Thermal Behavior of Automotive Lithium Batteries: Experimental Analysis and Digital Twin Development
Please login to view abstract download link
The lithium-ion battery is the key technology driving the development of the next era of transport electrification. In the future, most lithium-ion batteries will come from electric vehicles. A crucial aspect of the efficient use of these batteries is the optimization of the thermal management system. To this end, we have developed a numerical digital twin of the battery cell to characterize the thermal behavior of electric vehicle battery packs throughout their life cycle. The digital twin relies on Computational Fluid Dynamics (CFD) simulations to assess the performance of the external cooling system, integrating a model for internal thermal generation within the battery cell from Lagnoni et al. (2022, 2024) and simulating Conjugate Heat Transfer from within the cell to its external surface. The numerical digital twin of the battery cell is validated against wind tunnel experiments, considering both wind-off and wind-on conditions. The experimental setup used for the tests consists of a battery cycler based on a programmable charging system, specifically a 60 V–250 A Ametek SPS60x250-K02D, and a programmable discharging load, specifically a 60 V–500 A Zentro-Elektrik EL6000 Electronic Load. The two systems are remotely controlled via a GPIB standard interface using software developed in LabVIEW. Voltage and current are acquired through a data acquisition board, specifically a National Instruments DAQ 9219, which converts signals from analog to digital (ADC) with a maximum data acquisition rate of 100 ms. Temperature distributions are measured during charge and discharge cycles at constant C-rates using 30 K-type temperature probes, with a declared maximum expanded uncertainty of 0.3 K (for further information about the experimental setup, refer to Ceraolo et al., 2020). The digital twin effectively replicates the temperature gradients present on the battery with a satisfactory level of accuracy. The discrepancies between the thermal fields measured in experiments and those predicted by the numerical digital twin during charge and discharge cycles are less than a few percentage points. This accuracy allows the digital twin model to be used for performance assessment and optimization of the thermal management system for lithium batteries in automotive applications.
