Dissertation Defense - Miu Lun Lau

Automated Materials Spectroscopy and Topology Optimizations using Genetic Algorithms

Presented by Miu Lun Lau
Computing PhD, Computational Mathematics Science and Engineering emphasis
Attend via Zoom

Committee

Dr. Min Long (Advisor), Dr. Edoardo Serra, Dr. Michal Kopera

Abstract

This dissertation will focuses on implementation and development of genetic algorithm software for increasing the productivity and analysis rate for spectrum analysis, topology optimization and other. The software artifacts generated has resulted in a basic software framework calls \textit{Neo}. The framework has been further developed into different software for specific spectra method such as EXAFS, Astrophysics, Nano-Indentation, and many more.

For GA implementation that relates to EXAFS, we have demonstrated the usage of GA to correctly identify chemical species presented, for Copper metal species, Technetium compounds and in-situ of SnS2 batteries. The software artifact was able to generate accurate and physical reasonable EXAFS fittings that rapidly accelerates the analysis process.

For the Nano-indentation, our software also demonstrates better fits for graphite related material compared to conventional LSF fits. Additionally, our software has the abilities to rapidly analyze a number of indentation mapping, as shown in High Entropy alloy (HEA) result.

For the Heat exchanger optimization, we demonstrate the ability to optimize the shape of the heat exchanger based on the design parameter and the underlying physics informed model. The model is supplemented by adsorption parameters calculated in molecular dynamics.