2. Restraint system fundamentals

Every sled run, crash test, and component bench test in occupant safety ultimately interrogates a small set of physical devices. Before you can design a test matrix, instrument a dummy, or interpret a load curve, you must know exactly what each device is, how it is built, where it mounts, and which physical quantity it is supposed to control. This chapter treats the five core restraint hardware families — seat belts, pretensioners, load limiters, airbags, and child restraint systems (CRS) — as hardware objects, not as abstract safety features.

Seat belts: webbing, geometry, and the three-point path

A modern three-point belt is a load path, not a single part. The webbing itself is typically polyester, woven to roughly 46–50 mm width for the lap and shoulder sections, with a tensile strength in the range of 25–30 kN per strap in a new, undamaged state. The physical test object is the complete assembly: retractor, webbing, tongue and buckle, anchorages, and the D-ring or height adjuster that defines the shoulder path.

What matters for testing is the geometry the belt imposes on the occupant. The lap belt should sit across the pelvic bones, not the abdomen; the shoulder belt should cross the sternum and clavicle. In a frontal sled test, the belt geometry determines the initial contact point with the dummy and therefore the chest deflection and pelvis excursion you will measure. A useful teaching example: a 50th-percentile male dummy in a standard frontal configuration typically shows shoulder belt force peaking between 6 and 9 kN in a 50 km/h sled pulse, but that number is meaningless without the retractor lock-up time and the pretensioner firing time.

Retractors come in two relevant types for testing: emergency locking retractors (ELR), which lock on vehicle deceleration or webbing payout rate, and automatic locking retractors (ALR), which lock continuously once webbing is pulled out. ELR lock-up is triggered by a vehicle-sensitive pendulum or ball, or by a webbing-sensitive mechanism. In a test, you must verify which mechanism is active and at what threshold, because a retractor that locks too late allows excess payout and increases occupant excursion.

Pretensioners: removing slack before the load arrives

A pretensioner is a pyrotechnic or mechanical device that retracts webbing or moves the buckle to remove slack in the first 10–20 ms of a crash. There are three common physical configurations:

• Retractor pretensioner — a pyrotechnic charge drives a piston or ball chain that spins the retractor spool backward, pulling webbing in.
• Buckle pretensioner — the buckle stalk is pulled downward by a cable driven by a pyrotechnic actuator, tightening the lap portion.
• Anchor pretensioner — the outboard lap anchor is pulled rearward, tightening the lap belt directly.

For the test engineer, the pretensioner is defined by three parameters: stroke (how much webbing or buckle travel it provides, typically 60–120 mm), firing time (when the signal arrives relative to the crash pulse, often 5–15 ms after T0), and force (the retraction force, which must be enough to overcome friction but not so high that it injures the occupant). A pretensioner that fires too early can pull the occupant into an out-of-position state before the airbag deploys; one that fires too late provides no benefit.

Load limiters: capping the force the belt delivers

A load limiter is a force-limiting element in the belt path — usually a torsion bar inside the retractor, a tearing stitch in the webbing, or a deformable metal element. Its purpose is to cap the belt force at a level the occupant's chest can tolerate, typically 3–6 kN for a frontal restraint. The physical object is the limiter itself, and its behavior is defined by a force–displacement curve: it yields at a threshold force, then maintains that force while it deforms over a controlled stroke.

In a test, the load limiter interacts directly with the pretensioner. The pretensioner removes slack and starts the load; the load limiter then controls how much force reaches the chest. If you test a belt with a 4 kN limiter but the pretensioner stroke is too short, the occupant may still bottom out on the limiter before the airbag is in position. The observable result is a belt force plateau at 4 kN with continued chest deflection — a sign that the limiter is working but the overall restraint timing is wrong.

Airbags: inflators, cushions, and deployment timing

An airbag module consists of an inflator (pyrotechnic, hybrid, or stored-gas), a cushion (woven nylon or polyester), a housing, and a cover. The physical test object is the module as installed, because deployment is highly sensitive to the mounting angle, the cover tear pattern, and the surrounding structure.

Key parameters for testing are:

In a frontal sled test, the airbag must be fully inflated before the occupant's chest reaches the module. If the dummy contacts a partially inflated cushion, you will see a sharp chest acceleration spike and a belt force that drops as the bag absorbs the load. The test engineer's job is to verify that the deployment timing, cushion pressure, and venting match the occupant's kinematics.

Child restraint systems: a different physical object

A CRS is not a scaled-down adult belt. It is a separate structure — shell, harness, buckle, adjuster, and often a base — that must manage the child's motion through its own load path. The physical test object is the CRS as installed in a vehicle seat, with the vehicle belt or ISOFIX/LATCH anchors providing the primary attachment.

For testing, the critical parameters are:

• Harness tension — typically checked with a tension gauge; a loose harness allows excessive forward excursion.
• Shell stiffness — determines how much the CRS deforms under load.
• Rebound control — how the CRS returns after the crash pulse; excessive rebound can cause head contact with the vehicle interior.
• Anchor geometry — ISOFIX connector spacing and latch engagement must be verified before every test.

A common failure mode in CRS testing is a harness that appears tight but slips under load because the adjuster does not lock. The observable symptom is a sudden increase in forward head excursion beyond the expected value, with the harness webbing showing a slack loop after the test.

How the objects interact in a single test

In a frontal sled test, the sequence is roughly: at T0 the crash pulse begins; the retractor locks within 10–20 ms; the pretensioner fires at 5–15 ms and removes 60–120 mm of slack; the occupant begins to move forward; the load limiter engages at its threshold force and holds the belt force at 3–6 kN; the airbag reaches full inflation at 20–40 ms and provides a second load path through the chest and head. The test engineer observes the belt force curve, the chest deflection, the head acceleration, and the airbag pressure. If the belt force plateaus too early, the load limiter is too aggressive; if the chest deflection exceeds the injury criterion, the limiter is too high or the pretensioner is too slow.

Understanding these physical objects — their construction, their mounting, their timing, and their interaction — is the prerequisite for every subsequent chapter. The next time you look at a sled test setup, you should be able to point to each device and state what it is supposed to do, what it actually does, and which measurement tells you whether it worked.