Drag Reduction Study on Vehicle Shape Optimization Using Gradient-based Adjoint Method

Abstract

<div class="section abstract"><div class="htmlview paragraph">Given the increasingly stringent environmental regulations, most automotive manufacturers were confronted with tougher exhaust emission and energy consumption standards, thus, improving fuel economy has been the top priority for OEMs during the past few years. In this context, it is quite essential to improve the aerodynamic characteristics, especially drag reduction in vehicle shape development, considering its close correlation with fuel consumption and E-range. Of all the optimization approaches, the gradient-based adjoint method has currently received growing attention for its proven effectiveness in calculating the drag sensitivity with respect to geometry parameters, which is indispensable for subsequent shape modification. In this work, we aim to utilize the adjoint approach to optimize the vehicle shape for a lower drag on the DrivAer models. First, the exterior flow field was solved as the primal solution; then, surface drag sensitivity was calculated by adjoint solver, which helps to confirm the most drag-sensitive regions need to be modified in the next step. After that, the surface mesh was morphed based on the Normal Optimal Displacement results; and then the flow field of the new design was resolved to validate the drag reduction effect. Results showed that it’s around 24 counts of C_D decrease for the Notchback, which means 9% drag optimization. This paper evaluated the effectiveness of adjoint method for aerodynamic optimization of a generic vehicle, which indicates more extensive and promising application of this approach in the early stage of vehicle development for its high efficiency as well as strong robustness.</div></div>