Toward the Usage of Deep Learning Surrogate Models in Ground Vehicle Aerodynamics

Abstract

This study introduces a deep learning surrogate model designed to predict the evolution of the mean pressure coefficient on the back face of a Windsor body across a range of yaw angles from 2.5∘ to 10∘. Utilizing a variational autoencoder (VAE), the model effectively compresses snapshots of back pressure taken at yaw angles of 2.5∘, 5∘, and 10∘ into two latent vectors. These snapshots are derived from wall-modeled large eddy simulations (WMLESs) conducted at a Reynolds number of ReL=2.9×106. The frequencies that dominate the latent vectors correspond closely with those observed in both the drag’s temporal evolution and the dynamic mode decomposition. The projection of the mean pressure coefficient to the latent space yields an increasing linear evolution of the two latent variables with the yaw angle. The mean pressure coefficient distribution at a yaw angle of 7.5∘ is predicted with a mean error of e¯=3.13% when compared to the WMLESs results after obtaining the values of the latent space with linear interpolation.

Investigation on the Reduced-Order Model for the Hydrofoil of the Blended-Wing-Body Underwater Glider Flow Control with Steady-Stream Suction and Jets Based on the POD Method

“…The core idea of the POD method is to find the optimal standard orthogonal basis functions in the mean-square sense from a set of time-series spatial data, based on the specified information of the sampled flow field data, which usually come from the experimental or numerical simulation results. As a result, fewer orthogonal basis expansions are used to approximate the description of the higher-order data, ultimately enabling the reconstruction and prediction of the flow field [25,26].…”