Search Results

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.

Although considerable efforts have been made with respect to the reduction of fuel consumption of trucks during the last decades, the diminishing natural resources as well as the evolution of the truck traffic require continuous improvements in the field of aerodynamics. Indeed, the forces generated by the air on the trucks may originate, depending on weather, road type, truck type, dimension, etc., up to 50% of the fuel consumption. In order to analyze the influence of proportion variations (mainly related to the length) and add-on devices on the aerodynamic performance of a truck, a representative model was first generated. This simplified geometry of a tractor-trailer was based on the geometrical data of six European OEMs: Daimler, Iveco, and MAN (tractors), Kögel, Krone and Schmitz Cargobull (trailers). The model included a reduced level of details (exterior mirrors, wheels, simplified underbody and engine block).