Abstract
Inhomogeneity in Li-ion cells leads to an underutilization of its energy and non-uniform degradation. In our previous work, the proof of concept was established for a setup with 4 tiny Varta cells connected in parallel, enabling to decouple the effect of the two parameters influencing inhomogeneous behavior. This is to mimic the thermal and potential gradients occurring in large-format commercial Li-ion cells. To represent the thermal gradient, the cells are placed at different temperature and is called T-dep setup; to represent the potential gradient, the cells have different shunt resistors connected in series to each of the tiny cells in the setup and is called V-dep setup. In this work, the above setup is used to perform long-term accelerated cycling tests (about 1100 cycles under 100 % Depth of Discharge, DOD) and their State of Health (SOH) is analyzed at periodic intervals. Two different degradation phenomena are primarily observed, namely reduction and oxidation of the solvent occurring at negative and positive electrodes, respectively. The solvent reduction results in a continuous growth of a passivation layer called Solid Electrolyte Interface (SEI). The coupled electrochemical model, from our previous work, is integrated to an aging model in this work. The model is parameterized using preliminary cycle aging data performed on individual pristine cells. The model captures the experimentally observed cell capacity fade trends, in both setups, quite well. A comparison of the cell interfacial resistance increase reveals that the model captures the experimental behavior quite well until about 400 cycles and underestimates after this point. Using this validated model, a comparison of capacity fade and impedance increase, of cells in setup and individually cycled cells, reveals that the setup conditions do not necessarily aggravate the inhomogeneous aging. Further analyses show that there is a reversal of cell utilization upon cycling, meaning degradation of less aged cells seems to accelerate compared to their more aged counterparts. Additional simulation analyses under narrow DOD cycling indicate that overall setup degradation rate increases but the inhomogeneity is not aggravated compared to individually cycled cells. All in all, V-dep setup aging results only in a reduced driving mileage while T-dep setup aging results in both reduced driving mileage and acceleration/deceleration capability.