A recent uptick in battery-related fires has drawn attention to the challenge of identifying defects that can cause these catastrophic malfunctions but are rarely obvious to the naked eye. In hopes of preventing the dangerous glitches that can cause batteries to overheat and catch fire, researchers from Drexel University have developed a standard testing process to give manufacturers a better look at the internal workings of batteries.
In a paper recently published in the journal Electrochimica Acta, the group presented methods for using ultrasound to monitor the electrochemical and mechanical functions of a battery — which would immediately reveal any damage or flaws that could lead to overheating and even cause “thermal runaway.”
“While Li-ion batteries have been studied for nearly half a century and commercialized for over 30 years, we have only recently developed tools that can see inside with high resolution,” said Wes Chang, Ph.D., Assistant Professor, Primary Investigator, Battery Dynamics Lab in Drexel’s College of Engineering, who supervised the project. “In particular, ultrasound has been adapted from other fields, such as geophysics and biomedical sciences, for battery diagnostics only in the past decade. Because it is such a new technique in the battery and electric vehicle industries, there is a need to teach battery engineers how it works and why it is useful.”
The team’s recent work strives to do this by demonstrating a low-cost, accessible benchtop ultrasonic tool that it hopes can be easily implemented and used by battery engineers, including those who work at automotive companies producing EVs.
Here is an exclusive Tech Briefs interview, edited for length and clarity, with Chang.
Tech Briefs: What was the biggest technical challenge you faced?
Chang: We've been developing a new technique for battery diagnostics, and it's based on sound waves. So, it's adapting ultrasound normally used in the biomedical field for battery aging studies. To your question, this is always a problem with developing a new method: Battery scientists are normally chemists or material scientists, not so much mechanical engineers. Ultrasound is something that is commonly used in biomedical sciences and is also used in geophysics, it's really a mechanical probe. So, the biggest challenge for me was building a team that has people with expertise in mechanical design and signal processing and then trying to leverage those skills to build something that can measure battery chemistry. It's a very cross-disciplinary endeavor.
Ultimately, you have a tool that measures elastic modulus and density, which is correlated with battery aging and battery performance. The challenge was then making this technique user friendly and easy to understand for battery scientists.
Tech Briefs: Do you have any set plans for further research, work, etc.? If not, what are your next steps? Where do you go from here?
Chang: We do have set plans for upcoming work. The method as it stands is either a 1D or 2D technique. The 1D refers to when you scan the center of a batter; you send a soundwave through the center of the battery, and, as you cycle the battery, that soundwave changes shape. The way it changes shape is correlated to a change in elastic modulus and density.
You can do the same thing in 2D by just scanning the electrode, and that gets you an image of the battery. So, it tells you not just when but also where something is happening. So, 3D then refers to the ability to not only get a scan of the battery but also layer by layer resolution. This refers to the geometry.
So, a battery, from the outside, looks pretty simple. It's like a pouch or something like a cylinder. But inside it's many layers of electrodes that are stacked together. If you see a change in 2D, you normally attribute it to some area. But the question is, is that a defect or is that a possible failure mode. Is it happening through every electrode or is it happening on one specific electrode? That's what 3D tries to get at. You can imagine it is quite challenging because it requires us to figure out how to decompose that waveform into its individual components, like how it interacts with every single layer.
So, as a next step, what we're doing is building better algorithms that allow us to decompose the waveform into the effect of each layer. That's what will get us to 3D resolution. We think it's possible because this is already a capability that x-rays have. And, in some cases in biomedical sciences, they've been able to get the 3D version. It is definitely at the forefront of ultrasound technology, but that's where we want to be.
Tech Briefs: Is there anything else you'd like to add that I didn't touch upon?
Chang: I want to emphasize that ultrasound for batteries has been around even when I was a Ph.D. student — we were developing it in my old lab. So, really, the emphasis here is making a relatively new tool easy to use for battery scientists. And I just want to emphasize that the main outcome was basically us building this platform at the battery, battery startup SES AI , at their R&D site; we were able to train one of their engineers to use it on a daily basis. To me, that is the most fruitful and motivating aspect of our work — to see it being directly translated to industry usage. So, it's not only isolated to SES, but we're also speaking with other startups and a few large automotive manufacturing companies that either already have an ultrasound tool or are looking to get one. We're speaking with them to help them understand the tool better and make it more like a plug-and-play capability that any R&D lab and industry has for batteries.
