Spectral/hp element methods for industrial applications: flow past inverted wings in ground effect

Abstract

Computational Fluid Dynamics (CFD) is the basis of most modern industrial aerodynamic pipelines. With increased reliance on CFD processes, the demands on the accuracy and fidelity of the codes, grow hand-in- hand. When considering complex industrial flows, like those considered in automotive and motorsports applications, the accuracy and fidelity of traditional turbulence models like Reynolds Averaged Navier Stokes and Detached Eddy Simulation and the efficiency of legacy low- order spatial discretisations come into question. High-order techniques, like Spectral/hp Element Methods, offer distinct advantages in computational efficiency, hoping to bridge the gap and make higher-fidelity scale-resolving techniques such as Large Eddy Simulation more accessible to the average engineer. This thesis is presented in two parts. The first part investigates the current state-of-the-art of Spectral/hp Element methods. It uses industrial benchmarks including the SAE Notchback body and the Imperial Front Wing to highlight some of the restrictions of the methodology towards wider industrial applications. Some of the restrictions highlighted include the meshing intricacies in mixed h/p approaches, geometric and meshing issues, purely explicit advection schemes, stabilisation and algorithmic constraints including preconditioning. The second part looks at the benchmark cases. It focuses on inverted wings in ground effect. A single element in ground effect, the GA(W)-1 profile was studied with and without freestream forcing at 0,3,5 degree angles of attack. It was noted that the freestream forcing changed the transition mechanism - highlighting the need for freestream noise in resolved simulations. Finally, a multi-element front wing was studied utilising the Imperial Front Wing geometry at a ride height of 0.36h/c initially. Intergral quantities and spatial distributions, along with studies investigating the spanwise width using auto-correlation and the interplay between transition mechanisms using cross-spectral phase presented. Ride height was then swept and frequency analysis presented showing the implications on key flow structures.