2. Tether architecture and material selection

An airbag tether is not a single component but a small structural system: woven webbing, stitched terminations, reinforced anchor points, and the load path that connects them. In CAE validation, the tether is often the most highly loaded and least forgiving element in the restraint assembly. A tether that survives a quasi-static pull test can still fail during deployment because the loading is dynamic, the geometry is folded, and the load path passes through stitch lines that were never intended to carry the full deployment force. This section covers how tether architecture is defined, how webbing and stitching are selected, and how the load path is designed so that CAE models represent the real structure rather than an idealized one.

Webbing Types and Selection Criteria

Tether webbing is typically woven from nylon 6.6 or polyester multifilament yarns. Nylon 6.6 offers higher elongation and energy absorption, which is useful when the tether must absorb deployment energy without transferring a sharp peak load to the anchor. Polyester provides lower creep and better dimensional stability under sustained load, which matters for tethers that remain under tension after deployment. The choice is not simply "nylon for energy, polyester for stiffness"; it depends on the deployment temperature, the expected strain rate, and whether the tether is a primary load-bearing member or a positioning element.

Common webbing constructions include flat woven webbing, tubular webbing, and reinforced edge webbing. Flat webbing is the baseline for most tethers because it is thin, flexible, and easy to stitch. Tubular webbing distributes load more uniformly across the cross-section and resists edge abrasion, but it is thicker and can create packaging problems in tight fold patterns. Reinforced edge webbing adds a selvage or a woven-in cord at the edges to resist tear propagation; this is valuable when the tether passes over a sharp bracket or a seam. The selection should be driven by the load case, not by availability: a tether that carries 8 kN in a 20 ms deployment event needs a different construction than one that carries 2 kN in a slow positioning function.

Key parameters to record for CAE input include the webbing width, thickness, linear density, breaking strength, elongation at break, and the weave orientation relative to the load axis. A 50 mm wide nylon 6.6 webbing with a breaking strength of approximately 22 kN and an elongation at break of 18–22% is a representative example for a passenger-side tether. These values are illustrative and must be replaced with the actual supplier datasheet for the specific part number.

Stitching Patterns and Joint Efficiency

Stitching is where most tether failures originate. A stitched joint is not a homogeneous connection; it is a series of load transfers between the webbing plies through the thread. The joint efficiency—the ratio of joint strength to webbing strength—depends on the stitch pattern, thread type, stitch density, and the number of rows.

The most common patterns are the box-X stitch, the parallel row stitch, and the zigzag or W-pattern. The box-X pattern is widely used for tether terminations because it confines the stitch zone and prevents the webbing from peeling apart at the edges. Parallel rows are simpler to sew but can unzip progressively if the first row fails. The zigzag pattern distributes shear across a wider area and is often used where the tether must remain flexible after stitching.

Thread selection follows the webbing: bonded nylon thread (size 69 or 92) is typical for nylon webbing, while bonded polyester thread is used with polyester webbing to avoid differential shrinkage. Stitch density is usually specified as stitches per inch (SPI); 6–8 SPI is common for structural tether joints, with 8–10 SPI for high-load terminations. The number of rows is determined by the required joint strength: a single row of box-X stitching may achieve 60–70% joint efficiency, while a three-row configuration can reach 85–90% in a controlled pull test.

Anchor Points and Load Introduction

Anchor points are the interfaces between the tether and the airbag module, the vehicle structure, or the cushion. Each anchor introduces a stress concentration and a potential failure mode. Common anchor types include sewn loops around a metal bar, stitched tabs bolted to a bracket, and direct stitching to the cushion fabric.

The design rule is to match the anchor strength to the tether strength and to avoid abrupt stiffness changes. A sewn loop around a bar should have a radius large enough to prevent webbing cutting; a minimum bar diameter of 8–10 mm is typical for 50 mm webbing. Bolted tabs require reinforcement washers and a stitch pattern that spreads the bolt load into the webbing rather than concentrating it at the hole. Direct stitching to the cushion is the most weight-efficient but the least tolerant of misalignment; the stitch line must be oriented so that the load is carried in shear, not peel.

For CAE, each anchor should be represented with its actual geometry and boundary condition. A fixed constraint at the anchor node is rarely correct; a distributed coupling or a contact definition that allows the webbing to rotate around the bar is closer to reality. The anchor stiffness affects the load path and the peak force seen by the tether, so it should be calibrated against a component test if available.

Load-Path Design and CAE Representation

The load path is the sequence of elements that carry force from the airbag cushion to the vehicle structure. A well-designed load path is continuous, has no sharp changes in stiffness, and avoids loading stitches in peel or cleavage. In practice, this means routing the tether so that the primary tension is aligned with the webbing weave, the stitch lines are loaded in shear, and the anchor is loaded in a direction that the bracket was designed to carry.

Figure 1 shows a representative tether load path with the key elements and the CAE representation for each. The diagram is a signal-chain style view: the cushion pressure generates tension, the tether carries it, the stitch joint transfers it, and the anchor reacts it into the structure. Each transition is a potential weak point and a place where the CAE model needs explicit representation.

In the CAE model, the load path should be checked in three stages. First, a quasi-static pull of the tether assembly to verify the joint efficiency and anchor stiffness. Second, a dynamic deployment simulation with the tether represented as a membrane or seatbelt element with rate-dependent material data. Third, a post-deployment residual strength check if the tether remains loaded. The first stage calibrates the model; the second predicts the deployment response; the third confirms that the tether has not been degraded by the event.

Practical Example: Tether Joint Calibration

Consider a 50 mm nylon 6.6 tether with a box-X stitch termination, three rows, 8 SPI, bonded nylon thread. The webbing breaking strength is 22 kN. A quasi-static pull test at 100 mm/min gives a joint failure load of 18.5 kN, which is 84% joint efficiency. The failure mode is thread breakage in the first row, followed by progressive unzipping.

For the CAE model, the stitch joint is represented as a connector with a force-displacement curve calibrated to the test: initial stiffness 4.5 kN/mm, peak force 18.5 kN at 4.1 mm displacement, then a softening branch to 6 kN at 8 mm as the stitches unzip. The anchor is represented as a distributed coupling to a rigid bar with a rotational stiffness of 0.8 N·m/deg, based on a separate bracket test. The webbing is modeled with a rate-dependent material: at 0.01 s⁻¹ strain rate, the modulus is 1.2 GPa; at 10 s⁻¹, it rises to 1.8 GPa. These values are illustrative and must be replaced with actual test data for the specific webbing and thread combination.

When the calibrated model is run in a deployment simulation, the peak tether force is predicted at 14.2 kN, which is below the 18.5 kN joint capacity but above the 12 kN anchor bracket yield limit. The model therefore predicts anchor bracket yielding as the first failure, not stitch failure. This changes the design action: the bracket needs reinforcement, not the stitch pattern. Without the calibrated joint and anchor models, the simulation would have predicted a different failure mode and led to the wrong corrective action.

Design Checklist for Tether Architecture

The tether architecture is only as good as its weakest interface, and the CAE model is only as good as its representation of that interface. By selecting webbing and stitching from the load case, designing the load path to avoid peel and stress concentration, and calibrating each interface separately, the simulation becomes a predictive tool rather than a confirmation exercise. The next step is to translate this architecture into a validated model that can be used for deployment timing and occupant restraint studies.