Influence of contact resistance on thermal behavior of pouch-cell battery modules under partial direct liquid cooling: A numerical study

Direct liquid cooling (DLC) using dielectric fluids is emerging as a highly effective strategy for thermal management in high-performance lithium-ion battery systems, particularly under demanding operating conditions. However, most existing thermal models neglect heat generation from passive components and electrical contact resistances, which can significantly affect prediction accuracy during fast charging and discharging. This work presents a validated 3D multi-scale numerical model of a pouch-cell battery module cooled via a partial immersion DLC approach. The module, composed of four 60 Ah cells in a 2s2p electrical configuration and in a 1s4p hydraulic arrangement, is modeled using a multi-domain framework that integrates electrochemical and thermal phenomena. All model input parameters were experimentally measured in our laboratory, ensuring high physical fidelity. Importantly, the model incorporates ohmic heating in passive components and heat generated by contact resistance, factors often overlooked in existing literature. Validation against experimental measurements demonstrates high accuracy in predicting both transient and steady-state temperature profiles, including spatial temperature distributions within and between cells. Results reveal that passive component heating can momentarily account for up to 46 % of total heat generation under high C-rate charge-discharge cycles, while contact resistance contributes up to 12 % during semi-fast charging. These findings highlight the critical need to include these sources in thermal models to ensure accurate predictions and support design improvements. The proposed approach offers valuable insights for enhancing thermal performance, reliability, and safety of pouch-cell battery modules in electric vehicle applications. • Validated 3D multi-scale numerical model for pouch-cell modules with partial DLC • All model parameters experimentally measured for high physical fidelity • Passive component heating can reach 46 % of total heat at high C-rates • Inclusion of ohmic and contact resistance heating improves thermal prediction accuracy • 5.