FMEA-Graph-Based Approach for the Identification of Critical Cascading Failure Events in Li-Ion Battery Packs

Recent fire incidents involving electric vehicles have raised concerns about the reliability of lithium-ion battery systems. These events highlight the need for detailed failure hazards analysis to better understand how failure events correlate and propagate within the battery pack. As complex, hierarchical systems comprising numerous interacting multi-physics components, battery packs present unique challenges for effective failure analysis. This work introduces a systematic methodology to identify the most critical sequences of failure events within battery packs. The methodology utilizes the functional relationships between components and applies a “failure-cause-to-failure-mode” assumption to map failures and their associated risk priority numbers (RPNs), identified through Failure Modes and Effects Analysis (FMEA), onto a directed graph. The analysis then extracts the failure sequences with the highest risk factors, offering insights into the most critical paths of failure propagation. This approach provides a powerful tool to enhance system reliability during the design phase and supports fault diagnosis strategies by facilitating root cause identification.