Numerical investigation of unsteady flow past rudimentary landing gear using DDES, LES and URANS

Massively unsteady separated flow past a four-wheel rudimentary landing gear is computed based on three different numerical approaches. The methods include Delayed Detached Eddy Simulation (DDES) and the Unsteady Reynolds-Averaged Navier-Stokes (URANS) method based on the two-equation Shear Stress Transport (SST) model as well as Large Eddy Simulation (LES) based on the dynamic model. Using computational fluid dynamics software ANSYS (R) CFX (R), surface features of both the mean and unsteady flows are studied and compared with experimental data, such as time-averaged pressure, surface flow patterns, sound pressure level and so on, while flow field characteristics like instantaneous vorticity and turbulent kinetic energy are obtained to assess the quality of different numerical methods. The accuracy of DDES in predicting the landing gear flow is assessed both aerodynamically and acoustically from an engineering point of view. As expected, URANS can predict the attached flow near the wall well, but fails to obtain reasonable fluctuations in detached regions. Owing to poor near-wall grid resolution, LES predicts some non-physical separation in the area where the flow was originally attached, which adds superfluous turbulence fluctuations. The results of DDES have the advantages of both, and are in good agreement with the experimental results, which characterize the unsteady properties of the flow better. The feasibility of the CFX-DDES method is demonstrated in predicting the unsteady flow for landing gears with moderate grid scales. What's more, because of its numerical robustness and low dissipation, DDES also obtains better nearfield noise distribution which shows the potential in noise prediction for engineering applications. Additionally, a detailed analysis aiming at exploring the troublesome mechanisms of noise source generation is also exhibited using DDES. It shows the possibility that strongly unsteady interactions between vortices and landing gear structures may contribute to noise generation.