4. Airbag fabric and structural mechanics

An airbag is not a rigid pressure vessel. It is a sewn, woven, coated textile structure that inflates in tens of milliseconds, and its mechanical response is dominated by the fabric's membrane behavior, its permeability to gas, and the seams that hold the panels together. In deployment simulation, the fabric model determines how the bag pressurizes, how it unfolds, and how it interacts with the occupant. Getting the material model wrong produces a bag that either inflates too stiffly or leaks unrealistically, and the resulting occupant kinematics will be meaningless regardless of how well the gas dynamics are solved.

Shell Kinematics and the Membrane Assumption

Airbag fabric panels are thin — typically 0.2 to 0.4 mm thick — and carry load almost entirely in their own plane. This justifies the membrane (shell) assumption: bending stiffness is negligible, and the internal stress state is a two-dimensional tensor in the tangent plane of the surface. In an explicit finite element solver, the bag is discretized with membrane or thin-shell elements, and the kinematics are described by the deformation gradient F mapping the reference (flat, cut-pattern) configuration to the current (inflated) configuration.

The strain measure most commonly used is the Green–Lagrange strain:

E = 0.5 * (F^T · F − I)

where I is the identity tensor. For a woven fabric, the principal material directions are the warp (machine direction, 0°) and weft (cross direction, 90°). These directions are tracked through the deformation, and the constitutive response is evaluated in this local frame. A critical practical point: the cut pattern defines the reference orientation, so the warp/weft axes must be assigned consistently across every panel. If a panel is rotated 90° during meshing, the anisotropic stiffness will be applied in the wrong direction and the bag will inflate asymmetrically.

Membrane Stresses and Fabric Constitutive Behavior

Woven nylon 6.6 (the dominant airbag fabric) exhibits strongly nonlinear, anisotropic, and direction-dependent behavior. Under low strain, the weave crimp straightens and the fabric is compliant; as yarns straighten and lock, stiffness rises sharply. A common engineering idealization is a piecewise-linear orthotropic membrane:

These values are illustrative starting points for a simulation study, not measured data for a specific fabric. Real characterization requires biaxial tensile testing, because uniaxial tests on woven fabric give misleading moduli — the crimp interchange between warp and weft means the fabric's response depends on the ratio of biaxial loads. A biaxial test rig applies independent warp and weft tensions and records strain, producing the data needed to calibrate an orthotropic or fabric-specific material model.

For the membrane stress resultants (force per unit width), the constitutive relation in the local warp/weft frame is:

N_warp = t · (C11 · E_warp + C12 · E_weft)
N_weft = t · (C12 · E_warp + C22 · E_weft)

where t is the fabric thickness and Cij are the orthotropic stiffness components derived from the moduli and Poisson's ratios. The solver integrates these resultants over the element area to obtain nodal forces.

Fabric Permeability and Gas Leakage

No airbag fabric is perfectly gas-tight. Gas escapes through the weave interstices (porosity) and, for coated fabrics, through pinholes and seam perforations. Permeability is the single most influential parameter for the inflation pressure history. A bag that is too permeable vents too quickly and may not develop enough pressure to protect the occupant; a bag that is too tight over-pressurizes and can cause injury.

Permeability is typically characterized by a mass flow rate per unit area as a function of the pressure differential across the fabric:

m_dot = A · K · (Δp)^n

where m_dot is the mass flow rate, A is the fabric area, Δp is the pressure difference, and K and n are empirically fitted constants. For many airbag fabrics, n is close to 1 (linear Darcy-like flow) at low pressure differentials but rises toward 1.5–2 at higher differentials as inertial effects become significant. In simulation, this is implemented as a distributed leakage term applied to each element, with Δp evaluated from the local bag pressure minus ambient.

Seam Behavior

Seams are the weakest and most complex part of the airbag structure. They are created by stitching, welding, or adhesive bonding, and they introduce three effects that must be represented in simulation:

• Stiffness discontinuity: A seam is locally stiffer than the surrounding fabric because the stitch line constrains relative motion of the panels. This creates stress concentrations at the seam ends.
• Strength reduction: The seam efficiency — the ratio of seam strength to parent fabric strength — is typically 60–90% for sewn seams. The stitch holes also act as stress raisers.
• Leakage path: Gas can escape through the stitch perforations, adding a localized permeability term along the seam line.

In a shell-element model, seams are usually represented by one of three approaches: (1) shared nodes along the seam line with a local thickness or stiffness multiplier; (2) tied contact between overlapping panel edges; or (3) dedicated seam elements with their own constitutive law. The choice depends on whether the seam is expected to carry significant load or merely to close the bag geometry.

Material Models for Deployment Simulation

Three classes of material models are commonly used, in increasing order of fidelity:

1. Linear orthotropic elastic: Fastest, adequate for initial geometry checks and occupant-position studies where the bag is assumed to inflate fully and uniformly. Cannot capture fabric lock-up or failure.
2. Piecewise-linear orthotropic with failure: Captures the crimp-to-locked transition and allows element deletion when failure strain is exceeded. This is the workhorse for production deployment simulation.
3. Fabric-specific continuum models: Account for yarn reorientation, crimp interchange, and rate dependence. Used in research and in applications where the inflation transient is critical (e.g., out-of-position occupant loading).

The following SVG illustrates the load path from gas pressure through the fabric membrane to the seams and the resulting deformation and leakage mechanisms.

Calibration Workflow and Practical Checks

A defensible fabric model requires more than picking moduli from a table. The following workflow is typical for a deployment simulation study:

1. Characterize the fabric: Perform biaxial tensile tests at load ratios representative of inflation (e.g., 1:1, 2:1, 1:2 warp:weft). Fit the orthotropic or piecewise-linear model to the measured stress–strain curves.
2. Measure permeability: Use a standardized permeability test (e.g., ASTM D737 or a dedicated airbag fabric test) at pressure differentials spanning the expected inflation range. Fit K and n.
3. Characterize seams: Test seam specimens in tension to determine seam efficiency and failure mode. If seam leakage is significant, measure it separately.
4. Validate on a sub-scale problem: Simulate a simple inflation of a flat circular or rectangular bag and compare the pressure–time history and inflation shape against a physical test or a known analytical solution.
5. Check mesh convergence: Refine the membrane mesh until the peak pressure and inflation time converge. Seam regions often need local refinement because of stress concentrations.

When reviewing results, the key observables are the internal pressure history, the bag volume history, the membrane stress resultants at critical locations (seam ends, panel corners), and the leakage mass flow rate. If the bag inflates but the pressure never reaches the expected plateau, permeability is likely too high or the gas inflow is too low. If the bag over-pressurizes and seams fail early, the seam strength or the fabric failure strain is likely underestimated. These symptoms, mapped back to the material parameters, are the practical feedback loop that makes the fabric model trustworthy.