Machine Learning Based Aerodynamic Shape Optimization

Author

Noe Martinez Jr., The University of Texas Rio Grande Valley

Date of Award

5-2022

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Mechanical Engineering

First Advisor

Dr. Isaac Choutapalli

Second Advisor

Dr. Robert Freeman

Third Advisor

Dr. Horacio Vasquez

Abstract

The coefficient of pressure distribution for various 2D airfoil geometries were found using source – vortex panel methods. The data obtained in these simulations was used in multiple machine learning models which would predict the airfoil geometry from a given coefficient of pressure distribution. The neural networks employed were fully connected feedforward networks with Levenberg – Marquardt backpropagation and one model employed Bayesian Regularization. A novel tool for optimizing airfoil shape for a given coefficient of pressure distribution was created which performed well during testing. These models serve as the first step in minimizing the conflict between aerodynamic and stealth design for both low-speed and high-speed aircraft.

Comments

Copyright 2022 Noe Martinez, Jr. All Rights Reserved.

https://go.openathens.net/redirector/utrgv.edu?url=https://www.proquest.com/dissertations-theses/machine-learning-based-aerodynamic-shape/docview/2699728240/se-2?accountid=7119

Recommended Citation

Martinez, Noe Jr., "Machine Learning Based Aerodynamic Shape Optimization" (2022). Theses and Dissertations. 914.
https://scholarworks.utrgv.edu/etd/914