Computational Fluid Dynamics-based Design Optimisation of an Immersion Cooling Battery Thermal Management System

Effective design of battery thermal management system (BTMS) is essential to avoid system complexity, extra costs, and catastrophic battery failure. BTMS plays a significant role in prolonging the life span and ensuring the safety of lithium-ion batteries (LIB) in electric vehicles (EVs) and hybrid electric vehicles (HEVs). Accordingly, direct, also referred as immersion cooling is a potential solution which achieves a high heat transfer rate due to direct contact between the coolant and battery cells. In this work, an optimisation framework for LIB-based BMTS designs is investigated using biodiesel as a coolant by means of a combined computational fluid dynamic (CFD) with a surrogate modelling approach. Palm biodiesel is used as a dielectric coolant, proven to preserve LIBs within the temperature range 20 – 40 ℃, and prevent thermal runaway. The design of the BTMS is formulated in terms of two geometric and one operating design variables: inlet width, battery gap and coolant flow rate. Support Vector Regression (SVR)-based surrogate modelling approach using a Design of Experiment (DOE), and a permutation genetic algorithm is used to establish optimal process parameters.