4. Instrumentation for PAB impact testing

Passenger airbag (PAB) impact testing produces a short, violent event: the bag deploys in roughly 20–50 ms, the Hybrid III dummy's head and chest load the bag within another 30–80 ms, and the entire measurement window of interest is often under 200 ms. Every sensor on the sled or body-in-white must therefore capture high-amplitude, high-frequency transients without clipping, without aliasing, and without losing synchronization with the high-speed video. This chapter covers the five instrumentation families that make a PAB test defensible: accelerometers, load cells, pressure sensors, high-speed cameras, and the data acquisition system (DAS) that ties them together.

Accelerometers

Accelerometers dominate PAB instrumentation because occupant injury criteria are defined in terms of acceleration. The two most common types in this application are piezoresistive (MEMS) and piezoelectric. Piezoresistive sensors are preferred for dummy head and chest measurements because they respond to DC (true static acceleration) and are used in the standard Hybrid III head and chest accelerometer packages. Piezoelectric sensors are used where high-frequency content matters, such as airbag module housing acceleration or steering-column intrusion.

Selection is driven by range, sensitivity, and frequency response. A typical dummy head accelerometer has a range of ±2000 g and a sensitivity near 0.5 mV/g, while a chest accelerometer may be ±500 g. The sensor's mounted resonance should be at least five times the highest frequency of interest; for PAB work, the useful band is usually DC to 1650 Hz (CFC 1000 filtering per SAE J211), so a sensor with a mounted resonance above 10 kHz is safe. Mounting is critical: a stud-mounted accelerometer with a thin layer of cyanoacrylate or a rigid adapter block preserves high-frequency fidelity, whereas a wax or tape mount can introduce a 1–3 kHz resonance that corrupts the data.

Load Cells

Load cells measure the force transmitted through a structure. In PAB testing they appear in three places: the steering column or instrument-panel reaction structure, the seat belt anchorages, and the dummy's femur or tibia load cells. Most are strain-gauge bridges with a full Wheatstone configuration, giving an output of 1–3 mV/V at rated load. A 5 kN column load cell with 2 mV/V sensitivity and 10 V excitation produces 20 mV at full scale, which is well within the DAS input range but requires a low-noise amplifier and a stable excitation supply.

Because the airbag inflator fires with a pyrotechnic impulse, load cells on the module reaction structure see a sharp spike at t = 0 followed by the occupant loading pulse. The cell must be mechanically isolated from the inflator housing so that the spike does not ring the cell's own resonance. A common practice is to mount the load cell between the module bracket and the cross-car beam with a compliant washer that damps frequencies above 2 kHz.

Pressure Sensors

Pressure sensors are used inside the airbag cushion to characterize the inflation profile and to detect bag-bottoming. Two technologies are common: piezoresistive absolute sensors (0–300 kPa range) and piezoelectric dynamic pressure sensors. The piezoresistive type is preferred because it captures the static cushion pressure before and after deployment, which is needed to compute the bag's effective stiffness. A typical cushion pressure trace rises from ambient to 150–250 kPa within 15–25 ms, then decays as the vent holes and fabric porosity bleed gas.

Sensor placement matters more than sensor accuracy. A single pressure tap at the cushion center may miss local over-pressure near the occupant contact patch. Best practice is to use two or three taps: one near the inflator outlet, one at the geometric center, and one on the occupant-facing surface. Each tap must be flush with the inner fabric to avoid a stagnant volume that low-pass filters the signal.

High-Speed Cameras

High-speed video provides the spatial context that no transducer can: bag unfolding, occupant kinematics, and contact timing. For PAB testing, a frame rate of 1000–2000 fps at 1280×720 resolution is typical, with a shutter time of 1/2000 s or shorter to freeze motion blur. Two or three cameras are usually synchronized: one side view of the occupant, one front view of the bag, and optionally one overhead view of the instrument panel.

Lighting is the limiting factor. At 2000 fps with a 1/2000 s shutter, the exposure per frame is 0.5 ms, so high-output LED panels or metal-halide lamps are required. A common setup uses four 1000 W LED panels at 1.5 m distance, giving roughly 20,000 lux on the subject. Camera synchronization is achieved through a common trigger pulse from the DAS, with a timing accuracy better than 1 ms.

Data Acquisition Systems

The DAS is the backbone of the test. It must sample all channels simultaneously at a rate high enough to avoid aliasing. SAE J211 recommends a minimum of 10 times the highest frequency of interest; for CFC 1000 filtering, that means a sample rate of at least 20 kHz per channel. In practice, PAB tests use 20–50 kHz per channel with 16-bit resolution.

Key DAS specifications for PAB work include:

• Channel count: 32–64 channels for a full occupant test (head, chest, pelvis accelerometers; column, belt, and femur load cells; cushion pressure; inflator current).
• Anti-alias filtering: a low-pass filter at 0.4 × sample rate, typically a 4-pole Butterworth, applied before digitization.
• Excitation: regulated 5–10 V DC for strain-gauge bridges, with less than 0.1% drift over the test duration.
• Trigger: a pre-trigger buffer of at least 100 ms so that the inflator fire signal and any pre-impact events are captured.
• Time synchronization: a common time base shared with the high-speed cameras, usually via IRIG-B or a dedicated trigger pulse.

A typical channel list for a PAB test is shown below. The table is an example configuration, not a fixed standard; the exact channel set depends on the test matrix and the injury criteria being evaluated.

Signal Chain and Synchronization

The figure below shows the complete signal chain from sensor to recorded data, including the trigger and time-sync path that keeps transducers and cameras aligned.

Verification and Troubleshooting

Before a PAB test is considered valid, the instrumentation must pass three checks. First, a shunt calibration on every strain-gauge channel confirms the bridge and amplifier are functioning. Second, a tap test on each accelerometer verifies the mounted resonance and cable integrity; a shift in resonance above 10% from the baseline indicates a loose mount. Third, a time-zero check confirms that the inflator current channel and the camera trigger pulse align within 0.5 ms.

Common failure symptoms and their causes:

• Clipped accelerometer signal: sensor range too low or amplifier gain too high. Check the peak g level from a previous test and select a sensor with at least 2× margin.
• 50/60 Hz noise on load cells: poor grounding or a ground loop between the DAS and the test fixture. Use a single-point ground and shielded twisted-pair cable.
• Pressure trace with a slow rise: a stagnant volume at the pressure tap. Re-drill the tap flush with the cushion inner surface.
• Camera and DAS out of sync: trigger cable length mismatch or missing IRIG-B connection. Verify with a LED flash visible in the camera frame and a simultaneous pulse on the DAS.

With the instrumentation chain verified, the test engineer can trust that the recorded accelerations, loads, pressures, and video frames represent the same physical event within a known uncertainty. That trust is the foundation for the CAE correlation work in later chapters, where simulation results are compared channel-by-channel against these measurements.