Verification Study of a CFD-RANS Code for Turbulent Flow at High Reynolds Numbers

This article surveys the verification and validation methods of the in-house codes. A practical aspect of verification and validation procedure of an in-house code is provided. In order to solve Reynolds-averaged Navier-Stokes (RANS) equations of the compressible flow field, a computational fluid dynamics (CFD) computer program is developed. NASA Langley's CFL3D and FUN3D, flow solvers are adopted to verify and validate the developed code. Two test cases are studied; two-dimensional (2-D) zero pressure gradient flat plate and non-zero pressure gradient 2-D NACA 0012 airfoil, each which uses 5 systematically refined meshes. First, the code and solution verification for both cases are investigated. The observed order of accuracy and numerical uncertainty results agree with those of CFL3D, FUN3D, Cobalt, and RavenCFD flow solvers. For model validation, an experimental data is used as well. The predicted skin friction coefficient over the turbulent flat plate surface shows good agreement with the experiment and with the flow solvers. Furthermore, skin friction, pressure, drag, and lift coefficients for the 2-D NACA 0012 airfoil at three angles of attack ( alpha =0o, alpha =10o, and alpha =15o) are estimated. The results agree well. The presented evidences indicate that the code is verified and validated correctly.