Connor Miller

Author: Connor Miller | Major: Mechanical Engineering

My name is Connor Miller, and I am a recent graduate of the College of Engineering with a bachelor’s degree in mechanical engineering. For the spring 2022 semester, I continued my research that I began with Dr. Han Hu, Assistant Professor in the department of Mechanical Engineering, in which I used generative design to create heat exchangers that utilized more complex geometries that are just recently available to be manufactured because of technological advancements in additive manufacturing in hopes of designing a heat exchanger that can perform as well as, if not better than, heat exchangers using traditional designs.

This semester, I used what I learned last semester about the nTopology platform to generate multiple designs of heat exchangers. I was given a heat exchanger by Dr. Hu that was to act as my control heat exchanger that I would compare the results of the thermal analysis from my heat exchangers. I modeled the control heat exchanger in both Solidworks and nTopology and ran similar thermal analysis which yielded similar results. This was done to verify the validity of the analysis done in nTopology as well as gain a better understanding of the processes nTopology used in their thermal analysis. Now that I knew nTopology was a reliable source for my thermal analysis, I started designing heat exchangers with the more complex geometries. The geometry used for the cooling portion of the heat exchanger was a walled triply periodic minimal surface, or walled TPMS for short. The walled TPMS structure used a gyroid fill type to make its unique design. I made the cooling portion the same volume as the fins on control heat exchanger and made the base the exact same as well. This was so that when comparing results, the cooling structure is what would determine how well the heat exchanger, not its dimensions.

After the first walled TPMS heat exchanger was designed, I generated new designs by adding a ramp modifier that would adjust the thickness of the cooling structure based on the thermal analysis of the heat exchanger before it. After a few generations, I had a heat exchanger that performed very similarly to the control exchanger, however because of time constraint, I was not able to continue and reach a heat exchanger that outperformed the control. However, when plotting the surface area of the walled TPMS heat exchangers verses their thermal resistance, the trendline showed that a walled TPMS heat exchanger with a similar surface area would outperform the control heat exchanger.

The biggest challenge I faced through this entire process was the handling the computational requirements needed to run the thermal analysis for the heat exchangers. My mentor helped a lot by giving me access to computers with better computational power than the computers I was originally using. However, because of the complex geometries, running a thermal analysis would in some instances would still take hours to complete, even with more powerful computers.

I found out about this project when I started looking for research opportunities last spring through the Mechanical Engineering department. I was provided a list that had several different projects, and this is the one that caught my eye, so I reached out to Dr. Hu for more information. When he asked why I wanted to get involved in with this project, I mentioned that the way technology is evolving and how machine learning is being used more and more in the world was something I was interested in learning about. Also, I mentioned that when I graduate, I want to work as a design engineer, and by learning how to leverage machine learning in the design process, I feel I could become an asset to any future company that I work for. Since then, he has helped keep me on track with this project and was the one who suggested I use nTopology to help along the way.

This project has given me the opportunity to work on a project that was initially outside of my range of knowledge, so, while stressful, it is super exciting and rewarding to see how far along it has come. Unfortunately, I will not be continuing with this project and instead will be starting a full time Mechanical Engineering position with Lockheed Martin out of Camden, AR. I would like to take the opportunity to again thank Dr. Hu, the Mechanical Engineering department, the College of Engineering, and the Honors College for their unwavering support throughout this entire experience.