This conference paper was submitted for presentation at the NAFEMS World Congress 2025, held in Salzburg, Austria from May 19–22, 2025.

Abstract

This study investigates the virtual optimization of battery modules, focusing on both manufacturing processes and operational performance. The physical connection between the battery cell, the cooling system, and the housing is crucial for optimizing the overall performance of the battery pack. Therefore, a thorough examination of the adhesive used to bond these components is essential. Manufacturing Process Optimization: Battery cells are bonded to the housing and a carrier plate, which often includes an integrated water-cooling circuit. The bonding adhesive significantly impacts cell heat dissipation, module strength, and retention. This process is simulated using the commercial CFD software PreonLab, which is based on the Smoothed Particle Hydrodynamics (SPH) method. Through simulation, production time and adhesive distribution can be optimized. The investigated virtual manufacturing process includes: Applying adhesive to the panel '“ assessing the time required for the process and adhesive distribution. Applying the cells to the plate '“ evaluating the forces on the cells and the final distribution of the compressed gap filler. Different adhesive application variants are tested virtually, resulting in varied outcomes for production time and thermal conduction between the cooling plate and cells. Battery Performance Optimization: The charging time (e.g., 20% to 80% SOC) is often used to assess the performance of electric vehicles. In this study, the impact of different gap filler variants on the fast-charging behavior of battery modules is evaluated using AVL CFD software AVL FIRE M. Through the spatial resolution in CFD, we enable monitoring localized thermal and electrochemical effects, ensuring optimal battery health management and fast-charging performance. Using a physical, detailed 3D battery model as a virtual twin, the behavior of each battery cell and the entire system was analyzed. This investigation reveals how different bonding options affect cell temperatures, fast-charging behavior, and thermal management. It demonstrates that simple 1D system simulations with idealized models are insufficient for accurately predicting fast charge behavior in complete battery packs. Conclusion: This investigation provides valuable insights into optimizing production processes and predicting fast-charging times in battery packs. By employing CFD methods, we enhance the efficiency of thermal management systems through spatial resolution, enabling precise monitoring of localized effects. Additionally, the use of virtual development techniques significantly reduces costs and streamlines the early stages of development through virtual testing and optimization.