Abstract
Vehicle aerodynamics plays an important role in reducing fuel consumption. The underbody contributes to around 50% of the overall drag of a vehicle. As part of the underbody, the wheels and wheelhouses contribute to approximately 25–30% of the overall drag of a vehicle. As a result, wheel aerodynamics studies have been gaining popularity. However, a consensus of an appropriate turbulence model has not been reached, partially due to the lack of experiments appropriate for turbulence model validation studies for this type of flow. Seven turbulence models were used to simulate the flow within the wheelhouse of a simplified vehicle body, and results were shown to be incongruous with commonly used experimental data. The performance of each model was evaluated by comparing the aerodynamic coefficients obtained using computational fluid dynamics (CFD) to data collected from the Fabijanic wind tunnel experiments. The various turbulence models generally agreed with each other when determining average values, such as mean drag and lift coefficients, even if the particular values did not fall within the uncertainty of the experiment. However, they exhibited differences in the level of resolution in the flow structures within the wheelhouse. These flow structures are not able to be validated with currently available experimental data. Properly resolving flow structures is important when implementing flow-control devices to reduce drag. Results from this study emphasize the need for spatially and time-resolved experiments, especially for validating large eddy simulation (LES) and detached eddy simulation (DES) for flow within a wheelhouse.
