Ensure airbags deploy properly

Due to the explosive nature of an airbag system, testing its components requires careful attention to detail. In the case of the airbag inflator, you must take care not to trigger the device while testing it. On the other hand, testing a complete airbag system requires that you do trigger the system so you can observe the airbag inflate. Because testing destroys the airbag in a controlled explosion, you must ensure the setup is safe and that you gather all the data you need when the airbag deploys. You don't get a second chance to inspect a particular airbag, and running extra tests—which destroy additional airbags— becomes expensive.

Although you can test every component in an airbag inflator using automated test stations (see "Don't blow it," p. 7), you obviously can't test the deployment of every airbag in a production run. After you deploy an airbag in a test, you can't remanufacture it and ship it.

To help ensure product quality, airbag-system manufacturers test samples taken from production lots. Automakers also may have suppliers perform lot-acceptance testing (LAT) as part of their purchase-contract obligations. Although the number of samples per lot varies, typical tests involve one airbag, randomly selected from a production lot.

Airbag test systems include equipment that triggers the initiator and measures the response time as

Fig 1 Tests of an airbag system place the unit in a realistic position in a special test chamber. Courtesy of Microsys.

• a deployment chamber, equipped with the appropriate interlocks for operator safety,
• a ventilation system that removes dust and toxic residue after a test,
• a PC that controls the test and analyzes test data,
• one or two high-speed cameras that record the airbag deployment, and
• data-acquisition hardware that records test parameters. As an option, the test system also may include sensors that measure chamber temperature and internal air pressure.

When running a test, a technician installs a sample airbag system in a fixture (Fig. 1) housed in a deployment chamber. The technician then closes the deployment chamber, selects the appropriate test sequence, and runs the test. The test system then programs and triggers the cameras and "fires" the airbag.

Each airbag contains an inflator that provides the gas that inflates an airbag on command from electronic sensors. The inflator incorporates one or two initiators that start a rapid chemical reaction—a controlled explosion—that actually generates the gas. Initiators can generate the gas at different rates, depending on the decelleration detected during a crash by other electronic systems.

Fig 2 Three images show the deployment of an airbag as it inflates. Courtesy of Microsys.

The signal that fires an initiator is typically a 3-ms to 5-ms square wave, with an amplitude of 1.75 A. To ensure an initiator fires properly, the test system measures the voltage across the initiator's ignitor wire. As the wire heats, its resistance varies, and when the initiator fires, the explosion opens the wire and the current drops to zero. In an airbag system that uses two initiators, the test system will sequence the firings according to the manufacturer's specifications.

Once the airbag has deployed, the test system downloads the video data from the cameras. For production tests, this data includes approximately 100 video frames taken at 1000 frames/s during the first 1/10 of a second as the airbag inflates.

An experienced technician, or possibly an engineer, will analyze the video frames to ensure the airbag deployed properly. Fig. 2 shows three images from a typical inflation sequence, as acquired by a side-view camera.

The technician or engineer will step through the frames, or view them in slow motion, noting any impediment to the proper deployment. If the airbag included dual initiators, the technician will look for signs that firing the second initiator caused the airbag to inflate at a different rate. In most automated test systems, the ability to synchronize the video frames with the acquired data makes it easy to correlate observations with electrical or other test data.

If an airbag fails to inflate properly, the technician can send the test data to a design engineer or quality engineer for further analysis. When diagnosing the failure, the engineer will look for things that may have mechanically impeded the deployment, such as housing materials in the airbag system that didn't fully swing out of the way. He or she will also look at the initiator's current waveforms and the images to determine if that component worked properly. Only after looking at this information can the engineer decide if a production lot requires further testing.

Although this inspection method has proven effective, you can expect to see better inspection methods in the near future. Instead of relying on a technician or engineer to check images, test software will automatically analyze images to determine the characteristics of airbag deployment. This software will apply image-analysis techniques similar to those currently used to analyze how crash-test dummies move in a collision. In effect, this software takes the technician out of the inspection process and improves the repeatability and accuracy of the inspection test results.

"Testing Dual Airbag Inflators and Modules with the Model 2790 SourceMeter Switch System," Application note 2378, Keithley Instruments, Cleveland, OH, 2002. www.keithley.com/kei_assets/downloads/9665.PDF.

"Why Test Airbags?" Case study, Microsys Technologies, Mississauga, ON, Canada, 2002. www.micro-sys.com.