Evaluating internal battery functions and responses is difficult. By its nature, the battery is a closed box, so it is tough to see what’s going on inside, even though the key external parameters — terminal voltage, current flow, and gross temperature — are easy to measure. Researchers have used a variety of sophisticated techniques including MRI scans and Raman spectroscopy to help them see what’s transpiring inside in real time, with some impressive results. Still, it’s a real challenge to observe and quantify the intimately related electrical, chemical, and thermal events.

There’s another difficult aspect of battery testing: How do you consistently introduce desired faults so you can see how the battery “reacts” (especially an issue with Li-ion chemistries and their well-known thermal runaway and fire issues)?

Researchers at NASA’s Johnson Space Center and the Department of Energy’s National Energy Renewable Laboratory have devised a simple-looking patented approach (U.S. Patent # 9,142,829) and are looking to license it (“Internal Short Circuit Testing Device to Improve Battery Designs”). Of course, just because it looks simple doesn’t mean that it is or that it was simple to develop (and if it really is simple, well, that’s even better.) The objective is to install within the battery a latent, quiescent fault that can be triggered on demand.

In brief, they construct a sandwich of a very thin copper or aluminum disk with a diameter of about 20 mm, a copper puck, polyethylene or polypropylene separator, and a 50-micron layer of wax (Figure 1). They implant the device in a battery cell and can create an internal short circuit by exposing the cell to higher temperatures and melting the wax. (Using different wax formulations allows control over the temperature at which the “fault” is triggered.)

This multilayered 20-mm disk made of aluminum and copper with an interposed meltable insulator can be implanted in a battery and then thermally triggered to create an internal short circuit. (Image: NASA)
This multilayered 20-mm disk made of aluminum and copper with an interposed meltable insulator can be implanted in a battery and then thermally triggered to create an internal short circuit. (Image: NASA)

This wax is wicked away by the separator, cathode, and anode, so the remaining metal components can come into contact and develop an internal short circuit (Figure 2). Full construction details, application insight, and operating sequence of events are detailed in the 16-page patent, of course.

Note that the device can be placed anywhere within the battery (and can be used with both spirally wound and flat-plate cells). It can be located between the cathode and anode; between the cathode and the negative electrode; between the anode and the positive electrode; and between the positive and negative electrodes, as well as other locations. Each provides a test setting for different types of internal shorts. There’s even an option for providing a small hole, which can be sized to modulate the short-circuit current: a larger hole for a solid short or a smaller one for restricted current flow.

The result is what the researchers say is a reliable testing method to safely test battery-containment arrangements for thermal runaway, and I don’t doubt it. I am sure they have verified that it’s a valid technique for battery R&D test. Of course, it is obviously not for sampling in factory-production test, as the cost and risks of adding it outweigh the benefits.

Battery Testing When triggered, the unit induces a short circuit, which can be used to study battery behavior, possible thermal runaway, and other undesired consequences. (Image: NASA)

Still, as with all test setups, there’s the question of to what event the test arrangement is invisible to the device under test (DUT). In an ideal world, the test setup would have no impact, but that’s often not possible: Product test is often a macro version of Heisenberg’s Uncertainty Principle for atomic-scale events. Whether it’s observing the internal real-time operation of a battery, measuring load current and this power via a sense resistor, assessing RF power, or almost anything related to photons, the reality is that getting data while not affecting the DUT is often a major challenge.

Have you ever had a power-centric or other situation in which the only practical observation arrangement perhaps adversely affected the DUT or system more than you considered acceptable?