Application and validation of HUNS3D flow solver for aerodynamic drag prediction cases

Abstract

Computational fluid dynamics (CFD) simulations are extensively used in scientific research and in the design and analysis phase. In order to establish confidence and quantify the results of these simulations, verification and validation process should be conducted. In the present work the validation process for in-house CFD code HUNS3D (Hybrid Unstructured Navier-Stokes 3D) was carried out by simulating drag prediction cases and comparing results with the experiment and other CFD solvers. This procedure will ensure the solver accuracy and its ability to replicate the flow physics. Firstly, two dimensional test cases selected from AIAA Drag Prediction workshop were simulated. NACA-0012 airfoil which usually has a blunt trailing edge is slightly modified so that a sharp trailing edge is obtained. For grid convergence study, a total of five grids were investigated. From coarsest-to-finest the grids have the dimensions of 113 × 33 - 1793 × 513 cells respectively. The grid aspect ratio was kept constant for all the cases. An essential requirement for these test cases was to run them at incompressible condition (M = 0.15) and at Reynolds number of 6 million. As part of a convergence study, asymptotic convergence analysis was also performed. Next the study was extended for drag prediction of DLR-F6 Wing-Body-Nacelle-Pylon (WBNP) configuration. The lift and drag values were calculated at transonic flow conditions by solving RANS equations on unstructured hybrid grid. Three different test cases were performed by changing the angle of attack and the results were compared with the available experimental data. The simulations were converged to a value of 1e-8 for residuals. The predicted drag value was found in close agreement when compared with available results. The static stall simulation for NACA-0012 resulted in an overall difference of less than 8% for lift coefficient, when compared with the experimental data. For purpose of validating turbulence models, the simulations were conducted at different angles of attack (0°, 10°, 15°). Lift and drag coefficient were calculated by employing Spalart-Allmaras (SA) turbulence model and Shear Stress Turbulence (SST) turbulence model. These results were compared with those obtained by CFL3D solver and maximum difference of 6% was found. For DLR- F6 configuration the predicted drag values compares well with other CFD solver and experimental data showing consistent trend and comparable wing pressure distributions.