An Implicit Algorithm Suitable for Solving Viscous Flow Problems on Unstructured Mixed Element Grids

王刚,赵俊波,叶正寅
DOI: https://doi.org/10.3969/j.issn.1000-2758.2004.06.025
2004-01-01
Xibei Gongye Daxue Xuebao/Journal of Northwestern Polytechnical University
Abstract:Batina's implicit algorithm[1], and later that of Tomaro et al[2] are quite good; with suitable changes, Tomaro's algorithm can be made applicable to more difficult problems. Our implicit algorithm is different from theirs in three respects: mixed element grid, viscous flow, and introduction of adaptive local time stepping and residual averaging. In this paper, an implicit algorithm for solving the Navier-Stokes equations on three-dimensional unstructured mixed element grids is presented. The spatial discretization is based on the cell-centered finite-volume method. The first-order linearizing of flux vector and maximal eigenvalue splitting of flux Jacobi matrix are employed to construct an implicit Gauss-Seidel relaxation procedure. Adaptive local time stepping and residual averaging are used to accelerate convergence. In this paper, in order to check up the feasibility and efficiency of our implicit algorithm, we give two numerical examples of viscous flow problems around the ONERA M6 wing and DLR-F6 wing-body configuration respectively. For all calculations, Spalart-Allmaras one equation turbulence model is implemented. For each numerical example, comparison of convergence histories between our implicit algorithm and standard explicit 4-stage Runge-Kutta method are presented, which indicates that the convergence speed of implicit algorithm is several times faster than the explicit one's. For ONERA M6 wing configuration, the computed pressure distributions agree well with the experimental data, which show that the numerical method of this paper can well simulate the shock induced separation on the outboard portion of the wing. And for DLR-F6 wing-body configuration, numerical results show that our method has excellent accuracy in predicting the aerodynamics influence of pylon and nacelle. Our algorithm is particularly useful in handling complex flow problems such as shock-boundary layer interaction and interference flow between different aircraft components.
What problem does this paper attempt to address?