Aerodynamic investigation of wheel and wheelhouse flows

Rising fuel costs and the destructive impact of ground vehicles on the environment has driven the need for car manufacturers to reduce fuel burn and emissions, where the goal is to develop vehicles to be more efficient than previous generations. It is universally accepted that a low drag is a prerequisite for a fuel efficient vehicle, with the main contributor of drag force generated by flow separations dictating pressure differences between the front end and rear. With wheels and wheelhouses proven to account for up to 30% of a vehicle's drag, it has therefore become essential to acquire a deeper understanding of how the development of the flow field effects the exhibited forces. Thus, the aim of this research was to improve on the existing knowledge of the complex flow features found around a wheel and wheelhouse and to examine how modifications to the configuration affects these features and the production of drag.

To confirm the robustness of the numerical methods used in capturing the complex flow phenomena and justify the use of a stationary ground plane, two validation studies were performed. The first investigated the aerodynamic behaviour of the flow around a rotating and stationary 60% scale isolated wheel, with and without the use of a moving ground plane. A bespoke rotating wheel rig was designed and manufactured at Loughborough University with wind tunnel tests performed over a range of pre to post critical Reynolds numbers. Force coefficients were calculated using balance measurements and flow field data were obtained using Particle Image Velocimetry (PIV) where the unsteady flow field data generated was used to validate unsteady CFD predictions. These were performed using STAR-CCM+ and a k-omega SST turbulence model. Improved Delayed Detached Eddy Simulation (IDDES) was shown to outperform other models by capturing an increased amount of finer detailed, high frequency vortical structures. The CFD showed good agreement with the experimental results thus presented a validated numerical methodology. When comparing stationary to rotating configurations, both methodologies illustrated large scale structural differences in the surrounding flow due to changes in separation and wake structure and were corroborated by existing literature. Importantly, the CFD showed minimal difference between a stationary and moving ground plane simulations with a rotating wheel. This is evidence that, provided the wheel is rotating, valid research or preliminary design experiments can be performed without the complexity of a moving ground plane.

In continuation, an accompanying quarter car wheelhouse structure with a fully-pressure-tapped wheelarch was also designed and manufactured at Loughborough University for the second validation study as well as further testing where both wheel and wheelhouse structures were used in combination. Similar to the previous study, experimental techniques were used with balance, PIV and additional pressure data acquired and compared to computationally obtained results. The CFD again showed good agreement with all trends of the experimental data, including the variation in pressure exhibited upon the wheelarch, inferring that successful capture of the complex flow features found within a wheelhouse cavity had been achieved.

Once confidence was ensured in the numerical methods used, further tests which investigated different configurations of the wheel and wheelhouse model were then undertaken. This included changing the lateral displacement of the wheel, where results showed that up to a 10 mm displacement outboard of the housing, overall drag decreased. The reduction in housing drag was attributed to a reduction in the size of outboard longitudinal vortex structures narrowing the width of the outboard shear layer. For the wheelhouse overhang study, possible due to the front overhang of the housing structure being manufactured to allow modularity in length, significant drag benefits were found to be achievable with even the minimum of overhang extensions tested, CD = -0.021. The rate of change of reduction was shown to reduce the longer the overhang length.

Finally, designs were presented which simulated wheel and wheelhouse flows that better represented real world geometries, such as applying a radius on the wheel and the inclusion of wheel deflectors and housing ducts to alleviate wheelarch pressure. Wheel deflectors, widely accepted to work at reducing the drag generated by wheels due to the movement of the stagnation region, was shown in this study to not only divert air around wheels, but also beneficially re-balance the entire cavity flow regime, reducing total drag by up to CD = -0.012 and the drag generated by the wheelarch by half.