Search Results

Wheels on passenger vehicles cause about 25% of the aerodynamic drag. The interference of rims and tires in combination with the rotation result in strongly turbulent wake regions with complex flow phenomena. These wake structures interact with the flow around the vehicle. To understand the wake structures of wheels and their impact on the aerodynamic drag of the vehicle, the complexity was reduced by investigating a standalone tire in the wind tunnel. The wake region behind the wheel is investigated via Particle Image Velocimetry (PIV). The average flow field behind the investigated wheels is captured with this method and offers insight into the flow field. The investigation of the wake region allows for the connection of changes in the flow field to the change of tires and rims. Due to increased calculation performance, sophisticated computational fluid dynamics (CFD) simulations can capture detailed geometries like the tire tread and the movement of the rim.

A wind-tunnel experiment investigating unsteady flow phenomena around a generic notchback during single crosswind gusts is modeled with the open source CFD package OpenFOAM®. The overall objective is to assess the capability and accuracy achieved by the simulation tool with respect to its potential for industrial usage. Transient yaw simulations apply a sliding interface between two computational grids, which are generated using the commercial software Spider®. It is shown that a stable simulation process is feasible but requires long computation times. The physical accuracy of the investigated phenomena depends on the computational grid and on the turbulence model used. Although the obtained aerodynamic loads qualitatively correspond with the experimental results, the absolute values are not satisfactory when working with a coarse grid with 6.2 million cells. Then, characteristic surface pressure distributions and their transient development differ from the experimental data.

In motorsports, aerodynamic development processes target to achieve gains in performance. This requires a comprehensive understanding of the prevailing aerodynamics and the capability of analysing large quantities of numerical data. However, manual analysis of a significant amount of Computational Fluid Dynamics (CFD) data is time consuming and complex. The motivation is to optimize the aerodynamic analysis workflow with the use of deep learning architectures. In this research, variants of 3D deep learning models (3D-DL) such as Convolutional Autoencoder (CAE) and U-Net frameworks are applied to flow fields obtained from Reynolds Averaged Navier Stokes (RANS) simulations to transform the high-dimensional CFD domain into a low-dimensional embedding. Consequently, model order reduction enables the identification of inherent flow structures represented by the latent space of the models.