Off-campus Michigan Tech users: To download campus access theses or dissertations, please use the following button to log in with your Michigan Tech ID and password: log in to proxy server

Non-Michigan Tech users: Please talk to your librarian about requesting this thesis or dissertation through interlibrary loan.

Date of Award


Document Type

Campus Access Master's Thesis

Degree Name

Master of Science in Mechanical Engineering (MS)

Administrative Home Department

Department of Mechanical Engineering-Engineering Mechanics

Advisor 1

Zequn Wang

Committee Member 1

Kai Zhou

Committee Member 2

Shangyan Zou


Global warming has led to increased research in renewable energy and the need for efficient energy storage systems. Lithium-ion batteries are a promising solution, but their performance degrades at high temperatures. To improve thermal management, researchers are exploring the use of phase change materials (PCMs) combined with fin structures. Different fin geometries impact heat dissipation. The goal of this study is to perform a reliability-based design optimization of a battery thermal management system for a desired reliability and temperature level. The design geometry consists of four components that include the lithium-ion cell at the core having a fin structure with a PCM module attached to it, and an acrylic shell on the outside. The geometric design variables include the dimension of the outer radius of the battery shell (overall diameter of the battery) and three dimensions of a T-shaped fin structure. Along with the four design variables, two uncertainty parameters of battery heat generation that happens at the core and the ambient convective heat transfer coefficient on the outer surface are considered for the reliability based design optimization. Latin Hypercube Sampling is used to generate sample points for thermal analysis that is done using ANSYS Mechanical APDL. These data points are used to train a machine learning model to predict temperatures for unknown design samples during the optimization process. The optimization is done using a type of an evolutionary algorithm. Initially the optimization problem was formulated using a single objective function that was minimized to find the optimal design configuration. The results of this optimization encouraged to pursue the ix possibility of multiple optimal solutions and formulate a multi-objective optimization problem.