2. Airbag deployment physics and occupant restraint fundamentals

An airbag is not a cushion that gently catches an occupant. It is a short-duration gas-management system whose entire useful life is measured in tens of milliseconds. Understanding deployment physics means understanding four coupled variables: pressure (how much gas is inside the cushion), venting (how fast that gas escapes), fabric (how the cushion stretches, leaks, and survives the event), and timing (when the cushion arrives relative to the occupant's motion). Correlation robustness studies exist precisely because these four variables interact, and small changes in any one of them shift the outcome of the restraint event.

The Deployment Timeline

Consider a frontal impact at 50 km/h. The vehicle begins decelerating at roughly 200–400 g within the first 10–20 ms. The occupant, restrained primarily by the seat belt, continues forward. The airbag control unit (ACU) must decide to fire, the inflator must generate gas, and the cushion must fill the gap between the module and the occupant before the occupant reaches the steering wheel or instrument panel.

A representative timeline for a driver frontal airbag looks like this:

Time (ms)	Event	Dominant physics
0	Impact begins; ACU accelerometer sees deceleration	Crash pulse onset
10–20	Fire decision; squib current applied	Algorithm threshold crossing
20–30	Inflator generates gas; cushion begins to unfold	Pyrotechnic gas generation
30–50	Cushion reaches full deployment; internal pressure peaks	Pressure rise, fabric unfolding
50–80	Occupant contacts cushion; pressure rises again	Momentum transfer, venting begins
80–150	Cushion vents gas; occupant decelerates	Controlled energy absorption

The critical insight is that the cushion must be in position before the occupant arrives, but it must not be so stiff that it injures the occupant. This is the fundamental tension in airbag design: early enough to be useful, soft enough to be safe.

Pressure: The Gas Budget

Internal cushion pressure is governed by the ideal gas law applied to a rapidly changing volume. When the inflator discharges, gas mass enters the cushion at a rate determined by the propellant burn. Pressure rises until the cushion fabric unfolds and the vent holes open. The peak pressure depends on:

• Inflator output — total gas mass and mass flow rate, typically characterized by a tank pressure curve.
• Cushion volume — a larger cushion requires more gas to reach the same pressure.
• Vent area — larger vents bleed pressure faster, lowering peak pressure but also reducing the cushion's ability to absorb energy.
• Fabric permeability — gas that leaks through the weave acts as a distributed vent.

In a typical driver airbag, peak internal pressure during deployment can reach 200–400 kPa (absolute) for a few milliseconds, then drop to 30–80 kPa during occupant loading. These are order-of-magnitude figures for illustration; actual values depend on inflator type, cushion geometry, and ambient temperature.

Venting: Controlled Energy Dissipation

Venting is the airbag's primary mechanism for absorbing occupant energy. As the occupant compresses the cushion, gas is forced out through vent holes and through the fabric itself. The work done on the gas equals the energy removed from the occupant. If venting is too aggressive, the cushion bottoms out and the occupant contacts hard structure. If venting is insufficient, the cushion acts like a rigid balloon and the occupant rebounds with high force.

Vent design involves trade-offs:

• Vent hole diameter and count — the primary tuning parameter. A common approach is two to four vents of 20–40 mm diameter for a driver airbag.
• Vent location — vents near the occupant contact zone open earlier; vents on the side or rear open later.
• Active vents — some systems use tethers or flaps that open only at certain pressures, providing a two-stage venting profile.

In correlation studies, vent area is often the parameter with the highest sensitivity. A 10% change in effective vent area can shift peak occupant chest deflection by several millimeters, which is significant when the injury threshold is defined in millimeters.

Fabric: The Structural Envelope

The cushion fabric must contain high-pressure gas, survive the deployment event without tearing, and fold into a compact module. The two dominant materials are nylon 6,6 and polyester, typically coated with silicone or neoprene to control permeability.

Fabric behavior matters in three ways:

1. Permeability — uncoated fabric leaks gas through the weave. Coating reduces permeability but adds mass and stiffness. Permeability is often expressed as a leak rate at a reference pressure, and it changes with temperature and strain.
2. Stiffness and strain rate sensitivity — at deployment strain rates (10²–10³ s⁻¹), fabric behaves differently than in quasi-static testing. Correlation models must account for this rate dependence.
3. Fold pattern and seam design — how the cushion is folded determines the unfolding sequence and the initial pressure rise. A poorly folded cushion can deploy asymmetrically, causing the occupant to load one side of the cushion before the other.

Timing: The Occupant's Clock

Timing is not just "when the airbag fires." It is the synchronization of three clocks: the crash pulse, the inflator output, and the occupant's forward excursion. The ACU algorithm uses acceleration data to estimate crash severity and decide fire time. A typical fire time for a 50 km/h frontal impact is 10–20 ms after impact. But the occupant's position at the moment of cushion contact depends on belt geometry, seat position, and occupant size.

Out-of-position (OOP) scenarios are the hardest timing cases. If an occupant is leaning forward when the airbag deploys, the cushion may contact the occupant before it is fully inflated, or the occupant may be in the path of the deploying cushion. This is why OOP testing and suppression logic exist, and why timing robustness is a central concern in correlation studies.

How the Four Variables Couple

The SVG above shows the coupling explicitly. The inflator delivers gas mass; the cushion volume and fabric permeability determine how pressure builds; the vent path bleeds pressure; and the occupant's motion feeds back into the cushion as a compression load. Timing determines when the occupant enters this loop.

In a correlation study, you cannot tune one variable in isolation. If you increase vent area to reduce peak pressure, you also reduce the cushion's energy absorption, which may increase occupant chest deflection. If you increase fabric permeability to soften the cushion, you may reduce the pressure available to keep the cushion inflated during occupant loading. The design space is multidimensional, and robustness studies map how sensitive the outcome is to each dimension.

Practical Correlation Checkpoints

When building or validating a deployment model, the following checkpoints help isolate which physics is driving a mismatch:

• Pressure-time history at the cushion center — compare peak pressure, time-to-peak, and decay slope. A model that matches peak but not decay usually has a venting or permeability error.
• Occupant chest deflection and acceleration — these are the injury-relevant outputs. A model that matches pressure but not deflection likely has a fabric stiffness or contact geometry issue.
• Deployment kinematics from high-speed video — check unfolding sequence and time to full deployment. Asymmetric deployment points to fold pattern or seam modeling errors.
• Temperature sensitivity — run the model at −35 °C, 23 °C, and 85 °C. Inflator output and fabric stiffness both change with temperature; a model that only matches at 23 °C is not robust.

The next step in a robustness study is to assign uncertainty distributions to these parameters — vent area, permeability, inflator burn rate, fire time — and propagate them through the model. But that only makes sense once the baseline physics is understood and the model is correlated at nominal conditions. The four variables described here are the foundation on which every subsequent uncertainty study is built.