Xujian Cui

Battery thermal management system (BTMS) has significant impacts on the performance of electric vehicles (EVs). In this research, a computational fluid dynamics (CFD) coupled multi-objective optimization framework is proposed to improve the thermal performance of the battery pack having metal separators. CFD is utilized to study the thermal and fluid dynamics performance of the designed battery pack. Input parameters include inlet air temperature, thermal conductivity of coolant, thermal conductivity of metal separator, and diameter of heat dissipation hole. Five vital output parameters are maximum temperature, average temperature, temperature standard deviation (TSD), maximum pressure, and volume of the pack. The support vector machine (SVM) model is used to replace the real output parameters of the battery pack. Sensitivity analysis results indicate that the diameter of heat dissipation hole is the main factor affecting the volume of the structure and the pressure drop, while the inlet air temperature has significant influence on the battery pack thermal behavior. The cooling efficiency and the uniformity of temperature distribution are mainly determined by the inlet air temperature. The decrease of inlet air temperature could lead to a rise of temperature standard deviation. The nondominated sorting genetic algorithm-II (NSGA-II) is taken to acquire the optimum set of input parameters. The obtained optimal scheme of battery pack can improve the cooling efficiency as well as reducing the volume cost and the energy consumption of the cooling system while such design may result in a higher level of nonuniformity of the temperature and pressure distribution.

An effective battery thermal management system (BTMS) is essential to ensure that the battery pack operates within the normal temperature range, especially for multi-cell batteries. This paper studied the optimal configuration of an air-cooling (AC) system for a cylindrical battery pack. The thermal parameters of the single battery were measured experimentally. The heat dissipation performance of a single battery was analyzed and compared with the simulation results. The experimental and simulation results were in good agreement, which proves the validity of the computational fluid dynamics (CFD) model. Various schemes with different battery arrangements, different positions of the inlet and outlet of the cooling system and the number of inlets and outlets were compared. The results showed that an arrangement that uses a small length-width ratio is more conducive to promoting the performance of the cooling system. The inlet and outlet configuration of the cooling system, which facilitates fluid flow over most of the battery pack over shorter distances is more beneficial to battery thermal management. The configuration of a large number of inlets and outlets can facilitate more flexible adjustment of the fluid flow state and can slow down battery heating to a greater extent.

Electric Vehicles (EVs) are estimated as the most sustainable solutions for future transportation requirements. However, there are various problems related to the battery pack module and one of such problem is invariable high-temperature differences across the battery pack module due to the discharging and charging of batteries under operating conditions of EVs. High-temperature differences across the battery module contribute to degradation of maximum charge storage and capacity of Li-ion batteries which ultimately affects the performance of EVs. To address this problem, a Finite Element Modelling (FEM) based Automated Neural Network Search (ANS) approach is proposed. The research methodology constitutes of the four stages: Design of air-cooled battery pack module, setup of the FEM constraints and thermal equations, formulating the predictive model on generated data using ANS and lastly performing multi-objective response optimization of the best fit predictive model to formulate optimum design constraints for the air-cooled battery module. For efficient thermal management of the battery module, an empirical model is formulated using the mentioned methodology for minimizing the maximum temperature differences, standard deviation of temperature across the battery pack module and battery pack volume. The results obtained are as follows: (1) The battery pack module volume is reduced from 0.003279m³ to 0.002321m³ by 29.21%, (2) The maximum temperature differences across the eight cells of battery pack module declines from 6.81K to 4.38K by 35.66%, and (3) The standard deviation of temperature across battery pack decreases from 4.38K to 0.93K by 78.69%. Thus, the predictive empirical model enhances the thermal management and safety factor of battery module.