---

When testing air bags, test engineers must assess whether or not an air bag passes a test based on what they observe during bag inflation. But the bags inflate so rapidly a human inspector cannot record all details. To capture useful images of an air bag as it breaks out of its container and expands fully, test engineers need fast cameras, good lighting, the proper lens, and a computer system that will synchronize the entire test process. (see “New Air Bags Challenge Test Engineers,”).

High-speed imaging has evolved along with air-bag system design. It wasn’t long ago that engineers used high-speed film cameras to capture images of tests. But the demand for faster testing and test repeatability led to a need for higher image rates, so engineers moved from film to digital cameras and systems to acquire images.

Digital Offers Advantages

Figure 1. A display of captured data shows the image of an inflating air bag along with information from sensors and information about the test system. The system should simultaneously display an image and the acquired sensor data.

Unlike film images, digital images are easy to save and retrieve, and computer monitors display them easily. And because digital images exist as computer files, test engineers can quickly analyze them and compare them to performance standards.

Until recently, imagers operating at a maximum of 1000 images/s—one image per millisecond—were adequate for imaging air bags mounted in steering wheels and in instrument panels. Those types of bags inflate in 30 to 50 ms, so 30 to 50 digital images would capture their inflation. That was enough information to determine the success or failure of a bag.

Today, side air bags can inflate in as little as 5 ms. But capturing that inflation at 1000 images/s yields only five images, which provide insufficient information for an accurate analysis. So, changes in air-bag design meant that imaging systems had to operate at faster image-acquisition rates.

Typical of the high-speed imagers Microsys uses when developing test systems is the Kodak EktaPro HS Motion Analyzer, which captures 4500 full images/s, or 0.22 ms per image. At this rate, during a 5-ms interval, a camera acquires 22 images. The high-speed image acquisition lets test engineers see precisely when the bag’s inflating charge ignited and when optimum inflation occurred.

Shutter Speeds Increase, Too

As image-acquisition rates have increased, electronic shutter speeds have decreased. A shutter speed of 0.5 to 1 ms yields acceptable results for slow actions but is too slow for today’s air bags. To ensure sharp images, we use a system that operates with a shutter speed of 200 to 300µs.

In addition to image-acquisition rate and shutter speed, engineers need to think about and control other aspects of an inspection setup to get good visual data:

• Lighting. High-speed imaging requires a lot of light. When we use the Kodak system running at 4500 images/s, we supply ten to twelve 1000-W photographic lights to illuminate the bag undergoing testing. We need to provide enough light to achieve adequate exposure at a lens setting of f/5.6 or f/8, each of which provides a good depth of field.

You also need to account for light reflection and absorption. A black instrument panel will absorb most of the light directed at it, whereas a light gray or tan panel will reflect most light. Because many air bags are white, they require less illumination to bring out details than would an instrument panel. So test engineers must vary the exposure, or the light, to get useful images of key details. Those details help determine whether there are any tears in a bag and, if so, what caused them.

• Lenses. In most cases, a 12-mm lens is optimum for high-speed imaging. With a camera-to-subject distance of 10 to 12 ft, and the lens set at f/5.6 to f/8, a 12-mm lens provides depth of field of 3 to 4 ft and a field of view of approximately 5x5 ft. Photographers familiar with the focal lengths of lenses used in 35-mm photography may think of a 12-mm lens as one that provides an ultra wide angle. But because the CCD array in a high-speed camera is much smaller than a frame of 35-mm film, the 12-mm lens actually produces an image similar in perspective to a “normal” 50-mm lens.

• Point of view. Early air-bag testing involved only side-view images that showed when an air bag began to break through an instrument panel or a steering wheel cover, and when it reached full inflation. Now that auto manufacturers provide side air-bag systems, testing involves side and top views. An imaging system must record both views simultaneously.

• Sensor synchronization. To capture multiple images simultaneously, the test system must accurately trigger several cameras. And the system must synchronize the image data with data acquired from sensors that measure the current in the air-bag ignition coil (squib), air-bag pressure, pressure experienced by a crash-test dummy, and so on. A test may acquire data from as many as 24 transducers.  T&MW

Colin Williams is Product Manager at Microsys Technologies, a company that designs automated test systems for the automotive, occupant safety industry. He has been in the test engineering field for 15 years.

New Air Bags Challenge Test Engineers

Early air bags were relatively simple, straightforward systems. Testing them involved checking whether they inflated and how fast. The development of new types of air bags means test engineers must answer many new questions:

• Does air-bag inflation create particle projectiles as the bag breaks through the instrument panel?
• Does it vent noxious inflating gases into the passenger compartment?
• Does a smart air bag—one that controls its inflation under computer control—inflate with the proper explosive levels depending on the characteristics of the seat’s occupant? Does the bag expose a child to a less-violent explosion?
• Does the bag inflate with the appropriate speed: 30 to 50 ms for a front air bag and 5 to 10 ms for a side air bag?
• How do temperature and humidity affect an air bag’s ability to break through the instrument panel?
• How do they affect inflation time? (A cold environment can increase deployment time by 50% to 100%.)
  
  
  

—Colin Williams



Copyright 1999, Test & Measurement World. Published by Cahners Business Information, Newton, MA.