4. Intrusion timing measurement

Intrusion timing is the foundation on which every restraint design window is built. Before you can argue that a side airbag fires late enough to avoid being struck by a closing door, or that a B-pillar reinforcement keeps the occupant compartment intact long enough for the bag to deploy, you must be able to measure when the intruding structure moves and how far it moves. This chapter covers the three measurement techniques that form the backbone of intrusion timing work: accelerometers, string potentiometers, and high-speed video analysis. Each has a distinct role, and in a well-instrumented test they are used together so that their strengths compensate for one another.

Defining the Measurement Reference

Before any sensor is mounted, fix the coordinate convention. For side impact work the standard practice is to define the intrusion direction as positive — that is, motion of the door or B-pillar toward the occupant is recorded as a positive displacement and a positive acceleration. The vehicle's longitudinal axis (X) runs fore-aft, the lateral axis (Y) runs across the vehicle, and the vertical axis (Z) runs up-down. A door inner panel moving inboard toward the seat is moving in the +Y direction. Every sensor channel, every video overlay, and every plot in this chapter uses this convention. If a sensor is mounted with its sensitive axis reversed, the sign must be corrected in post-processing and the correction documented, otherwise the timing analysis will invert and the restraint window will be computed backwards.

Accelerometers: Capturing the Onset

Accelerometers are the primary tool for detecting the instant of impact and the early acceleration pulse. A piezoelectric or piezoresistive accelerometer mounted on the door inner panel or the B-pillar lower section, with its sensitive axis aligned to +Y, produces a voltage proportional to acceleration. The key parameters are range, sensitivity, and bandwidth. For side impact door work, a range of ±500 g to ±2000 g is typical, with a sensitivity around 0.1 to 0.5 mV/g and a frequency response flat to at least 5 kHz. A common choice is a 1000 g range, 0.2 mV/g unit with a 10 kHz low-pass filter set in the data acquisition system.

The accelerometer's value is its time resolution. The first contact between the barrier and the door produces a sharp acceleration spike, and the time from that spike to the peak intrusion velocity is often only 5–15 ms. An accelerometer sampled at 20 kHz or higher captures this onset cleanly. However, acceleration alone does not give displacement directly — it must be integrated twice, and integration drift accumulates rapidly. A 1000 g bias error of even 0.5% integrated over 30 ms produces a velocity error of roughly 1.5 m/s and a displacement error of tens of millimeters. This is why accelerometers are used for timing and onset detection, not for absolute displacement.

String Potentiometers: Direct Displacement

A string potentiometer (draw-wire displacement sensor) solves the integration problem by measuring displacement directly. The sensor body is fixed to a rigid reference — typically the seat frame, the tunnel, or a dedicated rigid fixture — and the string is attached to the moving door or B-pillar point. As the structure intrudes, the string extends and the potentiometer outputs a voltage proportional to displacement. A typical unit for this application has a 250–500 mm range, a resolution of 0.1 mm, and a response fast enough to track intrusion over 30–50 ms.

The critical installation detail is the string's line of action. The string must be aligned with the intrusion direction (+Y) as closely as possible. If the string is at an angle, the measured extension is the projection of the true displacement onto the string axis, and the reading underestimates intrusion by the cosine of the misalignment angle. A 15° misalignment produces a 3.4% underestimate — small but not negligible when the design window is tight. More importantly, if the attachment point rotates during intrusion, the string angle changes and the error becomes time-varying. Mount the string anchor on a point that translates rather than rotates, or accept and document the error.

String pots also have a maximum extension rate. If the door intrudes at 10 m/s, the string must pay out at 10 m/s, and some units will lag or fail to track. Check the specified maximum velocity against the expected peak intrusion velocity before the test. In practice, string pots are excellent for the 0–30 ms window where intrusion velocity is moderate, and they provide the ground-truth displacement that anchors the accelerometer integration.

High-Speed Video: The Spatial and Temporal Cross-Check

High-speed video at 1000–5000 frames per second provides an independent measurement of both timing and displacement, and it is the only technique that shows what is moving. A camera viewing the door and B-pillar from inside the vehicle, or through a window opening, records the intrusion event frame by frame. With a calibrated target grid or tracked markers on the door inner panel and B-pillar, displacement can be extracted to within a few millimeters, and the frame timestamp gives timing to within one frame interval (0.2–1 ms).

Video analysis is slower and more labor-intensive than sensor data, but it catches failure modes the sensors miss: a door that separates from its hinges, a trim panel that detaches and moves independently of the structure, or a string pot whose string goes slack. When the accelerometer-derived displacement and the string pot disagree, the video usually explains why. A practical workflow is to use the video to validate the sensor mounting and to identify the exact frame of first contact, then use that frame time to align the accelerometer and string pot records.

Aligning the Three Records

The three measurement channels are only useful once they share a common time base. The standard approach is a common trigger: the barrier or impactor contact closes a switch that simultaneously starts the data acquisition system and marks the video. In post-processing, the first frame showing contact is identified, and its timestamp is set as t = 0. The accelerometer record is then shifted so that its first significant spike aligns with that frame. The string pot record is shifted by the same amount. If the video frame rate is 2000 fps, the alignment uncertainty is ±0.25 ms, which is acceptable for restraint timing work where windows are typically 5–20 ms wide.

A useful cross-check is to compare the string pot displacement at 10 ms and 20 ms against the video-tracked displacement at the same times. If they agree within 2–3 mm, the mounting and alignment are sound. If the string pot reads consistently low, suspect string misalignment or anchor rotation. If the accelerometer-derived displacement diverges from the string pot after 15 ms, suspect integration drift and re-zero the integration at the string pot's value at that time.

Practical Measurement Checklist

With the three records aligned and cross-validated, you have a defensible intrusion timing curve: onset time from the accelerometer, absolute displacement from the string pot, and spatial confirmation from the video. That curve is the input to the restraint design window analysis in the chapters that follow — the time at which the door reaches a given intrusion depth determines how long the airbag has to deploy before the gap closes.