Articles techniques et présentations

Battery packs connect electrochemical cells in series and parallel to meet target energy and power specifications. At the pack level, performance and lifetime are governed by coupled electrical, thermal, and aging heterogeneities that arise from nonuniform current distribution, thermal gradients, and intrinsic cell-to-cell variability. These inhomogeneities intensify over repeated cycling, driving divergent degradation pathways and ultimately reducing pack reliability. Understanding how pack electrical topology amplifies or suppresses these effects is critical for lifetime-optimized pack design. This work presents a COMSOL Multiphysics modeling framework to investigate the influence of electrical topology on the long-term evolution of cell-level inhomogeneity in battery packs. The model is built within COMSOL's Battery Pack Interface, where a reduced-order electrochemical model captures cell voltage response accounting for ohmic, activation, and concentration overpotentials relative to the open-circuit potential. A phenomenological aging submodel is further coupled to track capacity fade as a function of each cell's local current, voltage, and temperature histories over cycling. In the Battery Pack Interface, electrical interconnections between cells are explicitly represented to capture current redistribution dynamics across the pack. This electrochemical model is coupled to a thermal model, implemented using the Heat Transfer Interface to resolve spatial temperature distributions arising from Joule heating under pack-level thermal boundary conditions representative of realistic cooling strategies. The model is formulated to enable simulation of different pack configurations, including series, parallel, and hybrid series-parallel topologies. Electrical interconnections are explicitly represented to capture current redistribution effects. The coupled system of equations is solved to obtain the time evolution of current density, temperature, and state of health at the cell level across the pack. Model validation is performed in two stages. In the first stage, single-cell voltage, current, and temperature predictions are validated against first-cycle experimental measurements, followed by validation of the predicted capacity fade trajectory against long-term experimental cycling data. In the second stage, pack-level voltage, pack current, and cell-resolved temperature profiles are validated against pack-level experiments, and simulated pack aging is directly compared to experimentally measured pack degradation. Good agreement across both stages confirms the fidelity of the coupled multiphysics approach. This work demonstrates the capability of COMSOL Multiphysics to quantify topology-induced nonuniformities in battery packs and provides a simulation-based framework for optimizing pack architecture with respect to lifetime, reliability, and thermal management. The approach is extensible to more complex pack geometries and next-generation cell chemistries.