Crash Test Dummies Explained — How Frontal Impact Safety Is Scored

When a car hits a barrier at 64 km/h, the real work happens in the data. Here's how crash test dummies measure injury risk — and what those numbers mean for you.

Crash test dummies look simple, but they're packed with sensors. When a car is crashed at speed, the data from those sensors is converted into a score for each part of the body. That score decides the car's safety rating.

This article focuses on the frontal offset test in particular, where the dummy is scored across four body regions.

Note: The injury criteria below remain at the core of crash testing. Current ANCAP overall ratings combine these results with weighted percentage scores across four protection areas.

Four body regions, four scores

In the frontal offset impact, the dummy's body is divided into four regions, each scored out of four points:

• Head and neck

• Chest

• Knee, femur and pelvis

• Lower leg and foot

In the classic frontal scoring structure, the four scores are added to give a frontal impact total of up to 16 points. Each region is assessed against established injury criteria, so the scoring isn't subjective — it's driven by measured physical loads.

The key measurements

• Head: the HIC36 (Head Injury Criterion). A score below 650 represents a low risk of serious injury; above 1000 means a high risk. Peak head acceleration of 72 g is the "good" benchmark. Hard contact with the structure is assumed if acceleration exceeds 80 g.

• Neck: shear force, tension and extension moment are measured over time. The limits tighten as the crash progresses — for example, neck tension of 2.7 kN at the start falls to 1.1 kN by 60 milliseconds.

• Chest: chest compression of 22 mm or less is excellent, while 50 mm represents a high risk of injury. The viscous criterion (V·C) is also measured, capturing the speed of chest deformation.

• Knee, femur and pelvis: femur compression force of 3.8 kN is the high-performance limit; above about 9 kN is a fracture-level load.

• Lower leg and foot: the tibia index (0.4 is good, 1.3 is high risk) and tibia compression, plus rearward pedal displacement (100 mm is good, 200 mm is high risk). Each criterion traces back to real-world crash data collected over decades.

How scores are calculated

Since 2018, ANCAP (Australasian NCAP) has fully aligned its protocol, testing procedures, and scoring methodology with Euro NCAP. Therefore, its biomechanical data assessment and penalty systems operate in the same way:

• Two Limits (Sliding Scale System): Each measured parameter is bound by a Higher Performance Limit (at or better than which maximum points are awarded) and a Lower Performance Limit (worse than which results in zero points).

• Linear Interpolation: Any measurement falling between the higher and lower limits receives a score calculated through linear interpolation (a sliding scale score).

• Weakest Link Standard (Lowest-Scoring Parameter): For body regions evaluated by multiple parameters (e.g., lower leg assessed by tibia index and tibia compression), the overall region score is determined solely by the lowest-scoring criterion.

Additional Mechanism: Capping Limit. In addition to linear interpolation, both ANCAP and Euro NCAP enforce a Capping limit for critical body areas (head, neck, chest, etc.). If dummy sensor readings exceed extreme safety limits indicating a high risk of fatal or severe injury, the parameter is not only awarded zero points, but a penalty is applied that zeroes out or severely caps the score for that entire impact test.

The modifiers

The dummy data isn't the whole story. ANCAP and Euro NACP also apply modifiers that adjust the score:

• Steering wheel displacement — excessive rearward movement (beyond 100 mm is the EEVC limit) triggers penalties

• A-pillar displacement — more than 200 mm of rearward movement costs up to two points

• Door opening during the impact — every door that opens costs at least one point, because it raises the risk of occupant ejection

• Pedal displacement and footwell intrusion — measured and penalised where relevant

• Airbag stability — unstable contact with the airbag, or the head bottoming out, costs a point

Why the detail matters

All of this engineering precision exists for one reason: to make ratings fair, objective and meaningful. A five-star rating isn't a marketing sticker — it's the output of a scoring system designed to measure real injury risk. Weak links are penalised, and structures that hold up are rewarded.

The next time you see a crash test result, remember what's behind it. It's sensor data from head to toe, measured against injury criteria developed over decades — a system built on years of crash research. Whether it's an electric vehicle or a petrol car, the safety standard is the same. When a car performs well across all four body regions and all the modifiers, that's not luck. It's engineering.