Sahraei failure criteria for detection of short circuit in lithium-ion batteries

The increasing reliance on Li-ion batteries for Electric Vehicle (EV) propulsion brings forth significant safety challenges, particularly in ensuring battery resilience under mechanical abuse and accidental impacts. A critical concern is the potential for internal fracture within battery cells, which can trigger short circuits, leading to thermal runaway and explosions. Consequently, understanding battery behavior under various loading conditions and developing predictive models for failure is essential for enhancing safety and optimizing protective structures in EVs. This dissertation presents a comprehensive analysis of Li-ion battery behavior under multiple loading scenarios, supported by two novel modeling approaches: a universal homogenized model for cylindrical cells and the Sahraei Failure Criterion for short circuit prediction. The first part of this research focuses on the development of a universal homogenized model for 18650 cylindrical battery cells, capable of accurately predicting cell behavior under axial, lateral, and three-point bending loads. Unlike previous models that addressed one or two loading conditions, this model incorporates uncoupled axial and lateral property calibrations and employs anisotropic crushable foam modeling for enhanced accuracy. The model is validated through experimental data and demonstrates superior performance in predicting cell response, particularly in axial and bending scenarios. The second part introduces the Sahraei Failure Criterion, a universal failure model designed to predict internal fractures and short circuits in both cylindrical and pouch cells. This criterion, derived from microstructural simulations of the electrode-separator assembly, is implemented in commercial simulation software, including Altair RADIOSS and Ansys LS-DYNA. The failure model is validated under various loading conditions, such as hemispherical and rod indentations, in-plane loading, and three-point bending. By defining the jellyroll’s failure strain based on the interaction of compressive and tensile strains, the model accurately predicts the onset of internal fractures, providing a critical tool for battery safety analysis. Furthermore, real physics has been added to the base failure criteria. By testing battery cell’s mechanical response under different physical conditions, enhanced failure criteria have been built and validated. Together, these models offer a robust framework for understanding and predicting Li-ion battery behavior under mechanical stress, contributing to safer EV designs and more effective protective structures. This research advances the field by combining detailed material calibrations and computational modeling, offering a comprehensive solution to address the safety concerns associated with Li-ion batteries in electric vehicles.