Search

Recent advances are presented in the development, implementation, and application of a novel solution adaptive method, the multiresolution (MR) scheme. The MR method uses the difference in information between adjacent grid levels on a set of nested grids for determining active and inactive grid cells. This amounts to computing the wavelet decomposition of the solution, which is known to be a rich source of regularity information. Shocks, contact discontinuities, reaction fronts, or any other inviscid and viscous flow features are identified and tracked in a time-accurate fashion. In active regions, the underlying finite volume scheme is solved in the usual manner, whereas in smooth regions, an inexpensive interpolation of the numerical divergence replaces both flux computations and reconstruction. Thus, the simulation becomes significantly more efficient without any loss of accuracy compared to the finest grid available. The combined finite volume-MR method will be described followed by two-dimensional and three-dimensional time-accurate examples on unstructured meshes

A 2-D computational study of flow over a fully open sunroof of a car with a passenger compartment has been done on a scalar workstation. The critical velocity for resonance in the passenger compartment has been determined. A wind deflector angle study has been done to determine an adequate angle to suppress the resonance. A validation case of computing the resonance condition of the flow over a 3:1 depth to length rectangular cavity has also been done. These resonance studies show the potential of using scalar workstations for a CFD design study.

The incompressible Navier-Stokes solvers INS3DUP, developed at NASA Ames and USA-IN3, developed at Rockwell Science Center, have been enhanced to simulate fluid-flow problems involving conjugate heat transfer (conduction and convection). The codes have been validated by comparing with experimental data, other computations and analytic solutions. The potential for using these codes in the analysis of component heat transfer of power assemblies is shown.

The physics that governs turbulence in recirculating flows is very different from that governing attached flow turbulence. This necessitates a modeling approach which attempts to convey the physics as faithfully as possible. This paper points out some of the known physics of recirculating turbulent flows and presents a model which addresses these physical issues. Results are shown for a variety of problems with moderate and large recirculation regions. The performance of the backflow model is demonstrated through comparisons with experimental data

Calculations of hypersonic turbulent flows over a spherically blunted cone, bicone and a cavity have been carried out using a modified Baldwin-Lomax zero-equation model and a new one-equation k-L model. A detailed evaluation of the models is presented by comparing computed kinematic eddy viscosity, length scale, skinfriction, heat transfer rates and wall pressure distributions with those obtained from experiments.

The USA-series of CFD codes are based on unified solution algorithms including explicit and implicit formulations, factorization and relaxation approaches, time marching and space marching methodologies, etc., in order to be able to solve a very wide class of CFD problems using a single framework. Euler or Navier-Stokes equations are solved using a finite-volume treatment with upwind Total Variation Diminishing discretization for the inviscid terms. Recently, these codes have been enlarged to also unify different aerothermodynamic options (perfect gas, real gas including equilibrium and nonequlibrium chemistry). This paper describes aspects of the finite-rate-chemistry capability. 27 references.