2. Occupant safety fundamentals for DAB

Driver airbag (DAB) design is not judged by whether the bag inflates, but by whether the occupant's measured response stays inside regulated and consumer-rating corridors. Every geometry, vent, tether, and inflator output decision ultimately maps to a small set of physical injury metrics. This section establishes those metrics, how they are computed from measured signals, and why out-of-position (OOP) occupants change the design problem entirely.

From Physical Response to Injury Criteria

Crash test dummies do not report "injury." They report time histories: accelerations, forces, moments, and displacements. Injury criteria are mathematical reductions of those histories into scalars that correlate with real-world injury risk. For DAB work, four families dominate:

• Head: resultant translational acceleration, used to compute HIC; rotational velocity is increasingly tracked for brain-strain-related metrics.
• Neck: axial force (tension/compression), shear force, and bending moment, combined into Nij.
• Chest: sternal deflection, spinal acceleration (3 ms clip), and viscous criterion (VC).
• Kinematics: whether the occupant loads the bag in the intended posture, or interacts with a deploying bag.

The reference frame convention used throughout this course is the SAE J211 sign convention: vehicle-forward is +X, occupant-right is +Y, and downward is +Z, with accelerations positive in the direction of the respective axis. All criteria below assume this convention unless stated otherwise.

HIC: Definition and Practical Computation

The Head Injury Criterion is defined as:

HIC = max over (t1, t2) of [ (t2 - t1) * ( (1/(t2-t1)) * ∫ a_res dt )^2.5 ]

where:
  a_res = sqrt(ax^2 + ay^2 + az^2)   [g, filtered per SAE J211 CFC 1000]
  t2 - t1 <= 15 ms  (HIC15) or 36 ms (HIC36)
  a_res is the resultant translational acceleration at the head CG

Two implementation details cause most disagreements between teams. First, the acceleration must be filtered at the correct channel frequency class before integration; filtering after computing the resultant is not equivalent to filtering each axis first. Second, the search window is bounded — HIC15 is the current regulatory and consumer norm, and a 36 ms window will almost always produce a larger number, so the window must be stated explicitly.

A practical computation on a sampled signal looks like this:

# a_res: resultant head CG acceleration, g, sampled at 10 kHz
# dt = 1e-4 s
best = 0
for i in range(len(a_res)):
    for j in range(i+1, min(i + int(0.015/dt), len(a_res))):
        w = (j - i) * dt
        mean_a = sum(a_res[i:j]) / (j - i)
        hic = w * mean_a**2.5
        best = max(best, hic)
# best is HIC15 for this channel

For a typical 50th-percentile male driver in a 56 km/h frontal rigid-barrier test, a well-tuned DAB produces HIC15 in the range of roughly 200–500. Values above 700 are a warning sign; above 1000 is a regulatory failure in most markets. When HIC is high, the first diagnostic question is when the peak occurs — early peaks (before 30 ms) usually indicate bag-to-head contact while the bag is still stiff, while late peaks indicate bottoming-out or head-to-hardware contact.

Chest Deflection and the Viscous Criterion

Chest injury in DAB events is driven by belt loading and, in some geometries, by bag pressure on the sternum. The primary measured quantity is sternal deflection relative to the spine, measured by a chest deflection potentiometer or IR-TRACC in the dummy. Two criteria are used:

VC captures the rate-dependent softening of chest tissue: a fast, shallow compression can be more injurious than a slow, deep one. In DAB design, chest deflection is most sensitive to bag pressure at the sternum and to the timing of bag-to-chest contact relative to belt lock-up. Reducing inflator mass flow or adding a chest-band tether can lower peak deflection, but often at the cost of head excursion — a trade-off that must be resolved with the full restraint system, not the airbag alone.

Out-of-Position: The Design Case That Governs

An out-of-position occupant is one whose body is closer to the airbag module than the nominal seating position at the moment of deployment. OOP is not an edge case — it is a regulated test matrix (FMVSS 208, ECE R94, and consumer protocols) because a deploying bag can itself cause injury when the occupant is within the bag's expansion path.

The two governing OOP scenarios for DAB are:

• Static deployment (SDO): the occupant is stationary and the bag deploys into them. The relevant metrics are neck force/moment, chest deflection, and head acceleration from bag contact.
• Dynamic deployment: the occupant is moving forward (e.g., from pre-crash braking) as the bag deploys, which changes both the contact timing and the effective closing velocity.

The critical design variables for OOP are inflator output (mass flow rate and temperature), vent area and vent timing, tether configuration, and bag fold pattern. A bag that is gentle enough for OOP is often too soft for the nominal in-position case, which is why dual-stage or adaptive inflators are standard on modern DABs.

How the Criteria Interact in a DAB Load Case

The figure below shows the signal chain from deployment command through occupant response to the criteria that are evaluated, and how OOP changes the path.

Reading the Results: A Diagnostic Sequence

When a DAB load case fails, the criteria themselves tell you where to look. A practical sequence:

1. Check HIC timing. If the peak is before 25 ms, the head is meeting a stiff bag — reduce early mass flow or open vents earlier. If the peak is after 60 ms, suspect bottoming-out or contact with the A-pillar or steering wheel rim.
2. Check neck load direction. High tension with extension moment indicates the head is being pushed back and up by the bag; high compression with flexion indicates the head is being driven down into the bag.
3. Check chest deflection rate. A high VC with modest peak deflection points to a fast pressure rise at the sternum — a vent or tether change is more effective than reducing total inflator output.
4. Check OOP separately. If in-position passes but OOP fails, the fix is almost always in the early deployment phase: lower initial mass flow, delayed vent opening, or a different fold pattern that directs the bag away from the occupant.

The key takeaway is that HIC, Nij, chest deflection, and VC are not independent targets to be optimized one at a time. They are coupled through the bag pressure history, and the OOP cases constrain the early part of that history in ways the in-position case does not. The next section moves from these criteria into the physical DAB architecture that must satisfy them.