Dataset of "Impact of Particle Size Distribution on the Behaviour of Lithium-Ion Batteries under Dynamic Operation"

Physics-based electrochemical models are widely used to predict lithium-ion battery (LIB) behaviour under realistic operating conditions, where simplified laboratory profiles are insufficient. In this study, three models of different complexity were implemented and compared: the Doyle-Fuller-Newman (DFN) model, the Many-Particle Model (MPM) and the Many-Particle Doyle-Fuller-Newman model (MP-DFN). The models were parameterized using available datasets and adjusted to best represent a commercial LG M50LT cylindrical cell and validated against experimental data from constant-current/constant-voltage charging and discharge under the Worldwide Harmonised Light Vehicle Test Cycle (WLTC). The results show that while all models reproduce he overall voltage response, the many-particle formulations capture dynamic behaviour more faithfully, particularly diring WLTC discharge, where the particle size distribution strongly infulences predictive accuracy. These improvements are accompanied by highger computational demands, but the study provides new insight into the role of particle-level heterogeneity under realistic automotive conditions and offers guidance for selecting suitable modelling approcaches depending on application needs. The results of the individual models for charging and discharging correspond to the outputs of simulations carried out using the open-source Python Battery Mathematical Modelling (PyBaMM) package.