4. Restraint performance indicators
In a crash event, the restraint system is a coordinated chain of energy-management devices, not a collection of independent parts. The seat belt, pretensioner, load limiter, and airbag each act on the occupant within a narrow time window, and their combined effect determines the deceleration profile the body experiences. For injury analytics, four indicator families carry most of the diagnostic weight: belt force at the load limiter, airbag internal pressure, pretensioner firing timing, and the resulting occupant kinematics. This chapter explains what each indicator physically represents, how it is measured or derived, and how the four interact so that a dashboard can attribute an injury outcome to a specific restraint behavior rather than to a generic "restraint event."
Belt load limiting: force as a controlled ceiling
A load-limiting retractor or buckle anchorage deliberately yields once belt force exceeds a threshold, allowing a small amount of webbing to pay out. The purpose is to cap the force transmitted to the chest and shoulder while still coupling the occupant to the vehicle deceleration. The key indicator is the belt force–time history at the load-limiting element, typically measured with a load cell in the buckle or derived from retractor payout.
Two parameters dominate interpretation. The limit force level (commonly in the 3–6 kN range for front occupants, depending on regulation and vehicle class) sets the plateau the force trace reaches after the initial spike. The payout onset time and total payout distance determine how much additional occupant excursion the limiter permits. A force trace that plateaus cleanly indicates the limiter is functioning; a trace that spikes well above the design level and stays there suggests the limiter did not activate, which correlates with elevated chest deflection and rib injury risk. Conversely, an excessively low plateau with large payout can allow the occupant to travel too far forward, increasing head excursion and the risk of contact with the interior or an out-of-position airbag interaction.
Airbag pressure: the inflation energetics indicator
Airbag pressure is the most direct proxy for the energy delivered to the occupant by the frontal airbag. In test instrumentation, pressure is recorded by a transducer inside the module or derived from tank pressure and inflator mass-flow characteristics. The indicator of interest is the pressure–time curve, characterized by peak pressure, time-to-peak, and the venting decay after peak.
Peak pressure and time-to-peak together describe how aggressively the bag presents its surface to the occupant. A bag that peaks too early and too high can strike an occupant who is still moving forward at high velocity, producing high chest acceleration and head-neck tension. A bag that peaks late leaves the occupant traveling into a partially inflated cushion, which can bottom out against the module. Venting behavior matters as well: a bag that holds pressure too long behaves stiffly, while aggressive venting reduces the cushioning but may allow the occupant to contact structure. In analytics, airbag pressure is best interpreted alongside occupant position at the moment of peak pressure, because the same pressure curve produces very different loading depending on where the chest and head are.
Pretensioner timing: removing slack before the main event
The pretensioner retracts webbing to eliminate slack and couple the occupant to the restraint early in the crash pulse. Its critical indicator is firing time relative to the crash pulse, usually expressed as time after the trigger (e.g., 10–20 ms after algorithm decision) and the resulting webbing retraction distance. Timing is a double-edged parameter. Firing too late leaves slack that the occupant must take up through forward excursion, delaying effective restraint and increasing peak force when the belt finally engages. Firing too early, before the occupant has begun to move, can pull the occupant into an unfavorable posture or pre-load the belt against a body that is about to rebound, and in some configurations it can worsen chest loading by removing the natural compliance of the initial belt stretch.
In a dashboard, pretensioner timing should be shown as a marker on the same time axis as the crash pulse and the belt force trace, so the analyst can see whether retraction completed before the force plateau began. A useful derived indicator is the slack-removal interval: the time between pretensioner fire and the first sustained rise in belt force.
Occupant kinematics: the outcome the other three produce
Kinematics is the observable result of the restraint chain: forward excursion of the head and chest, torso pitch, and the timing of peak displacement relative to airbag and belt loading. It is measured from high-speed video, from accelerometer and angular-rate sensors on the occupant, or reconstructed from a validated occupant model. The indicators that matter for injury analytics are peak forward head excursion, peak chest forward excursion, and the phase relationship between those peaks and the restraint loading events.
The phase relationship is where the four indicators become a single story. If head excursion peaks while airbag pressure is still rising, the head meets a stiffening cushion; if it peaks after venting has begun, the head may pass through a softened bag toward structure. If chest excursion peaks before the belt force plateau, the limiter is engaging too late to shape the chest deceleration. These are the attributions a crash-response dashboard must support.
Reading the four indicators together
The example timeline above is illustrative, not measured data, but it demonstrates the analytical method. The pretensioner fires at roughly 20 ms, and belt force begins its sustained rise shortly after, indicating slack was removed before the occupant loaded the belt. The load limiter plateaus the force through the middle of the event. Airbag pressure peaks before head excursion peaks, meaning the head arrives as the bag is venting — a configuration that softens head loading but risks bottoming out if excursion is large. Chest excursion peaks while the belt plateau is still active, so the limiter is doing its job of shaping chest deceleration.
When the phase relationships break down, the dashboard should flag the specific mismatch rather than a generic "restraint anomaly." Three common patterns are worth encoding as diagnostic rules:
- Late pretensioner relative to belt force rise: slack-removal interval exceeds the expected window, suggesting delayed firing or excessive initial slack; correlates with higher peak belt force and larger chest excursion.
- Belt force above the limiter ceiling: plateau exceeds design level, indicating limiter non-activation or a jammed payout path; correlates with elevated chest deflection.
- Head excursion peak before airbag peak: the head meets a still-inflating bag, producing high head acceleration and neck tension; often paired with early pretensioner timing that pulled the occupant forward.
Practical measurement and attribution workflow
To make these indicators usable in a dashboard, the analyst needs a consistent derivation pipeline. A workable sequence for a frontal sled or vehicle test is:
- Time-align all channels to the crash pulse trigger (t = 0 at pulse onset), and document the alignment convention so signs and offsets are consistent across belt force, pressure, and kinematics.
- Extract belt force plateau level and onset time from the buckle load cell; compute slack-removal interval as the time from pretensioner fire command to the first sustained force rise above a defined noise threshold.
- Extract airbag peak pressure, time-to-peak, and venting half-time from the pressure trace.
- Extract head and chest peak excursion and their times from video tracking or occupant sensors; compute the phase offsets against airbag peak and belt plateau.
- Compare each derived indicator against the design envelope for the occupant class and impact severity, and record the mismatch pattern if any indicator falls outside.
The value of treating these four indicators as a synchronized set is that it converts a pass/fail restraint result into a causal narrative: which device acted, when, and how the occupant responded. That narrative is what a crash-response dashboard must surface, and it is the foundation for the injury-analytics chapters that follow.