Increasing electrification of vehicles has led to a growing demand for Li-ion batteries in automotive applications. Automotive vehicles experience cyclic mechanical loads during operation, which may include; acceleration, deceleration, shock, vibration and collision. Most of these events are not severe and do not have an instantaneous effect on the battery, however it is unclear if there is a cumulative effect over time. This cumulative effect may have a significant impact on performance and safety of the battery. The collective effect of these types of mechanical events is not well characterised and there exists a gap in knowledge. This is due to the difficulty in replicating real world conditions in the laboratory.

This project focusses on the study of Li-ion batteries during static testing (3-point bend) and dynamic testing (acceleration/deceleration pulses). Several 3 point bend tests were carried out where the cell was bent to up to 30 mm and the relaxation behaviour was captured by monitoring the reduction in load over time. The electrochemical impedance of the cells was also monitored during the relaxation period and then cycled between 2.7 and 4.2 V. After each cycle, EIS was performed in order to detect impedance variation. The gradual separation of the internal structure of the cells was also captured using X-ray tomography as demonstared in fig.1. This image reveals clear separation of the internal layers at the point of contact with the side rollers and this significant separation will directly impact the performance of the cell in long term. High precision coulometry was also used as another technique to predict the cycle life of the cells besides identifying any degradation mechanism due to the severe bending. For dynamic testing cells were exposed to up to 1000 g pulses, using a drop tower (10 kJ maximum energy input). Impedance, differential capacity, and capacity of the cells were analysed before the drop, after the drop and then periodically during storage at 35^oC.

Results from the high g impact tests showed that there was damage caused to the cells, but this could not be observed using EIS until after electrochemical cycling, which then manifested as an increase in the impedance of the cell. This suggests that the mechanisms are more complicated than simple damage caused by the impact, but that the impact causes damage which cannot be detected. These damages are then exacerbated by charging/discharging which then can be identified. We expect this to have significant implications for safety particularly that it may not be possible to detect damage caused by impacts without working the cells, hence this has an inherent risk. The long term effects of these incidents with regards to failure threshold are yet to be investigated.

Figure.1: CT image demonstrating clear separation of the internal layers of a pouch cell under 3 point bend loading.