2. Airbag system fundamentals

Before any deployment test or CAE correlation study can be meaningful, the analyst must understand what is actually being tested. An airbag module is not a single part but an assembly whose behavior during deployment is governed by the interaction of four subsystems: the housing and mounting structure, the inflator, the cushion, and the folding pattern. Each contributes distinct physics, and each introduces distinct sources of variability that later appear as scatter in test data or as divergence between physical and simulated results.

Module Architecture

A driver-side frontal airbag module typically consists of a stamped or cast metal housing (often aluminum or steel), a retaining ring, the inflator, the folded cushion with its tethers, a trim cover with a tear seam, and the electrical connector with its shorting clip. The passenger-side module is larger and usually mounts the cushion to a chute or door assembly rather than a simple cover. Side airbags add further variants: seat-mounted thorax bags, curtain bags routed along the roof rail, and inflatable seat belts.

The housing performs three functions: it reacts the inflator thrust load into the vehicle structure, it contains the folded cushion before deployment, and it defines the initial direction of cushion expansion. In CAE, the housing is usually represented as rigid or as a stiff elastic body, because its deformation is small compared with the cushion's. The retaining ring and fasteners, however, are common failure points in real deployments and should not be idealized away when the study targets module integrity rather than occupant kinematics.

The trim cover tear seam is a particularly important detail. Its opening pressure and tear propagation determine when the cushion first becomes visible and how it emerges. A seam that opens too slowly delays the cushion, while one that opens asymmetrically can deflect the cushion off-axis. In correlation work, the tear seam is often modeled with a simplified failure criterion rather than a full fracture model, and this simplification is one of the first places where simulation and test diverge.

Inflator Types and Their Output Characteristics

Inflators fall into three broad families, and the choice of family strongly shapes the pressure-time history that drives the cushion.

For CAE, the inflator is almost never modeled at the level of propellant chemistry. Instead it is represented by a tank test curve: mass flow rate versus time, or pressure versus time measured in a fixed-volume tank. This curve is the single most important input to any deployment simulation, and its accuracy sets the ceiling on correlation quality. A common practical error is to use a tank curve measured at one temperature for a simulation intended to represent another; inflator output is strongly temperature dependent, and a −35 °C deployment can differ substantially from a +85 °C deployment in both peak pressure and rise time.

Cushion Materials and Construction

Most frontal cushions are woven from nylon 6.6 yarn, typically in the range of 420 to 840 denier, with a silicone or polyurethane coating on one or both faces to control permeability. The coating is not cosmetic: it governs how much gas escapes through the fabric, which in turn controls cushion pressure and stiffness. Uncoated or lightly coated fabrics vent more aggressively and produce a softer, slower cushion; heavily coated fabrics retain pressure and produce a firmer cushion.

Two material properties dominate deployment behavior. The first is the biaxial tensile response of the coated fabric, which determines how the cushion inflates and how much it stretches before reaching its working shape. The second is the seam strength, since cushion seams are frequently the limiting feature under high internal pressure. In simulation, fabric is usually represented with a membrane element formulation and a nonlinear orthotropic material model; seams are represented either as tied constraints or as separate elements with reduced strength.

Venting is a design variable in its own right. Cushions may vent through the fabric itself, through dedicated vent holes, or through a combination. The vent area and its opening behavior determine the pressure decay after peak, which is what actually controls occupant loading in the later phase of the event.

Folded Configurations

The fold pattern determines the cushion's initial geometry, its unfolding sequence, and the direction in which it first emerges from the module. Three patterns dominate:

• Flat fold: the cushion is laid flat and folded in alternating layers. It produces a wide, predictable opening and is common in passenger modules.
• Roll fold: the cushion is rolled from the edges inward. It produces a controlled, progressive unfurling and is common in side and curtain bags.
• Z-fold / accordion: the cushion is folded back and forth in a stack. It produces a compact package and a rapid initial expansion, and is widely used in driver modules.

For CAE, the fold pattern must be represented explicitly, because the initial geometry and the contact between folded layers govern the early deployment phase. Two practical approaches exist. The first is to import the actual folded geometry from the CAD or from a folding simulation. The second is to model the cushion in its unfolded state and use an initialization procedure to compress it into the folded shape. The first approach is more faithful but requires clean geometry; the second is more robust but introduces artificial stresses that must be relaxed before the deployment run.

A frequent correlation problem originates here: if the folded geometry in the model does not match the physical fold, the cushion will emerge in the wrong direction or at the wrong time, and no amount of inflator tuning will fix it. The fold should therefore be verified independently, for example by comparing the model's early-time cushion outline against high-speed video of a deployment before any occupant interaction occurs.

What This Means for Deployment Testing and CAE Correlation

The practical consequence of the architecture above is that a deployment test measures the combined response of several subsystems, and a CAE model must reproduce each of them with appropriate fidelity. When a correlation study fails, the cause is usually traceable to one of four inputs: an inflator tank curve measured under the wrong conditions, a fold pattern that does not match the physical part, a fabric permeability or biaxial response that was assumed rather than measured, or a tear seam that was idealized as a simple failure threshold.

A disciplined workflow therefore begins by documenting these four inputs before any simulation is run. Record the inflator part number, lot, and tank test conditions. Record the fold pattern and, where possible, obtain the folded geometry from the physical part. Record the fabric specification including coating weight and measured permeability. Record the tear seam geometry and, if available, its measured opening pressure. With these four items in hand, the analyst has a defensible baseline, and any remaining discrepancy between test and simulation can be attributed to modeling choices rather than to unknown inputs.

With the module architecture, inflator output, cushion material behavior, and fold pattern established as the four governing inputs, the next stage of the course can address how these inputs are exercised in a physical deployment test and how the resulting measurements are prepared for comparison against simulation.