After writing my bachelor’s thesis in industrial mathematics on lithium batteries and their concentration-dependent parameters, I got the opportunity to work as a scientific assistant for Williams (Associate Professor in battery simulations).
My task was to investigate LFP–LTO batteries and how varying certain parameters would affect performance at different C-rates.
My job consisted of generating data from COMSOL simulations, analysing the results using Python, and writing a report in Overleaf.
The simulations were run using a pre-made model developed by Williams; I only had to make small adjustments to adapt it to the data we wanted to generate. All C-rates (0.5C, 1C, 3C, 5C, 10C) were divided into 5 time intervals, resulting in a total of 25 simulation files.
The simulations were performed on a virtual machine (Ubuntu, Linux). They turned out to be very large and computationally demanding, which led to several challenges. Because of this, I had to do a significant amount of troubleshooting, which resulted in a steep learning curve for both COMSOL and Linux.