Higher tightness requirements in automotive production

One reason for the growing importance of leak testing in automotive engineering is the requirements resulting from the goal of reducing emissions from fuel tanks and lines. The heat exchangers for exhaust gas recirculation systems for nitrogen oxide reduction must also be tested for leaks during production. Similarly, new refrigerants that are not harmful to the climate but are highly flammable, such as R1234yf, entail higher leak tightness requirements. The leak-tightness of components also plays an important role in technologies for increasing the efficiency of combustion engines: whether it is the intercooler for turbo engines or fuel injection systems. Modern diesel common-rail systems, for example, must be leak-proof at pressures of up to 3000 bar. Of course, quality assurance also extends to components that are directly relevant to safety, such as brake boosters or the gas generators of airbags. The reason for the millions of airbag recalls that have been going on for several years now: No air humidity may penetrate into pyrotechnical gas generators during their life cycle.

Example 1: Heat exchanger
The problem with testing coolers and heat exchangers is on the one hand their function-related temperature sensitivity and on the other hand their complex geometry. The results of a water bath test stand and fall anyway with the attention of the human tester. If, in addition, a complex rib structure of the test part ensures that water bubbles can escape from a leak but cannot rise because they are stuck between the ribs, the water bath test is of course in vain. Pressure drop testing also reaches its limits in the context of heat exchangers. The smallest leak rates that can ideally be determined by a pressure drop test are in the order of 10-3 mbar∙l/s. But even the smallest changes in temperature can lead to leakage. But even the smallest temperature changes alter the pressure differences that can be measured. This is because if the temperature rises during the test, leaks remain undetected; if the temperature falls, the pressure drop test detects phantom leaks. An example: If the temperature in a test part with 3 liters of volume and 2.5 bar air pressure rises by only 0.1 degrees during a measurement interval of 20 seconds, this increases the internal pressure to 2.50085 bar. This means that any leak rate will appear to be 0.13 mbar∙l/s (=^ 8 sccm) less than it actually is. The temperature difference of 0.1 °C thus causes a measurement error in the pressure drop test that is a hundred times greater than the limiting leak rate of the method.

In order to be able to reliably test heat exchangers and coolers today, test gas methods are used. The test gases used are usually helium or the commercially available forming gas, a non-flammable mixture of 5 % hydrogen and 95 % nitrogen. Automotive suppliers worldwide already perform more than 800 million leak tests on heat exchanger systems every year - from engine oil and exhaust gas recirculation coolers to radiators. Especially for small and medium-sized parts that only need to be tested against possible oil or water leaks, testing in the simple and cost-effective accumulation chamber is recommended. This involves measuring how much test gas escapes from the oil, water or charge air cooler and accumulates in the accumulation chamber over a specific time interval. In practice, helicopter testing in the accumulation chamber determines leakage rates of up to 1∙10-4 mbar∙l/s. In this way, water-tightness (leak rates less than 10-2 mbar∙l/s) and oil-tightness (leak rates less than 10-3 mbar∙l/s) can be guaranteed.

Example 2: The air conditioner
Changes in the field of air-conditioning systems used in cars are also tightening up the sealing requirements. This is because highly climate-damaging fluorinated greenhouse gases such as R134a are about to be phased out for good. As of January 2017, no new cars produced in the EU will be allowed to use R134a as a refrigerant in their air conditioning systems. However, alternative solutions such as the refrigerant R1234yf are flammable even at low temperatures and can react to form highly corrosive hydrofluoric acid when exposed to heat. Because R1234yf is more expensive than R134a, manufacturers also like to calculate with a lower reserve of refrigerant for their system, which also increases the leak tightness requirements. R1234yf is currently favoured by car manufacturers in Asia and the USA, while carbon dioxide (CO2) is a popular alternative among German car manufacturers. However, CO2 places special technical demands on an air-conditioning system because it is used at a significantly higher operating pressure of up to 120 bar. Either way, the tightness requirements for air conditioning systems and their components are increasing. The old rule of thumb of 5 g R134a, which may leak per year and connection point, is obsolete due to the new refrigerants.

