4. Regulatory and consumer-test requirements
An airbag CAE model is only as credible as the load cases it is correlated against. Before any mesh is refined or any tank test is scheduled, the engineer must answer a deceptively simple question: which deployment conditions actually matter for this program? The answer comes from a layered set of requirements — federal regulation, regional type-approval, consumer rating programs, and out-of-position (OOP) protocols — each of which imposes different initial conditions, occupant positions, and pass/fail criteria on the inflator and cushion. This chapter maps those requirements to the specific inflation-correlation work they demand.
The Regulatory Baseline: FMVSS 208 and ECE R94/R95
FMVSS 208 (49 CFR 571.208) is the U.S. federal occupant-crash-protection standard and the single most influential driver of airbag inflation behavior. Its key features for correlation work are:
- Rigid-barrier and offset tests at 56 km/h (35 mph) frontal, plus the 40 km/h (25 mph) unbelted rigid-barrier test, which produces the most aggressive occupant kinematics and therefore the highest demand on bag pressure and venting.
- Occupant size matrix: the 50th-percentile adult male (Hybrid III), the 5th-percentile adult female, and child restraints / 6-year-old and 1-year-old surrogates. Each changes the effective mass loading on the cushion.
- Inflation timing windows: the standard specifies that the airbag must be fully inflated before the occupant loads it, which translates into a required time-to-fire (TTF) and time-to-peak-pressure envelope. This is the primary correlation target for the inflator model.
- Unbelted occupants are the worst case for OOP-like loading because the torso travels further before contacting the bag.
ECE R94 (frontal impact) and ECE R95 (lateral impact) are the UNECE type-approval counterparts. R94 uses a 40% offset deformable barrier at 56 km/h with the Hybrid III 50th male, and its airbag requirements are expressed through occupant injury criteria (HIC, chest deflection, femur load) rather than explicit inflation timing. R95 addresses side impact with a moving deformable barrier at 50 km/h; for side airbags, the correlation target shifts from cushion pressure to deployment speed and coverage, because the available crush space is only 200–300 mm.
Consumer Tests: Euro NCAP and Its OOP Siblings
Euro NCAP raises the bar beyond regulation. Its frontal offset test (50 km/h, 50% overlap, deformable barrier) and the mobile progressive deformable barrier (MPDB) test at 50 km/h with 50% overlap are scored on a 0–16 point scale per body region. For airbag correlation, the critical Euro NCAP-specific requirements are:
- Dual-stage and adaptive inflators must be modeled with both stages, because Euro NCAP rewards lower inflation energy for belted occupants (reduced chest loading) while still meeting the unbelted FMVSS 208 case.
- Seat belt pretensioner and load limiter interaction changes the occupant's forward excursion, which in turn changes the bag's effective compression volume.
- Far-side occupant protection (introduced in the 2020+ protocols) requires the bag to manage a 50th-percentile occupant on the opposite side of the impact — a case where the bag is loaded off-center and the venting asymmetry dominates.
Other consumer programs impose their own variants: US NCAP (frontal 56 km/h rigid barrier, side MDB), IIHS (40 km/h moderate overlap, 64 km/h small overlap, and the side 2.0 test with a 1,900 kg MDB at 60 km/h), and China NCAP (C-NCAP frontal 50 km/h 100% overlap and 64 km/h 40% offset). Each changes the barrier stiffness, impact speed, and occupant size, so a single correlated inflator model must be validated across the full matrix rather than a single point.
Out-of-Position (OOP) Protocols: The Hardest Correlation Case
OOP testing simulates an occupant who is too close to the airbag module at deployment — a child leaning forward, an unbelted adult with hands on the wheel, or a rear-facing infant seat in the front passenger seat. Because the bag contacts the occupant during inflation, the loading is dominated by the inflator's early mass flow and the cushion's unfolding kinematics, not by the steady-state pressure. The principal protocols are:
| Protocol | Occupant / Position | Key Correlation Target |
|---|---|---|
| FMVSS 208 OOP (S13–S16) | 1-yr-old, 3-yr-old, 6-yr-old, 5th female; static, no belt | Injury criteria (HIC, neck, chest) during first 30 ms of deployment |
| ECE R94 OOP (Annex 8) | Child seats, adult surrogates, defined proximity | Bag-to-occupant contact force and timing |
| Euro NCAP OOP | Child in CRS, adult in "in-position" and "out-of-position" | Deployment-induced injury metrics, bag pressure rise rate |
| FMVSS 208 Low-Risk Deployment (LRD) | 1-yr-old and 3-yr-old in CRS, rear-facing | Peak pressure and mass flow at 5–15 ms |
The OOP case is where inflation correlation most often breaks down. A model tuned to the 50th-percentile belted occupant at 56 km/h will typically over-predict the bag's early pressure because it does not capture the unfolding resistance of the cushion against a nearby occupant. The fix is not a global pressure scaling factor; it is a re-calibration of the inflator's mass-flow curve and the cushion's fabric permeability in the first 20 ms.
From Requirement to Correlation Target: A Worked Mapping
Consider a driver-side frontal airbag for a vehicle sold in both the U.S. and Europe. The correlation matrix must cover at least four distinct load cases, each with its own target:
- FMVSS 208 unbelted 56 km/h — target: bag fully inflated by 30 ms, peak pressure 220–260 kPa (example range), occupant HIC < 700.
- ECE R94 offset 56 km/h — target: bag pressure at 40 ms within ±10% of sled measurement, chest deflection < 42 mm.
- Euro NCAP MPDB 50 km/h — target: dual-stage timing such that stage 2 fires at 15–20 ms, bag pressure rise rate matched to within 15%.
- FMVSS 208 OOP 1-yr-old — target: peak pressure during first 15 ms below the LRD threshold, mass flow curve matched to tank test within 5%.
The practical workflow is to build a single inflator model (tank-test-calibrated mass flow and temperature) and a single cushion model (fabric permeability, vent size, seam strength), then validate against all four cases. If the OOP case fails while the belted cases pass, the discrepancy is almost always in the early-time mass flow — the first 10 ms — which is exactly the region that tank tests at ambient temperature under-resolve.
What This Means for the CAE Model
The regulatory and consumer landscape is not a backdrop — it is the specification. Every correlation decision (mesh density in the cushion, fabric model, inflator mass-flow table, venting coefficient) should be traceable to a specific requirement. A model that is "correlated" only to a single 56 km/h belted sled test is not correlated; it is curve-fitted. The defensible approach is to define the requirement matrix first, then build the model to satisfy the most demanding case (usually OOP or unbelted FMVSS 208), and verify that the less demanding cases remain within tolerance.
The diagram above summarizes the mapping: regulatory, consumer, and OOP layers each impose distinct targets on the same physical model. The validation loop is not optional — if any case exceeds tolerance, the inflator mass-flow table or cushion permeability must be re-calibrated, and the process repeats until all cases converge. This is the discipline that separates a defensible CAE model from a numerical approximation.