Realistic Modelling of Li-ion Batteries for Calculating High Frequency Impedance

Abstract

Abstract This thesis primarily focuses on building a robust 3-D simulation model and implementing a versatile solver setup for a Lithium-ion (Li-ion) battery cell to study the High-Frequency (HF) behavior. Although this study primarily focuses on the impedance characteristics of battery cells, it also investigates their multidisciplinary behavior, including electromagnetic, thermal, and structural aspects. The developed simulation model is initially validated using a simple copper wire and a coin cell battery before being applied on the cylindrical cell model. The 3-D model includes all cell layers with defined dimensions and tolerances, and is finely meshed in critical regions to ensure convergence and analyze mesh dependency. The primary outcome of this study is the impedance as a function of frequency, swept across a range from 100 Hz to 100 MHz. The simulated data is compared with experimental test data of the same cell. This comparison helps understand the accuracy of the solver and study possible failure points. Another key outcome of this study is the development of a genetic algorithm (GA)-based computational framework designed to identify optimal solutions for cell design. This helps in balancing multiple user-defined objectives to achieve an efficient design solution with balanced trade-off among competing objectives. The conclusion of the thesis is that a 3-D simulation model gives a better estimation of the real-world HF behavior, than a 1-D model which has numerous assumptions on the material behavior and field propagation. The studied Li-ion cell indicates a drop in impedance behavior in the mid-frequency range, attributed to dominant capacitive and resistive effects. At higher frequencies, beyond the cutoff frequency, an increase in impedance is observed due to dominant inductive behavior. This model is used to visualize field propagation and can be further scaled to the module or pack level to analyze and improve the EMC performance of the system. However, with further modifications, the developed simulation model can be better optimized for detailed analysis of multidisciplinary parameters such as temperature, Joule heating, displacement, and stress.