4. Cushion geometry and pattern design

An airbag cushion is a pressure vessel made of coated fabric, and its performance is decided long before the first explicit dynamic run. The geometry that governs deployment kinematics, occupant contact pressure, and seam loading is produced by a chain of transformations: a 3D inflated shape is defined, that shape is cut into panels, the panels are flattened into 2D patterns, and the patterns are constrained by what the cutting table, sewing machine, and coating process can actually produce. This chapter walks that chain in the order a designer actually works, with the parameters and checks that matter at each stage.

4.1 Starting from the Inflated Shape, Not the Flat Pattern

Design intent should be expressed on the deployed (inflated) surface. For a driver-side cushion, the target is typically a flattened torus-like bag with a face diameter of roughly 650–720 mm and a depth of 250–320 mm at 1.5–2.0 bar internal pressure. For a side curtain, the target is a long, thin "sausage" or multi-chamber cushion spanning 1600–2200 mm along the roof rail. These inflated dimensions, not the flat blank, are what the occupant interacts with, so they are the correct starting point.

A practical workflow is to build the inflated surface as a parametric shell in CAD or as a coarse membrane FE model, then iterate the surface until the key dimensions and curvature match the target. At this stage, two quantities should be recorded because they drive everything downstream:

• Developable-ness. A surface that can be flattened without stretching is developable (zero Gaussian curvature). A sphere is not; a cylinder is. Real cushions are mostly non-developable, so flattening always introduces some distortion, and the designer must decide where to absorb it.
• Seam lines. Seams are not decoration; they are the boundaries between panels and the locations of stress concentration. Their 3D path determines the 2D pattern outline.

4.2 Panel Segmentation: How Many Pieces, and Where to Cut

Panel segmentation is the decision of how many fabric pieces form the cushion and where the seams run. It is driven by three competing pressures: structural efficiency (fewer seams = fewer leak paths and lower cost), manufacturability (flat panels must fit the cutting table and the sewing fixture), and deployment behavior (seam placement controls how the bag unfolds).

Common segmentations for a driver-side cushion are:

For a side curtain, segmentation is usually along the length: a series of chambers separated by internal seams, each chamber fed by a common inflator manifold. The chamber pitch (typically 80–150 mm) and the seam width (typically 8–15 mm finished) are the dominant geometric parameters.

4.3 2D Pattern Flattening: The Core Operation

Flattening converts the 3D panel surface into a 2D outline that can be cut from roll goods. The standard approach is a mesh-based flattening: the 3D panel is discretized into a triangle mesh, and the mesh is unfolded into the plane while minimizing a distortion energy. The result is a 2D mesh whose boundary is the pattern outline.

Two distortion metrics should be monitored during flattening:

• Length distortion along the seam direction. If a 3D seam edge of 200 mm becomes 208 mm in 2D, the sewn seam will be 4% too long and will pucker or wrinkle when the bag inflates.
• Area distortion across the panel. Large area distortion means the fabric is being asked to stretch or compress during forming, which changes local permeability and seam strength.

A practical acceptance criterion for a driver-side cushion panel is length distortion below 2% and area distortion below 3% in the central region, with slightly higher values permitted near the panel corners where the fabric can conform. If these limits are exceeded, the usual fix is to add a seam (increase panel count) or to adjust the 3D surface to be more developable.

4.4 Seam Layout and Allowance

Once the 2D outlines are known, the seam layout is defined. A seam has three geometric attributes: its path (where it runs), its allowance (how much fabric is consumed), and its type (sewn, welded, or bonded).

For a sewn seam, the allowance is the sum of the two panel edges plus the seam width. A typical lap seam on coated nylon uses a 10 mm seam width and a 12 mm allowance per panel, giving a finished seam of about 10 mm with 2 mm of trim. The pattern must include this allowance; forgetting it is one of the most common causes of a bag that is too small after sewing.

Seam path design also controls deployment. A seam that runs across the face of the bag will create a hinge; a seam that runs radially will create a petal. For a driver-side bag, the front panel is often sewn to the back panel along a circular path, and the resulting seam acts as the hinge that determines how the bag unfolds toward the occupant.

4.5 Manufacturing Constraints That Feed Back into Geometry

The pattern is not final until it passes the manufacturing constraints. The most important are:

• Roll width. Coated fabric is supplied in rolls of 1500–1800 mm width. A panel wider than the roll requires a seam, which changes the segmentation.
• Cutting table size. The cutting table limits the maximum panel dimension; large side curtains are often cut in two pieces and joined.
• Minimum seam width. The sewing machine and the coating thickness set a minimum finished seam width, typically 8–12 mm. Narrower seams risk stitch pull-out under pressure.
• Corner radius. Sharp corners concentrate stress and are difficult to sew. A minimum corner radius of 5–10 mm is common.
• Grain direction. The fabric's warp and weft directions have different stiffness and permeability. The pattern should be oriented so that the principal stress direction aligns with the stronger yarn direction.

4.6 A Worked Example: Driver-Side Cushion Panel

Consider a driver-side cushion with a target inflated face diameter of 680 mm and a depth of 280 mm. The designer chooses a three-panel segmentation: a circular front panel, a circular back panel, and a side gusset. The 3D surface is built as a revolved profile, and the gusset is the strip between the two circular panels.

Flattening the front panel gives a 2D outline that is approximately circular with a diameter of 690 mm (a 1.5% length distortion, acceptable). The gusset flattens to a strip 280 mm wide and 2140 mm long (the circumference of the 680 mm circle). The seam allowance is 12 mm per edge, so the cut gusset is 304 mm wide and 2164 mm long. The roll width is 1600 mm, so the gusset must be cut in two pieces and joined, adding one seam.

The resulting pattern is then checked against the manufacturing constraints: the front panel fits the cutting table, the gusset pieces fit the roll width, the seam width is 10 mm finished, and the corner radii are above 8 mm. The grain line is aligned with the circumferential direction of the gusset, which is the principal stress direction during deployment.

4.7 Verification Before Release

Before a pattern is released to manufacturing, it should pass a short verification sequence. First, re-measure the 2D seam lengths against the 3D seam lengths and confirm the distortion limits. Second, lay the 2D panels out on a virtual roll of the actual width and confirm that every panel fits without an unplanned seam. Third, check the grain line against the principal stress direction from a coarse inflation simulation. Fourth, confirm that all corner radii and seam widths meet the manufacturing minimums. Only then should the pattern be frozen and sent to the cutting table.

The geometry and pattern design stage is where most airbag performance is locked in. A cushion that is correctly segmented, accurately flattened, and constrained by real manufacturing limits will deploy predictably and pass validation with fewer iterations. A cushion that is not will consume simulation and test cycles trying to correct problems that originate in the pattern itself.