Performance and prediction of baffled cold plate based battery thermal management system

Abstract

• Heat transfer performance of cold plate with longer and more baffles is better. • Performance of baffled cold plate can be further improved with the increase of s - L . • Requirements for BTMS operating parameters under different conditions are revealed. • Characteristics of baffled cold plate and BTMS can be predicted by neural network. The development of NEV is restricted by the problem of battery thermal safety. BTMS should be lightweight, compact, and efficient. Baffled cold plate with small aspect ratio channels was designed for BTMS. The influence of structure and operating parameters on baffled cold plate and BTMS performance were explored through ANSYS simulation. BP neural network model was established to study the function of performance prediction and application in the fields of BTM. The results show that the comprehensive heat transfer performance of baffled cold plates with longer and more baffles is better. Comprehensive heat transfer performance and temperature uniformity of baffled cold plate can be improved through local connection method, and the optimization effect is enhanced with the increase in left local slit spacing. Under low discharge rate conditions, it’s easier to meet the needs of BTM. When the battery is discharged at 3 C rate, volumetric flow rate needs to reach 1 mL·s -1 at all inlet temperature conditions. Through data processing and modeling, BP neural network predictions of flow and heat transfer characteristics of baffled cold plate and temperature control performance of BTMS are proven accurate. It facilitates the design and optimization of baffled cold plates, and the design and control strategy formulation of BTMS.