[Thesis] Gradient-Free Aeroacoustic Shape Optimization Using High-Order Large Eddy Simulation

In this video, my Ph.D. thesis on developing an aeroacoustic shape optimization framework is explained. The framework utilizes the in-house high-order unstructured solver (HORUS) for flow simulation, based on the high-order flux reconstruction (FR) approach. The optimization algorithm used is the gradient-free mesh adaptive direct search (MADS). For far-field aeroacoustic shape optimization, the Ffowcs Williams and Hawkings (FW-H) formulations are applied.

The framework is parallelized, making the runtime of the optimization process independent of the number of design parameters, significantly improving efficiency. This technology has been tested on various problems, demonstrating its potential for designing the next generation of aircraft.

This explanation is simplified for undergraduate students, avoiding detailed mathematics. For comprehensive details, refer to the thesis and published papers.

Ph.D. Thesis: https://spectrum.library.concordia.ca...

Near-field aeroacoustic shape optimization at low Reynolds numbers
AIAA Journal: https://arc.aiaa.org/doi/abs/10.2514/...
Pre-print on arXiv: https://arxiv.org/pdf/2310.10574

Gradient-Free Aeroacoustic Shape Optimization Using Large Eddy Simulation
Pre-print on arXiv: https://arxiv.org/abs/2312.14167