2. Restraint system architecture and components

A restraint system is not a collection of independent safety devices; it is a coordinated energy-management chain. Every component — seat belt, pretensioner, load limiter, airbag, seat structure, and steering column — is tuned to act within a specific time window and force range so that the occupant decelerates in a controlled manner. Understanding how these elements interact is the foundation for any physical test or CAE correlation activity, because a mismatch in one component's timing or force level propagates through the entire occupant response.

Core Components and Their Functions

The seat belt (three-point restraint) remains the primary occupant restraint. Its webbing geometry defines the initial coupling between occupant and vehicle. Key parameters include webbing stiffness (typically expressed as force per unit elongation), the geometry of the D-ring and buckle anchor points, and the retractor characteristics. In a frontal impact, the belt converts occupant kinetic energy into webbing strain and controlled slip through the retractor.

The pretensioner removes slack from the belt system at the very beginning of a crash event. Pyrotechnic pretensioners typically fire within 10–20 ms after the crash signal, pulling 40–80 mm of webbing back into the retractor or pulling the buckle downward. This early slack removal ensures the occupant engages the belt before significant forward excursion occurs. Without pretension, the occupant would travel further before belt loading begins, increasing the risk of contact with interior structures.

The load limiter controls the maximum force transmitted through the belt to the occupant's chest. It is usually implemented as a torsion bar inside the retractor or as a stitch pattern in the webbing that tears at a calibrated force. A typical load limiter threshold for a frontal occupant is in the range of 3–5 kN at the belt, though this varies with occupant size, seating position, and regulatory or rating requirements. The load limiter works together with the pretensioner: pretensioner first, then load limiter releases force progressively.

The airbag provides a supplementary restraint surface that distributes deceleration forces over a larger area of the occupant's body. Frontal airbags deploy in 20–40 ms and are designed to be fully inflated before the occupant's head and torso reach the airbag surface. Side airbags (thorax, curtain, and combination) deploy faster, often within 10–20 ms, because the available distance between occupant and door trim is much smaller. The airbag's venting characteristics determine how quickly it deflates after peak inflation, which affects the rebound phase of occupant motion.

The seat is not a passive platform. Seat structure stiffness, recliner strength, and cushion foam properties all influence occupant kinematics. In rear impacts, the seat back and head restraint are the primary restraint elements. A seat with excessive compliance allows the occupant's torso to rotate rearward, increasing neck injury risk. The head restraint must be positioned close to the back of the head — a gap of more than 60 mm significantly reduces its effectiveness.

The steering column in frontal impacts acts as a secondary energy absorber. A collapsible steering column compresses or deforms under axial load, absorbing energy that would otherwise be transferred to the driver's chest through the airbag and belt. The column's collapse force is calibrated to complement the belt load limiter, so the combined force on the driver's chest remains within tolerable limits.

Interaction Across Impact Directions

The table below summarizes how the primary restraint elements engage in each impact direction.

In a frontal impact, the sequence is critical. The pretensioner fires first, removing slack. The belt then begins to load the occupant's pelvis and chest. As belt force rises, the load limiter activates, allowing controlled webbing payout. Simultaneously, the airbag inflates and provides a compliant surface for the head and upper torso. The steering column collapses if the driver's chest loads it through the airbag. If any of these events is delayed or out of sequence, the occupant may experience higher peak forces or increased excursion.

In a side impact, the available deformation distance is much shorter. The side airbag must deploy almost immediately, and the seat structure must resist lateral intrusion. The belt plays a limited role because the occupant's motion is primarily lateral, not forward. This is why side impact protection relies heavily on the airbag and the seat frame.

In a rear impact, the seat and head restraint are the primary restraints. The belt's locking retractor prevents the occupant from sliding forward after the initial rearward motion. The seat back must absorb energy without excessive rearward deflection, and the head restraint must limit neck extension.

Practical Example: Frontal Sled Test Correlation

Consider a frontal sled test at 50 km/h with a 50th percentile male dummy. The test setup includes a production seat, a three-point belt with pretensioner and load limiter, a driver airbag, and a collapsible steering column. The measured signals include belt force at the retractor, chest deflection, head acceleration, and steering column displacement.

In a typical correlation exercise, the CAE model is first checked against the physical test for pretensioner timing. If the pretensioner fires 5 ms late in the model, the belt force will rise later, and the chest deflection peak may shift. The load limiter threshold is then adjusted to match the measured belt force plateau. The airbag inflation curve is tuned to match the head acceleration pulse. Finally, the steering column collapse force is adjusted to match the column displacement.

A common failure symptom is a model that predicts lower chest deflection than the test. This often indicates that the load limiter is set too high in the model, or that the belt webbing stiffness is overestimated. Another symptom is a head acceleration peak that occurs too early, which usually means the airbag is inflating too fast or the pretensioner is firing too early. Troubleshooting involves checking the timing of each event against the test data and adjusting the corresponding parameter.

Visualizing the Restraint Interaction Chain

The SVG below illustrates the interaction chain for a frontal impact, showing the sequence of component activation and the resulting occupant motion. It is not a flowchart of steps but a signal-and-force relationship diagram: each component's output feeds into the occupant's deceleration profile.

The diagram shows that the pretensioner and load limiter act on the belt force path, while the airbag and steering column act on the head and chest through separate but coordinated paths. The seat structure provides the reaction platform for all of these forces. The final injury criteria — chest deflection, HIC, neck force, and femur load — are the integrated result of these simultaneous interactions.

Key Parameters for Correlation

When correlating physical tests with CAE models, the following parameters are most sensitive and should be checked first:

• Pretensioner firing time and stroke: A 5 ms delay can shift the entire belt force curve.
• Load limiter threshold: Directly controls chest deflection magnitude.
• Airbag inflation curve: Affects head acceleration timing and peak.
• Seat back stiffness: Influences torso rotation and belt fit.
• Steering column collapse force: Affects driver chest loading in frontal impacts.

These parameters are not independent. Changing the load limiter threshold without adjusting the airbag inflation curve can create a mismatch in the occupant's forward excursion. The correlation engineer must iterate across the full chain, using the test data as the reference for each component's timing and force level.

In the next stage of the correlation workflow, these component-level parameters become the calibration variables. The physical test provides the target signals, and the CAE model is tuned until the simulated occupant response matches the measured chest deflection, head acceleration, and belt force within acceptable tolerances. The architecture described here — the sequence of activation and the force paths — is what makes that tuning possible.