Destructive battery tests

Newly designed system for a reliable test environment

Abuse tests show how efficient and safe batteries remain when exposed to conditions outside of their intended use. These extreme tests usually result in thermal runaway of the battery. To protect personnel and equipment, a suitable test environment is required, which Weiss Technik has recently developed. The report explains the results of the beta tests.

Testing as a basis for battery use

The introduction of batteries and their use in automotive applications is subject to a series of tests. The test conditions are defined in national and international norms and standards. There are also internal company guidelines, particularly in the automotive industry, which may differ from the official test specifications and are used for development purposes in addition to certification. These battery tests are carried out at cell, module, pack or vehicle level.

Some tests are performed outside the specified operating range of the batteries and are used to demonstrate their safety and functional integrity. This often results in thermal runaway of the battery - which requires a suitable test environment.

Specially developed test system for destructive battery testing

The newly designed Weiss Technik test environment features tertiary explosion protection, which reduces the effects of an explosion or blast, such as pressure and heat release, to a harmless level. The test space is designed to withstand regular battery incidents (fires and gas explosions) without damage. It is certified as a pressure-shock-resistant design together with the pressure relief device according to the ATEX directive (hydrogen, gas group 2C).

The test system essentially consists of three components: Pressure relief mechanism, test space and rack. The test system has a modular design so that it can be equipped with appropriate additional options (test equipment) to suit the test requirements.

The test equipment is used to carry out thermal, mechanical and electrical tests. For thermal tests, the test chamber can be combined with an external air conditioning module that allows the desired temperatures to be controlled. A mechanical press can be passed through the underside for mechanical testing (e.g. penetration or integrity testing). Wall access ports, specially designed to withstand the extreme conditions, can be used to wire the test specimen or to feed measurement sensors into the test space, allowing electrical tests to be carried out.

Further options for the test system will be available in the future, such as a high speed camera, replaceable inner liner (contamination issues) or gas analysis. Work is also underway to develop and certify a secondary area with integrated waste gas treatment, which will be connected to the test space and minimise the release of pollutants into the environment.

The first investigations with the new test system on two different battery cells or a cell group are described below.

Set-up and description of the test trials

Figure 2 (see below) shows a schematic of the test set-up for the destruction tests. The test space (ExtremeEvent) is set up in a fire and pressure resistant laboratory room equipped with ventilation.

During the destructive safety test, the temperature is measured in the test space as well as on the wall surfaces and on the test specimen. The pressure in the test chamber is also measured at two points. The measuring sensors are located in the test space through the integrated wall access ports, preventing heat and material exchange with the environment.

Figure 3 lists the relevant properties of the tested battery cell in tabular form. A cell group consisting of a total of six battery cells was used for the test. Specifically, this is an NMC battery cell (chemical compound consisting of nickel-manganese-cobalt as active material for cathode), which has a comparatively violent reaction compared to other cell types.

The cells are held together with a clamping claw, as shown in Fig. 4, and heated evenly on a heating plate (1 kW) until thermal runaway of the cells occurs.

After a period of about 30 minutes, the cells reach almost 290 degrees Celsius on the underside (TDUT, bottom). On the upper side, the temperatures at this point are just over 100 degrees Celsius (TDUT,top), see Fig. 5. Shortly afterwards, there is an enormous outgassing (venting) and finally thermal runaway.

Within about 25 seconds, five cells gas out completely, with the gas igniting immediately due to the high cell temperature (thermal runaway). One cell (cell 02) enters the thermal runaway state significantly earlier (approx. 6 minutes). The small time offset, with the exception of cell 02, can be seen in the voltage curves and voltage drops U in Fig. 6. Two pressure peaks occur during the thermal runaway of the cells. The maximum pressure at both measuring points is approximately 15 mbar (Pbottom) and 17 mbar (ptop), which activates the pressure relief mechanism, causing the valves to partially open. During thermal runaway, the temperature rises to a maximum of over 660 degrees Celsius at the top of the cells. The air temperature (Tair) briefly reaches 435 degrees Celsius, and the inner surfaces of the side wall or test space door reach a maximum of approximately 190 degrees Celsius.

At the same time, measurements were carried out with a single cell (comparable cell type), which is currently under development and has been optimised in terms of performance and energy density. The capacity of approx. 190 ampere hours is significantly lower than that of the investigated cell group.

In contrast to the cell group, the gas and heat quantities released during thermal runaway are simultaneous - and not time-shifted as in the previous test (cell group). Again, the cell is brought to thermal runaway by means of a 1.5 kilowatt heating plate, one on each side of the cell. Thermal runaway shows a much more violent reaction, resulting in a spontaneous explosion and disintegration of the cell. Accordingly, the measurement results show that the reaction to the test space is significantly more severe. This can be clearly seen in the measured pressure peak.