2. Restraint design window definition

In side impact, the restraint design window is the bounded set of deployment conditions within which a side airbag, inflatable curtain, or seat-belt pretensioner can actually reduce occupant injury rather than add to it. Unlike frontal restraint design, where the occupant has substantial ride-down distance ahead of them, the side impact case offers only the trim gap between the occupant and the door structure — typically 120–200 mm at the thorax and less at the pelvis. The design window therefore has to be defined with three coupled quantities: a timing threshold (when the restraint must fire), an intrusion tolerance limit (how much structural deformation the restraint can absorb before it is overwhelmed), and an occupant injury criterion (the measurable outcome that defines success). These three are not independent; tightening one shifts the others, and the calibration task is to find the intersection that satisfies all three simultaneously across the crash pulse population the vehicle must cover.

Timing thresholds: from impact detection to full cushion pressure

The timing threshold is best expressed as a budget rather than a single number. For a typical B-pillar-mounted side airbag, the chain runs: impact event begins at t = 0; the side sensor (pressure sensor in the door cavity or accelerometer on the B-pillar) crosses its arming threshold at t_arm; the restraint control module confirms the crash and issues the fire command at t_fire; the inflator output begins at t_squib; and the cushion reaches useful pressure at t_ready. The design window requires t_ready to precede the moment the occupant's thorax would otherwise contact the intruding door trim, t_contact. In practice, engineers work with a margin:

For a representative 50 km/h MDB (moving deformable barrier) side impact per FMVSS 214 / ECE R95, t_contact at the thorax is commonly in the 18–28 ms range depending on door intrusion velocity. A pressure-based side sensor can typically arm within 3–6 ms of barrier contact, and a pyrotechnic inflator reaches 90% of peak pressure in roughly 8–12 ms at 20 °C. That leaves a realistic t_ready of 12–18 ms — tight but achievable. The critical engineering point is that the timing threshold must be validated at the cold end of the inflator's operating range (−35 °C), where output is slower, and at the hot end (+85 °C), where output is faster and over-pressure risk rises. A design window that only closes at 20 °C is not a design window.

Intrusion tolerance limits: how much door movement the restraint can survive

Intrusion tolerance is the maximum inward displacement of the door inner panel, measured at the occupant contact point, at which the restraint still provides a net benefit. It is not a single structural number; it is a property of the restraint–structure pair. Three quantities define it:

• Available gap — the free distance between the undeployed cushion and the occupant's thorax at the moment of firing. If the gap is too small, the deploying cushion can strike the occupant before it is fully inflated, producing "out-of-position" loading.
• Cushion stroke — the additional displacement the inflated cushion can accommodate before bottoming out against the door. A 60 L thorax bag typically provides 80–120 mm of effective stroke at 50 km/h intrusion rates.
• Structural intrusion rate — the door's inward velocity at the contact point, which determines how quickly the available stroke is consumed. A door intruding at 8 m/s consumes 80 mm in 10 ms.

The tolerance limit is reached when intrusion rate × time-to-bottoming exceeds cushion stroke. In a worked example: if the door intrudes at 7 m/s and the cushion provides 100 mm of stroke, bottoming occurs at roughly 14 ms after contact. If the occupant's thorax has already been loaded by the cushion for 10 ms by then, the remaining 4 ms of metal-to-bone contact is what the injury criteria must survive. This is why intrusion tolerance is expressed as a time-resolved limit, not a static displacement: the same 150 mm of intrusion is survivable at 4 m/s and marginal at 9 m/s.

Occupant injury criteria: the measurable success condition

The injury criteria close the window by defining what "restraint worked" means in measurable terms. For side impact, the relevant criteria are drawn from FMVSS 214, ECE R95, and the Euro NCAP side protocols:

These criteria are not interchangeable. A cushion tuned to minimize TTI by softening pressure may increase rib deflection because it bottoms out earlier; a stiffer cushion that controls deflection may raise VC. The design window is the region of (fire time, cushion pressure, vent area) space where all criteria pass simultaneously for the occupant population — 5th percentile female through 95th percentile male, and the OOP (out-of-position) cases required by FMVSS 208 for side restraints.

Putting the three together: a calibration worksheet

The following worksheet is a reproducible starting point for a thorax side airbag calibration on a mid-size sedan. All values are illustrative teaching parameters, not measured results.

Inputs (example):
  Barrier: MDB 950 kg, 50 km/h, 27° crab angle (ECE R95)
  Door intrusion rate at thorax: 6.5 m/s (from CAE)
  Occupant: Hybrid III 50th male, seated per FMVSS 214
  Available gap (trim to thorax): 165 mm
  Inflator: hybrid, 60 L cushion, vent 2 × 25 mm

Timing budget:
  t_arm (pressure sensor)        = 4 ms
  t_confirm (RCM logic)          = 2 ms
  t_squib to 90% pressure @20C   = 10 ms
  t_ready                        = 16 ms
  t_contact (CAE)                = 24 ms
  margin                         = 8 ms  -> PASS (>= 8 ms)

Intrusion tolerance:
  cushion stroke                 = 100 mm
  time to bottoming = 100/6500   = 15.4 ms after contact
  residual metal contact         = 24 + 15.4 - 16 = 23.4 ms? 
    -> recheck: bottoming at t_contact + 15.4 = 39.4 ms
    -> cushion loaded from 16 ms, so 23.4 ms of cushion stroke
    -> acceptable if VC and deflection pass in CAE

Injury check (CAE, example):
  TTI(d)   = 78 g   (limit 85)   PASS
  Rib defl = 38 mm  (limit 42)   PASS
  VC       = 0.82   (limit 1.0)  PASS
  Pelvis F = 3.9 kN (limit 5.0)  PASS

Two failure symptoms are worth recognizing. First, if t_ready exceeds t_contact − 8 ms at −35 °C, the window is open at room temperature but closed in cold weather; the fix is either a faster arming threshold or a dual-stage inflator with a cold-boosted first stage. Second, if rib deflection passes but VC fails, the cushion is deploying too fast into an already-contacting thorax; increasing vent area or delaying the second stage by 3–5 ms usually recovers the window without sacrificing TTI.

Reading the window in practice

The three panels above are meant to be read together, not sequentially. The timing budget on the left tells you whether the cushion can be ready before contact; the intrusion tolerance curve in the center tells you how long the cushion can hold the line once contact occurs; the feasible region on the right is the intersection of all injury criteria in the calibration space. A calibration that passes timing but sits outside the feasible region on the right is a restraint that fires on time but loads the occupant incorrectly — a common outcome when engineers tune fire time alone without re-checking VC and rib deflection.

Two practical rules follow. First, always define the design window at the worst-case occupant and the worst-case thermal state, then verify the nominal case sits comfortably inside. Second, treat the intrusion tolerance limit as a structural input, not a restraint output: if the door intrudes faster than the cushion can stroke, no amount of inflator tuning recovers the window, and the fix belongs in the body structure (B-pillar reinforcement, door beam) rather than the airbag. The design window is therefore a shared contract between the restraint team and the body-in-white team, and the timing thresholds, intrusion limits, and injury criteria are the three clauses of that contract.