Using Sound to STudy batteries
Feasible uses sound to detect the physical properties of batteries. We do this by sending ultrasonic pulses through a battery and recording the echoes. By applying advanced analytics to the data, we can then extract meaning from it.
A major advantage to our approach is that the acoustic behavior is sensitive to changes in the physical properties along the path of the sound waves. Thus, electrical, thermal, and strain-based methods, ultrasonic analysis can directly and actively probe the physical condition of internal battery components.
The speed of sound, c, in a material is a strong function of its density, ⍴, and stiffness or bulk modulus, B. The acoustic impedance, Z, of a material (or its willingness to transmit sound) is also strong function of ⍴ and B. Importantly, as a sound wave reaches the interface between two materials, the mismatch in Z between them determines how much of the sound wave gets reflected back (R) or transmitted through (T).
Real-time Analysis during use
In a basic measurement, the echoing behavior is recorded as time elapses after the initial input pulse (time-of-flight). As each reading takes much less than 1 ms and is non-desctructive, acoustic analysis of batteries can be carried out in real-time during use. The video below shows real data from a Li-ion battery as it is charged and discharged. The shifts in echoing behavior is clear, and demonstrates the sensitivity of sound to these subtle physical changes.
We ran similar tests on several types of commercial batteries with widespread adoption, including LCO and LFP pouch cells; alkaline, NiCd and NiMH AAs; and NMC and NCA 18650s. Every cell type showed repeatable and robust correlation between the acoustic behavior and charge state. This shows that our approach can be used to analyze batteries in a way that is mechanistically different from and complementary to standard diagnostic methods.
Many of the highly-publicized issues with battery performance and safety can be traced back to problems with quality control, defect detection, or in-use monitoring. This results from a lack of at-scale diagnostics that give direct information about internal battery structure.
Our technology meets a deep need in battery diagnostics, offering structural information with a measurement that is fast, non-invasive, and scalable. It can provide physical insights into the battery development, manufacturing, and in-use management life cycle that were previously impossible. Our method enables users to directly probe physical processes affecting performance. This could be used to aid cell design and development, improve manufacturing quality and yield, and increase efficiency and safety of battery usage.
If wildly successful, we will change how batteries are made, tested, and managed everywhere.