venant system captures flood wave propagation. Industrial examples include turbomachinery design, combustion chamber simulations, and blast-wave analysis relevant to research institutes and defense laboratories.
Despite strengths, the Roe solver has known limitations. It can violate the entropy condition at transonic rarefactions, producing non-physical expansion shocks unless an entropy fix is applied. It may fail to preserve positivity of density or pressure in near-vacuum or strong-shock situations, prompting positivity-preserving modifications. For MHD the standard Roe linearization can be non-unique and yield spurious solutions unless constrained by divergence-control methods associated with magnetic field solvers. Carbuncle instability and odd-even decoupling have been observed in high-Mach, grid-aligned shock simulations; remedies include hybrid fluxes, multidimensional dissipation, and careful mesh design used by computational groups confronting these artifacts.
Category:Numerical analysis Category:Computational fluid dynamics Category:Algorithms introduced in 1981