Identification of mechanism consistency for LFP/C batteries during accelerated aging tests based on statistical distributions

This study establishes a novel approach to investigate if accelerated aging tests can accurately model realistic cell aging in a short time while also maintaining the consistency of the involved aging mechanisms. As a trade-off between efficiency and consistent mechanism, the application of accelerated aging necessitates carefully selecting stress factors to identify the operational range and the significance of aging-related stress factors. Based on three levels of major stress factors designed for 43-month calendar aging tests and 10-month cyclic aging tests, this work aims at the stress ranking and indicating suitable operational intervals for commercial LFP/C batteries, taking two of the most popular lifetime distributions for batteries, namely log-normal and Weibull. Statistical distributions of lithium-ion batteries are attained from discharge capacity loss with nonlinear mixed-effects (NLME) models. Results prove that log-normal is the preferred model, and the right-skewed Weibull becomes more pronounced with deeper aging, especially in calendar aging. The evolution law of distribution parameters guided by the consistent acceleration factor was derived. The likelihood ratio parametric bootstrap approach based on the NLME model for life samples consistently yields that test conditions with the temperature above 47.5 °C and average state-of-charge (SOC) for cycling aging above 72.5% can result in different life behaviors. In contrast, the combination of SOC levels and higher temperatures does not lead to a change in the calendar aging mechanisms. The temperature is the most significant stress, followed by temperature-coupled cycle depth and SOC levels. This method can offer a reference to make reasonable test plans for detecting battery's performance to predict their life more accurately.