Ultrasound Non-destructive Evaluation for Battery Management Systems.

Recent research has shown that ultrasound can detect the changes in material elasticity that occur as batteries undergo charging and discharging, enabling accurate measurement of the battery's state of charge. Our study aims to build on this approach by implementing it in automotive batteries, allowing for continuous monitoring of the battery's charge level and structural health during normal use.

Lithium-ion battery cells are widely recognized as a crucial component of sustainable transportation solutions. Incorporating ultrasound for charge monitoring can enhance the performance of these cells. Traditional methods for measuring the battery state of charge (SOC) rely on tracking voltage and current, but such methods suffer from limited efficiency and accuracy.

In contrast, ultrasound enables direct SOC measurement at any time, independent of charge history, thus eliminating errors that may accumulate in successive measurements. This technology can provide highly accurate SOC readings, improve battery range estimation, and enhance the structural integrity of the battery.

Previously, successful demonstration of ultrasound charge monitoring on an individual battery cell was carried out in a laboratory environment. The changes in elastic properties and density of lithium-ion batteries affect the wave speed travelling through the test cell, and this wave speed is measured by determining the time taken by the longitudinal wave to traverse the cell.

This information is then used to determine the battery SOC. However, automotive batteries consist of several cells stacked together, and hence, our research aims to investigate the application of various ultrasonic techniques to multiple cells within a battery module and its impact on module design.

To achieve this, we will incorporate a built-in test system in a laboratory environment to collect data that will subsequently be analysed using signal processing and numerical modelling techniques in predictive machine learning. Initially, the equipment will include an ultrasonic pulse-generator, ultrasonic probes, a custom test bed, and an individual lithium-ion cell before moving on to multiple cells stacked in series.

Therefore, we aim to address the challenges posed by offline, single-cell, history-dependent, complex, and costly SOC prediction techniques. Additionally, we aim to embed an ultrasonic network configuration relevant to the automotive industry.

Our research aims to gain a comprehensive understanding of the physiochemical characteristics of lithium-ion batteries, with the goal of advancing the sustainable scalability of the automotive sector, particularly as electric vehicle deployment continues to increase. Specifically, we seek to provide new insights into the physiochemical changes of lithium-ion batteries, enabling more accurate determination of state of charge (SOC) and state of health (SOH) information.

Our approach aims to extract this information in a cost-effective manner, without the need for expensive equipment, thus facilitating battery management systems that can provide accurate outputs under driving conditions. By offering a viable alternative to current SOC/SOH determination methods, our work can contribute to minimising manufacturing costs, reducing material requirements, and achieving a simple yet effective solution that moves the automotive industry towards sustainability.