2. Injury assessment fundamentals

Global NCAP crash testing produces a stream of numbers — accelerations, forces, moments, deflections — but a test is only meaningful when those numbers are converted into a defensible statement about human injury risk. This chapter establishes the conceptual and practical bridge between what a crash test dummy measures and what a real occupant experiences. Every later topic in this course, from protocol-year thresholds to star-rating arithmetic, depends on the logic introduced here.

Why Dummies Are Not People, and Why We Still Use Them

A crash test dummy is a mechanical surrogate designed to be repeatable, durable, and instrumented. It is not a biofidelic replica of a human body. Its mass distribution, stiffness, damping, and joint behaviour are calibrated approximations. The value of a dummy lies in correlation: over decades, researchers have run matched dummy tests and human-subject or cadaver tests, and derived statistical relationships between measured dummy responses and observed human injuries. Those relationships are the injury criteria.

This has a critical practical consequence. A dummy measurement is never an injury by itself. A chest deflection of 50 mm is not "a broken rib"; it is a signal that, when passed through a validated risk curve, corresponds to a certain probability of a certain severity of thoracic injury. Understanding this distinction prevents two common errors: treating threshold values as binary pass/fail absolutes, and assuming that a dummy response that "looks fine" guarantees occupant safety.

The Structure of a Biomechanical Injury Criterion

Most injury criteria share a common architecture. They combine one or more measured physical quantities — acceleration, force, moment, deflection, or pressure — through a formula, and compare the result against a threshold derived from injury probability data. Three elements matter:

• The measured quantity and its location. Head acceleration is measured at the head CG; chest deflection at a specific sternal or rib location; femur force along the femoral axis. The criterion is only valid for the measurement point it was derived from.
• The time basis. Some criteria are peak values over the whole event; others are time-windowed (for example, a 3 ms clip) or cumulative (integrating over duration). Mixing time bases invalidates comparison.
• The risk reference. Thresholds are usually tied to a stated injury probability — commonly AIS 2 or AIS 3 — for a defined population. A threshold without its risk basis is ambiguous.

The Abbreviated Injury Scale (AIS) is the reference vocabulary. AIS grades injury severity from 1 (minor) to 6 (currently untreatable), per body region. Injury criteria are typically expressed as the dummy response corresponding to a chosen probability of exceeding a given AIS level. Global NCAP protocols generally target AIS 2 and above for most regions, with head and neck criteria often referenced to AIS 3 or higher due to the consequences of those injuries.

Head and Neck: HIC and Its Relatives

The Head Injury Criterion (HIC) remains the cornerstone. It is computed from the resultant translational acceleration at the head centre of gravity:

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

where a(t) is the resultant head acceleration in g,
and (t2 - t1) is constrained to a maximum window (typically 36 ms for HIC36,
15 ms for HIC15).

In practice, the calculation is performed on filtered acceleration data — the CFC 1000 filter class is standard for head acceleration in most protocols. The window search is exhaustive over the recording; the reported HIC is the maximum value found. A HIC15 of 700 is a widely used reference point associated with roughly a 5% risk of AIS 3+ head injury in the populations from which it was derived; protocol thresholds vary by year and by assessment area, which is why the exact number must always be read from the applicable protocol version.

HIC has known limitations: it ignores rotational acceleration, which is increasingly recognised as a driver of diffuse axonal injury and subdural haematoma. Newer criteria such as BrIC (Brain Injury Criterion) address rotational effects and appear in more recent protocol revisions. When a protocol specifies BrIC, the dummy must be instrumented to measure angular velocity at the head, and the criterion is computed from the resultant angular velocity magnitude.

Neck criteria combine axial force, shear force, and bending moment, usually expressed as Nij (the neck injury criterion) or as separate tension/compression and flexion/extension limits. Nij normalises each force and moment component against its own critical value and sums them, producing a dimensionless number where 1.0 corresponds to the reference injury risk.

Chest, Abdomen, and Pelvis

Thoracic injury is dominated by compression. The chest deflection criterion compares measured sternal deflection against a threshold, often expressed as an absolute value in millimetres or as a percentage of chest depth. The widely cited 63 mm figure for a 50th-percentile male is a reference, not a universal constant; protocols and dummy types (Hybrid III vs THOR) use different values and different measurement locations. Rib deflection rate and the viscous criterion (VC), which combines compression and compression rate, capture the rate-dependent behaviour of soft tissue and are used in some assessment areas.

Abdomen and pelvis criteria typically use force or pressure measurements — for example, abdominal pressure or belt force — with thresholds tied to AIS 2 organ injury. These are more variable across dummy designs and are an area where protocol-year differences are especially pronounced.

Lower Extremities

Femur force, tibia force and bending moment, and knee displacement form the lower-extremity set. The femur criterion is usually an axial compressive force limit; tibia criteria combine axial force and bending moment through an interaction formula. These are relatively well-correlated with real-world fracture risk because long-bone fracture mechanics are comparatively well understood.

From Measurement to Harm: The Translation Chain

The figure below shows how a raw dummy signal becomes an injury-risk statement, and where each step can go wrong.

Worked Example: Computing HIC15 from a Sampled Signal

Consider a simplified teaching example. A head accelerometer records a resultant acceleration pulse, already filtered to CFC 1000, sampled at 10 kHz. Suppose the pulse is approximately triangular: it rises from 0 g at t = 0 ms to 120 g at t = 10 ms, then falls linearly to 0 g at t = 30 ms. This is a constructed example, not a measured result, but it illustrates the mechanics of the calculation.

The HIC integral over a window [t1, t2] is the average acceleration multiplied by the window duration, raised to the power 2.5, then multiplied by the duration. For a triangular pulse, the average over the full 30 ms window is 60 g, giving:

HIC = 0.030 * (60)^2.5
    = 0.030 * 60^2.5
    60^2.5 = 60^2 * sqrt(60) = 3600 * 7.746 ≈ 27,885
HIC ≈ 0.030 * 27,885 ≈ 837

Because HIC15 restricts the window to 15 ms, the maximum over 15 ms windows must be found. Taking the window from 0 to 15 ms, the average acceleration is approximately 80 g (the pulse is above 80 g for part of this window), and the calculation yields a different value. The point of the exercise is procedural: the analyst must (1) confirm the filter class, (2) search all admissible windows, (3) apply the correct exponent and duration constraint, and (4) report the maximum. A common failure symptom is a HIC value that is implausibly low — often caused by applying a 15 ms window when the protocol requires 36 ms, or by using unfiltered data with high-frequency noise inflating the peak.

Reading Results Against Real-World Harm

When a protocol threshold is met, the correct interpretation is that the estimated risk of the referenced injury severity is at or below the level the protocol accepts — not that the occupant is safe. When a threshold is exceeded, the penalty reflects an elevated risk, and the magnitude of the exceedance matters. A HIC15 of 720 against a 700 limit is a marginal exceedance; a HIC15 of 1,200 is a qualitatively different situation. Analysts should always report the criterion value, the threshold, the risk basis, and the protocol year together, because the same number can pass one year's protocol and fail another's.

Body-Region Summary and Practical Checks

The essential discipline is consistency: the same measurement point, the same filter, the same window, and the same risk basis throughout an assessment. Once that discipline is in place, the criterion values become comparable, auditable, and meaningful as statements about real-world harm rather than isolated test artefacts.