Early warning of battery failure via spontaneous acoustic emission: a case study of overcharging in sodium-ion batteries

Exploring novel real-time online monitoring methods that can reflect the internal degradation state of batteries is the fundamental approach to improving the timeliness and accuracy of safety early warning for battery systems. This study investigates the feasibility of applying acoustic emission (AE) technology to characterize internal damage during the overcharging of commercial sodium-ion batteries. The experimental results indicate that the self-generated AE signals of batteries during overcharging are highly consistent with the battery degradation mechanisms. Both the temporal distribution of the AE signals and the degradation mechanisms can be categorized into four corresponding stages. In-situ optical AE experiments further validate the synchronization between two characteristic AE signals and internal damage as well as side reactions within the battery: high-frequency signals in the range of 150–300 kHz originate from the initiation and propagation of cathode particle cracks, while low-frequency signals between 5 and 100 kHz are associated with gas generation from electrode and electrolyte decomposition as well as bubble rupture processes. Accordingly, this study constructs a targeted signal processing workflow to extract the two types of characteristic events mentioned above, and introduces the Time-Weighted Event Density (TWED) index to quantify the aggregation degree of AE events. Cyclic experiments at different cut-off voltages show that the TWED value of the battery is highly correlated with its available capacity fading, which can be used as a key indicator for diagnosing accelerated mechanical damage of electrode materials and abnormal gas generation. This study demonstrates that monitoring the AE characteristics of batteries can effectively identify internal physical damage and side reactions, providing a promising non-destructive monitoring method for early warning of battery thermal runaway.