Thermal study on a LTO cell module: Experiment, 3D numerical analysis and model order reduction

Abstract

Understanding and controlling the temperature evolution of Lithium-ion batteries is crucial to maintain high performance, ensuring long cycling life and avoiding thermal abuse. This paper presents a numerical and experimental thermal analysis of an air-cooled industrial module provided with 20 prismatic lithium-titanate-oxide cells. First, a 3D numerical model is presented for studying the dynamic distribution of the module temperature when the cooling fans are turned on or off. The numerical results are validated against test bench measurements. The flow field investigation explains the uneven temperature distribution among cells. The computation in natural convection inside the module, i.e. with fans off, was resolved by means of a fine empirical tuning. Building on the results of the 3D model, a 0D lumped model has been developed resorting to a model order reduction (MOR) technique and an energy balance differential equation. The model was characterized by tuning the module experimental data coming from a straight-forward testing protocol. The 0D MOR model, implemented on Simulink, demonstrated capable of quickly predicting the highest cell temperatures, allowing an easy and precise control of the module temperature, with an error $<$ 1 °C.