3. Test laboratory infrastructure

Occupant restraint testing does not happen in a generic mechanical lab. It requires a purpose-built facility in which a vehicle body, sled buck, or component can be accelerated or decelerated along a controlled axis while a synchronized instrumentation suite records what happens to the occupant. This chapter examines the physical and data infrastructure that makes such tests repeatable: sled systems, crash halls, fixed barriers, tow tracks, and the data acquisition (DAQ) architecture that ties sensors, triggers, and analysis together.

Sled Systems: Reproducing the Crash Pulse Without a Full Vehicle

A sled system is the workhorse of restraint development. Instead of destroying a complete vehicle, engineers mount a representative body-in-white section, a buck, or a seat-and-restraint assembly on a rigid sled carriage and subject it to a deceleration (or acceleration) pulse that mimics a real crash event. The key advantage is repeatability: the same pulse can be run dozens of times with only the restraint configuration changing.

Sled types differ mainly in how the pulse is generated. A deceleration sled is propelled to a target velocity and then stopped by a programmable braking system—typically hydraulic, servo-hydraulic, or a combination of crushable metal honeycomb and hydraulic dampers. The braking force profile determines the shape of the deceleration pulse. An acceleration sled starts at rest and is accelerated by a pneumatic or hydraulic actuator, which is useful when the test specification calls for a positive acceleration pulse rather than a deceleration.

The pulse itself is defined by a few parameters: peak deceleration (in g), pulse duration (ms), and the integral of the pulse, which equals the velocity change Δv. For a frontal restraint test, a typical pulse might be a 50 km/h Δv with a peak of 30–35 g over roughly 100–120 ms. The sled's control system uses a feedback loop: a pre-programmed target pulse is compared against measured sled acceleration, and the braking or actuation profile is adjusted until the measured pulse falls within a corridor defined by the test standard.

Crash Halls and Fixed Barriers

A crash hall is the large, structurally reinforced space that houses full-scale impact tests. Its defining feature is a massive reaction mass—often a concrete block weighing hundreds of tonnes—against which a vehicle or sled is driven. The barrier face may be flat, angled, or fitted with a deformable element such as a honeycomb or a pole, depending on the test configuration.

For restraint testing specifically, the crash hall must provide:

• A level, straight run-up of sufficient length to reach target velocity before impact.
• A rigid barrier whose deflection under load is negligible compared with the vehicle crush.
• Overhead and floor-mounted lighting synchronized to the high-speed cameras.
• Safe separation between the test article, the cameras, and the operator control room.

Barrier design matters because any compliance in the barrier changes the effective crash pulse. A barrier that flexes absorbs energy that should have gone into the vehicle structure, altering the deceleration history and therefore the occupant loading. For this reason, barriers are periodically surveyed and their stiffness verified.

Tow Tracks and Propulsion

Full-scale tests need a propulsion method that delivers the test article to the barrier at a precise speed and attitude. Tow tracks serve this role. A steel cable or a rigid tow bar connects the vehicle to a drive system—historically a winch or a linear motor, now often a servo-controlled electric or hydraulic drive. The vehicle is guided by the track itself or by an on-board steering robot that holds a commanded path.

Two parameters dominate tow-track setup: release velocity and release attitude. Velocity is measured by a speed trap or by the vehicle's own accelerometer integrated over time; attitude (yaw angle at impact) is set by the steering robot or by the track geometry. For a frontal rigid-barrier test, yaw is nominally zero; for an oblique or small-overlap test, the yaw angle is a controlled variable and must be held within a tight tolerance.

Data Acquisition Architecture

The DAQ system is what converts a physical event into interpretable numbers. In restraint testing, the architecture typically has four layers: sensors, signal conditioning, digitizers, and the recording/analysis host.

Sensors. The core measurements are accelerations (vehicle, sled, and occupant locations), loads (seat belt force, load-cell-equipped anchors), displacements (belt payout, steering column stroke), and sometimes pressure or angular rate. Occupant accelerometers are mounted on the head, chest, and pelvis of the ATD (anthropomorphic test device).

Signal conditioning. Charge-mode accelerometers require charge amplifiers; strain-gauge load cells require bridge excitation and amplification. Anti-aliasing filters are applied here, with a cutoff set according to the channel class—commonly 1,000 Hz for occupant accelerations and 600 Hz for belt loads, following SAE J211 conventions.

Digitizers. Signals are sampled simultaneously across all channels. A typical system samples at 10–20 kHz per channel with 16-bit resolution, giving sufficient oversampling relative to the anti-aliasing filter. Channel-to-channel skew must be kept in the low microseconds; otherwise, timing comparisons between, say, belt force and head acceleration become unreliable.

Recording and analysis host. The host stores raw and filtered data, applies the J211 filter class, and computes occupant injury criteria such as HIC, chest deflection, and Nij. A common trigger architecture uses a pre-trigger buffer: the system records continuously into a circular buffer and, on a trigger event (barrier contact, sled brake release, or a manual arm signal), freezes the buffer so that pre-impact data is preserved.

Example channel map (frontal sled test, 24 channels):
  CH01-03  Sled accelerometer X/Y/Z        10 kHz, CFC 1000
  CH04-06  Head accelerometer X/Y/Z        10 kHz, CFC 1000
  CH07-09  Chest accelerometer X/Y/Z       10 kHz, CFC 1000
  CH10-12  Pelvis accelerometer X/Y/Z      10 kHz, CFC 1000
  CH13-14  Belt load (shoulder/lap)        10 kHz, CFC 600
  CH15     Belt payout displacement        10 kHz, CFC 600
  CH16     Steering column stroke          10 kHz, CFC 600
  CH17-18  Seat anchor loads               10 kHz, CFC 600
  CH19     Sled velocity (integrated)      10 kHz
  CH20     Trigger / time-zero             10 kHz
  CH21-24  Spare / camera sync             10 kHz

Time zero is a critical convention. It is usually defined as the moment of barrier contact or the instant the sled brake is released, and every channel must share that reference. If time zero drifts between channels, injury criteria computed from different signals will be misaligned.

Putting the Infrastructure Together

The diagram below shows how the major subsystems connect for a typical frontal sled test. The sled carries the buck and ATD; the braking system generates the pulse; the DAQ front-end digitizes sensor signals; the host records and analyzes; and the trigger synchronizes cameras and data.

Practical Verification and Common Failure Modes

Before a sled test is considered valid, the infrastructure itself must be verified. A pre-test checklist typically includes: pulse corridor run with a dummy mass, channel calibration check (shunt or known input), time-zero alignment across all channels, camera frame synchronization, and a dry-run trigger to confirm the pre-trigger buffer captures the intended window.

Common failure symptoms and their likely causes:

Understanding this infrastructure is what separates a test that produces numbers from a test that produces trustworthy evidence. The sled, the hall, the barrier, the tow track, and the DAQ chain are not independent—each one constrains what the others can achieve, and the validity of every occupant injury criterion depends on all of them working within specification.