3. Regulatory and rating frameworks

Before a single Excel formula can decide pass or fail, the evaluation logic must be anchored to a specific regulatory or rating framework. Each framework defines its own set of injury criteria, its own measurement channels, and its own thresholds — and these definitions are not interchangeable. A chest deflection of 50 mm may be a comfortable pass under one regime and a marginal failure under another, not because the occupant is different but because the criterion definition, the scaling, and the limit value differ. This chapter maps the five frameworks most commonly encountered in body-structure and restraint engineering, focusing on what each one measures, how it defines the criterion, and where the pass/fail thresholds sit.

FMVSS 208 — Frontal Occupant Protection

FMVSS 208 is the U.S. frontal crash standard. Its injury assessment is built around the Hybrid III 50th-percentile male dummy in the driver and right-front passenger positions, evaluated in a 56 km/h (35 mph) rigid barrier test, plus optional unbelted and offset conditions. The regulation defines a set of injury criteria (IC) that are computed from dummy channels and compared against limits. The core set is:

• HIC15 — Head Injury Criterion over a 15 ms window, limit 700 for the 50th male.
• Chest deflection — peak sternal deflection, limit 63 mm (driver) / 63 mm (passenger) for the 50th male.
• Chest acceleration — 3 ms clip, limit 60 g.
• Femur axial force — limit 10 kN per leg.
• Neck — Nij (combined tension/compression and bending moment), limit 1.0.

The critical implementation detail is that FMVSS 208 uses HIC15, not HIC36. The window length changes the computed value materially: a short, sharp pulse can produce a high HIC15 while the HIC36 stays low. In Excel, this means the HIC formula must be parameterized by window length, and the framework selector must drive that parameter. A common error is to hard-code a 36 ms window and then apply it to a 208 dataset — the resulting HIC will be systematically understated.

FMVSS 214 — Side Impact Protection

FMVSS 214 governs side impact. The primary test is the moving deformable barrier (MDB) at 54 km/h (33.5 mph) striking the vehicle at 27° (crabbed barrier), with the ES-2re dummy in the driver position and the SID-IIs in the rear. The injury criteria are:

• Head — HIC36, limit 1000.
• Chest — rib deflection (upper, middle, lower), with a limit of 44 mm for the ES-2re; the SID-IIs uses a different deflection limit (typically 41 mm) and a viscous criterion (VC) limit of 1.0 m/s.
• Pelvis — pubic symphysis force, limit 6 kN (ES-2re).
• Abdomen — force limit 2.5 kN (ES-2re).

Note the framework switch: 214 uses HIC36, not HIC15. The dummy type also changes the deflection limit and introduces VC as a separate criterion. In an automated pass/fail sheet, the dummy type must be a lookup key that selects both the limit set and the criterion definitions — the ES-2re and SID-IIs rows cannot share a single deflection threshold.

ECE R94 and R95 — European Frontal and Side

ECE R94 (frontal) and R95 (side) are UNECE regulations. R94 uses a 56 km/h 40% offset deformable barrier (ODB) test with the Hybrid III 50th male, and its criteria closely mirror FMVSS 208 but with important differences: HIC36 is used (limit 1000), chest deflection limit is 50 mm (not 63 mm), and the chest VC limit is 1.0 m/s. R95 uses the AE-MDB barrier at 50 km/h with the ES-2 dummy, with rib deflection limits around 42 mm and a VC limit of 1.0 m/s.

The practical consequence is that a vehicle validated to FMVSS 208 at 63 mm chest deflection will not automatically pass R94 at 50 mm. The Excel model must therefore treat the chest deflection limit as a framework-dependent parameter, not a constant. The same physical channel (chest deflection in mm) feeds two different limit cells depending on which regulation row is active.

Euro NCAP and C-NCAP — Rating Protocols

Euro NCAP and C-NCAP are consumer rating programmes, not type-approval regulations, but they are frequently the binding constraint in development because their thresholds are stricter and their scoring is continuous rather than binary. Euro NCAP frontal (Mobile Progressive Deformable Barrier, MPDB, and full-width) uses HIC15 with a limit of 650 for a 5-star score, chest deflection limits around 42–50 mm depending on dummy and position, and a femoral force limit of 6 kN. C-NCAP (2021 revision) uses a 50 km/h MPDB and a 64 km/h ODB, with HIC15 limit 650, chest deflection limit 50 mm, and a femoral force limit of 6 kN, plus a separate penalty structure for restraint system reminders.

Because rating protocols award points on a sliding scale, the pass/fail automation for these frameworks is usually implemented as a threshold band rather than a single limit: a value below the "full score" threshold earns maximum points, a value between full-score and zero-score thresholds earns a linear interpolation, and a value above the zero-score threshold earns zero. The Excel model should therefore store three numbers per criterion — full-score limit, zero-score limit, and the interpolation rule — rather than a single pass/fail value.

Where the Frameworks Diverge — and Why It Matters for Automation

The table below summarizes the criterion-level differences that most often cause errors when a single evaluation sheet is reused across frameworks.

Implementing Framework Selection in the Evaluation Sheet

A robust structure separates three layers: a framework definition table (one row per framework × criterion, holding the limit, the window length, and the scoring rule), a measurement table (raw channel data per test), and an evaluation table that joins the two on framework ID and criterion ID. The join key is what prevents cross-framework contamination. In Excel, this is naturally a lookup:

=INDEX(FrameworkLimits[Limit],
       MATCH(1,
         (FrameworkLimits[FrameworkID]=$B$2)*
         (FrameworkLimits[CriterionID]=$A6),0))

With the framework ID in $B$2 and the criterion ID in column A, the limit is pulled dynamically. The HIC window length is stored in the same table and passed to the HIC calculation, so a single workbook can evaluate the same raw pulse under 208 (15 ms) and R94 (36 ms) without editing formulas. The pass/fail cell then reduces to a comparison against the pulled limit, with the scoring rule (binary or interpolated) selected by a second lookup on the framework's scoring type.

The SVG below shows how the three layers connect and where the framework-dependent parameters enter the evaluation.

Practical Verification of a Framework Switch

To confirm the model behaves correctly, run a controlled check with a known input. Take a chest deflection channel of 48.2 mm and evaluate it under three frameworks in sequence. Under ECE R94 (limit 50 mm) the verdict should be PASS; under FMVSS 208 (limit 63 mm) PASS; under FMVSS 214 with the ES-2re dummy (limit 44 mm) FAIL. If all three return the same verdict, the framework lookup is not actually driving the limit — most likely the limit cell is hard-coded or the MATCH is returning the first row regardless of framework ID. A second check on HIC: feed a pulse that yields HIC15 = 720 and HIC36 = 640. Under FMVSS 208 (HIC15, limit 700) the verdict must be FAIL; under ECE R94 (HIC36, limit 1000) it must be PASS. If both return PASS, the window-length parameter is not reaching the HIC calculation.

The frameworks differ in window length, limit value, dummy-specific thresholds, and scoring philosophy. None of these differences can be safely collapsed into a single "typical" limit. The evaluation sheet must carry the framework as a first-class input, resolve every limit through a lookup keyed on framework and criterion, and record the resolved limit in the output. That structure is what makes the pass/fail automation trustworthy across the full set of regulations and rating protocols a modern vehicle programme must satisfy.