Leak test for the production of automotive components

In an increasing number of application scenarios, only modern test gas methods provide the accuracy and reliability required in the manufacturing process - be it heat exchangers for exhaust gas recirculation systems for nitrogen oxide reduction, or charge air coolers for modern turbo engines. Precise leak testing is also essential, especially at the various stages of the production process for traction batteries for EV/HEV vehicles. This is the only way to ensure the operational safety and service life of the expensive battery - from the leak tightness of the individual battery cell to the cooling circuit and the entire housing.

1. water bath or bubble test
30 years ago, the simple but unreliable water bath test was still the common method used by most suppliers. In practice, its detection limit is a leakage rate of 5∙10-2 mbar∙l/s. However, the significance of this test stands and falls with the daily form of the human tester who has to detect the bubbles. In addition, especially in the case of complex geometry, such as heat exchangers, air sometimes escapes from the test part but does not detach and rise as an air bubble. Then it is not perceptible. In addition, components - if they can tolerate moisture at all - often have to be dried at great expense after the water bath.

2. pressure drop test
Some suppliers still rely on the supposedly inexpensive pressure drop test. The test part is filled with air up to a defined overpressure and any pressure changes are measured over a defined time interval. If the pressure drops, a leak exists. The smallest leak rates that can ideally be determined in this way are of the order of 10-3 mbar∙l/s. A major problem with the method, however, is its susceptibility to temperature fluctuations. If the temperature rises by only a fraction of a degree during the test, leaks often go undetected; if, on the other hand, the temperature falls, the pressure drop test detects phantom leaks. Attempts to compensate for temperature changes reduce this effect, but do not eliminate it completely.

3. helium test in the vacuum chamber
Test gas methods are far more reliable here. A vacuum leak tester such as the LDS3000 from Inficon, for example, is highly sensitive, allows very short cycle times and is therefore recommended for installation in automated test systems in production lines. In the vacuum method, the test specimen is placed in a hermetically sealed chamber in which pumps create a vacuum. The test item is also evacuated and filled with the test gas helium, allowing it to escape through any leaks into the vacuum of the chamber, where it is detected by the leak tester. In this way, even larger components can be tested against the smallest limiting leak rates - under optimum conditions down to 1∙10-12 mbar∙l/s. Typical applications of vacuum testing are tests on air-conditioning components - whether evaporators, condensers or filling valves (down to 10-5 mbar∙l/s) -, on common rail injection systems (10-6 mbar∙l/s) or on fuel systems and fuel tanks (10-5 mbar∙l/s). Vacuum testing is also the method of choice for battery cells. Whether they are prismatic, round or pouch cells: Once the manufacturer has filled them with the flammable electrolyte and sealed them, he must test the tightness of the battery cell cases against leakage rates of 10-5 to 10-6 mbar-l/s. It is therefore not uncommon for 3 to 5 % helium to be added as a test gas to the electrolyte filling of the cells.

4. test in the simple accumulation chamber, also with forming gas
However, especially for small and medium-sized parts that are only tested against possible oil (10-3 mbar∙l/s) or water leaks (10-2 mbar∙l/s), testing in the accumulation chamber, which is as simple as it is cost-effective, is recommended. This measures how much test gas escapes from an oil, water or charge air cooler, for example, and accumulates in the accumulation chamber over a certain time interval. In practice, an instrument such as Inficon's LDS3000 AQ detects leak rates of up to 1∙10-5 mbar∙l/s in the accumulation chamber. Instead of helium, the LDS3000 AQ can use less expensive forming gas, a nonflammable, commercially available mixture of 95 % nitrogen and 5 % hydrogen. Using the accumulation method, this device detects liquid leaks as reliably as only helium vacuum testing could in the past - and at a cost similar to that of simple air testing.

5. the dynamic robot sniffing leakage search
Sniffer leak detection with test gas is not only intended to identify leaks, but also to locate them precisely. It is often also used to determine whether the joints between components that have already been assembled are leak-tight, for example before vehicle installation on pre-assembled modules of the air conditioning system or if the leak-tightness of a traction battery housing to the outside - in accordance with IP67 or IP69K - must be guaranteed. A distinction is made between static and dynamic sniffer leak detection. In the latter, the sniffer probe moves over a larger area of the test part. Dynamic robotic sniffer leak detection is particularly demanding from a technical point of view. In this case, the probe is attached to the tip of a robot arm and is automatically guided over the test part. Inficon devices such as the XL3000flex (with the test gases helium or forming gas) or the Protec P3000XL (with helium) are predestined for this application scenario. This is because they draw in the air to be tested with a very high gas flow of 3000 sccm, whereas conventional multi-purpose leak detectors operate with only 60 sccm. The high gas flow allows the probe to maintain the required safety distance from the surface of the test part (often 5 or 6 mm) during dynamic and automatic sniffing leak detection, while still being able to identify leaks 100 percent even at higher feed rates of 10 cm/s or more. Typical limiting leak rates to be tested against in dynamic robotic sniffing leak detection are 10-2 mbar∙l/s for water tightness, 10-3 mbar∙l/s for oil tightness and 10-4 mbar∙l/s for tightness against liquid fuels.

6. multigas sniffer leak detection after installation
Even after the actual installation of the air conditioning components, the OEM still has to test the tightness of the joints - or even check the refrigerant circuit of the installed traction battery for leaks. In such cases, a multi-gas leak detector such as the Ecotec E3000 from Inficon can use the refrigerant in question - whether CO2 or R1234yf - itself as a test gas and directly detect any leakage. In fact, such a multi-gas leak detector is the only way to test the filling valve for leaks after it has been closed. It can also detect fuel leaks in the common rail connections. The smallest leak rate that can be detected with the Ecotec E3000 corresponds to 0.05 g/a of the refrigerant R134a or a helium leak rate of 10-6 mbar∙l/s.