A refrigerant loss of 5 g/a corresponds to a helium leakage rate of 4∙10-5 mbar∙l/s. Currently, most air-conditioning components - whether evaporators, condensers or charging valves - are still tested against leakage rates in the order of 10-4 to 10-5 mbar∙l/s. For air-conditioning hoses, for example, a heli-um test is used in the vacuum chamber. The helium test in the high-density vacuum chamber measures how much helium escapes from the test part into the vacuum of the chamber through a leak. Compared to the test in the simple accumulation chamber, the test in the complex vacuum chamber has the advantage of shorter cycle times. The sensitivity is also higher - under optimum conditions, leak rates of down to 1∙10-12 mbar∙l/s can be determined using the vacuum method.

Although car manufacturers expect the supplier to carry out quality assurance and to check for leaks, a leak test is still necessary after installation of the air-conditioning system at the three to six connection points of the air-conditioning system that the car manufacturer has installed himself. The car manufacturers are therefore concerned to keep the number of such joints as low as possible, especially as these points are difficult to access, particularly in the case of more expensive vehicles with more elaborate panelling towards the interior. The leak test of the joints is then usually carried out during final assembly with a sniffer leak detector. In the past, helium or forming gas was used as the test gas, but today sniffer leak detectors often detect the respective refrigerants directly by measuring traces of leaked R134a, R1234yf or CO2.

Example 3: Fuel components and injection systems
In general, fuel systems, fuel tanks and fuel lines are subject to increasingly stringent sealing requirements. One driver for this are strict US and especially Californian regulations for the prevention of hydrocarbon emissions. This also makes the use of permeable plastics problematic. Fuel tanks and fuel lines are therefore tested today by many manufacturers against leakage rates of 10-4 to 10-6 mbar∙l/s - which makes an integral leak test with test ga-ses necessary. For smaller parts such as fuel injectors or motorcycle tanks, helium testing in the accumulation chamber is a good choice. However, because the detection limit of the accumulation method depends on the free volume of the test chamber, very large parts are tested using the vacuum method.

Injectors and gasoline pumps, for example, are tested with helium in the accumulator chamber against leakage rates in the range of 10-4 to 10-5 mbar∙l/s. But modern common-rail injection systems often have even higher leakage requirements due to the particularly high pressures at which they are operated - the limit leakage rates here range up to 10-6 mbar∙l/s.

Example 4: Airbags
It is well known: Quality problems with gas generators of airbags have recently led to several - and never-ending - vehicle recalls. Today, if suppliers want to exclude the possibility of moisture penetrating pyrotechnical gas generators, they usually test against a leakage rate of 10-6 mbar∙l/s. Often, a special test gas method is used for this purpose. Often a special test gas method is used for this purpose: the so-called bombing. In the bombing method, the detonator is first subjected to a helium overpressure in a pressure chamber, so that the test gas penetrates through any leaks into the interior of the test part. The detonator is then placed in a vacuum chamber. After evacuation of the vacuum chamber, the helium that has penetrated the test part can escape into the chamber and be measured there by mass spectrometers. The leak tightness requirements for cold gas generators for airbags are somewhat higher than for pyrotechnical generators. They usually contain a helium-argon mixture. In order for this gas mixture to inflate the airbag when it is released, it is under a high pressure that must be maintained for at least ten years - some manufacturers even calculate more than 15 years. The leak tightness of cold gas generators is therefore also tested in the vacuum chamber, often against a leakage rate of 10-7 mbar∙l/s. Here, of course, the helium content of the escaping he-lium-argon mixture can be detected directly, without the detour via bombing.

Test gas methods remain indispensable
Whether it's emission reduction, increasing the efficiency of combustion engines, or vehicle safety, there are many reasons for the growing importance of leak testing in automotive engineering. For suppliers and vehicle manufacturers, there is no way around modern, test gas-based methods. The drives of the future will not change this decisively. Leak tests are also necessary at all stages of production of batteries for EV/HEV vehicles. Each individual battery cell - which is later assembled into battery modules and then into battery packs - must be reliably protected against the ingress of air and moisture. For this reason, the individual battery cells are already tested against limit leakage rates of 10-5 to 10-6 mbar∙l/s in the vacuum chamber. In this way, leak testing with test gases for automotive production probably remains
will continue to be indispensable in the future.