📚 Course ContentsChapter 4 Dummy variants and selection

4. Dummy variants and selection

Anthropomorphic test devices (ATDs) are not interchangeable. Each family was designed around a specific regulatory or research question, and its geometry, mass distribution, joint stiffness, and instrumentation channels encode those assumptions. Selecting the wrong ATD does not merely reduce data quality — it can invalidate the entire test relative to the standard or research objective being addressed. This section maps the four ATD families most commonly encountered in occupant-protection testing and provides a defensible selection procedure.

Hybrid III: the regulatory baseline

The Hybrid III 50th percentile male is the reference occupant for FMVSS 208 frontal testing and remains the most widely used ATD worldwide. Its defining characteristics are a segmented steel-and-vinyl-rubber structure, a curved flexible lumbar spine, and a head-neck complex tuned to match human frontal impact response. Instrumentation is well standardized: head accelerometers in a triaxial array, a six-axis upper-neck load cell, chest accelerometers and a chest deflection potentiometer, a six-axis lumbar load cell, and femur and tibia load cells.

Hybrid III variants cover the 5th percentile female and the 95th percentile male, plus the CRABI and Hybrid III child family (3-, 6-, 10-year-old and smaller). The critical practical point is that Hybrid III is a frontal ATD. Its neck and lumbar response, and the absence of shoulder or pelvis biofidelity in lateral loading, make it unsuitable for side impact. Using a Hybrid III in a pole side test produces numbers, but those numbers have no established injury-risk correlation.

THOR: advanced frontal biofidelity

THOR (Test Device for Human Occupant Restraint) addresses known Hybrid III limitations: a more human-like face and mandible, a four-rib chest with improved deflection measurement, a multi-segment lumbar spine, and — in the THOR-50M configuration — an articulated pelvis and improved knee-thigh-hip load paths. THOR is instrumented far more densely than Hybrid III, typically including chest-band potentiometers, an abdominal pressure sensor, and additional spinal load cells.

THOR is a research and regulatory-development ATD. It is not yet the compliance device for FMVSS 208, so a THOR test cannot be substituted for a certification test. Its value lies in restraint-system optimization, where the richer channel set reveals load paths that Hybrid III averages away. The cost is setup complexity: THOR requires careful umbilical routing, more data channels (often 60+), and stricter seating procedures because its articulated pelvis is sensitive to initial position.

WorldSID: the lateral reference

WorldSID 50th percentile male is the side-impact ATD specified in UN R95, FMVSS 214 (as an option alongside SID-IIs), and most Euro NCAP side protocols. Its design is genuinely bilateral — the same ATD can be seated on either the struck or non-struck side — with a ribcage of curved ribs instrumented for deflection, a six-axis pelvis load cell, and shoulder and thorax load cells. The 5th percentile female WorldSID is increasingly required in side protocols because small occupants sit closer to the door trim.

WorldSID's arm position is a frequent source of test-to-test variation. The standard specifies the arm down against the door for the struck-side occupant; deviating from this changes thorax loading substantially. Always confirm the arm position against the applicable protocol before running.

Child dummies: size, mass, and restraint interface

Child ATDs are grouped by nominal age, but the selection driver is the restraint system, not the age label. The ECE R44/UN R129 families (Q-series and P-series) and the US Hybrid III child family differ in geometry and instrumentation. Key selection rules:

Rear-facing CRS: use the ATD specified by the CRS manufacturer and the protocol; Q-series dummies are standard in UN R129, Hybrid III child dummies in FMVSS 213.
Booster seats: the ATD must match the booster's certified mass group; a 10-year-old Hybrid III in a booster designed for a 6-year-old produces non-representative belt geometry.
Instrumentation: child ATDs carry head accelerometers and chest deflection; some carry abdominal load cells. Confirm channel availability before writing the data-acquisition plan.

A common error is selecting an ATD by age when the protocol specifies by mass group. A 22 kg child may fall into either Group II or Group III depending on the standard; the ATD, not the child's age, must match the protocol's group definition.

Selection criteria in practice

Selection is driven by four questions, applied in order: (1) What standard or protocol governs the test? The protocol names the ATD and variant — this is non-negotiable for compliance. (2) What is the impact direction? Frontal → Hybrid III or THOR; lateral → WorldSID or SID-IIs; rear → Hybrid III with rear-impact neck configuration. (3) What occupant size is being evaluated? Match percentile to the design occupant or the worst-case occupant for the restraint. (4) What instrumentation does the research question require? If the question concerns chest-band load distribution, THOR or WorldSID is required; Hybrid III cannot answer it.

The figure below summarizes the decision path and the primary instrumentation each family provides.

Worked selection example

Consider a program evaluating a new side airbag for a compact SUV under Euro NCAP side mobile deformable barrier (MDB) and pole tests. The protocol fixes WorldSID 50M for the MDB test and WorldSID 50M for the pole test; the 5th percentile female WorldSID is additionally required in the far-side pole configuration. The team also wants to understand rib-fracture risk for a 95th percentile occupant, which is outside the protocol. Here the correct move is to run the compliance tests with the mandated WorldSID variants and add a separate research test with a WorldSID 95M (where available) or a THOR-based lateral configuration, clearly labeled as non-compliance data. Mixing the two data sets in a single report without labeling is a common and serious error.

For a frontal restraint optimization study, the reverse applies: run the FMVSS 208 certification matrix with Hybrid III 50M and 5F, then run a parallel THOR-50M matrix to resolve chest-band load distribution. The THOR results inform design; the Hybrid III results certify it. Both are needed, and neither substitutes for the other.

Never mix ATD families within a single comparative data set without explicit labeling. Injury criteria (HIC, chest deflection limits, Nij) are ATD-specific; a HIC of 700 from a Hybrid III and a HIC of 700 from a THOR are not equivalent, and a table that places them side by side without annotation will be misread.

Setup implications of ATD choice

ATD selection cascades into the instrumentation and positioning plan. Hybrid III requires a relatively simple umbilical and a well-defined H-point procedure. THOR requires additional channels, a pressure-sensor abdominal band, and a more careful pelvis seating check. WorldSID requires bilateral arm-position verification and, for the struck-side occupant, confirmation that the arm is down against the door trim per protocol. Child ATDs require CRS-specific installation checks and, for rear-facing seats, support-leg or ISOFIX confirmation before the ATD is placed.

The practical rule: select the ATD from the protocol first, then read the ATD-specific seating procedure, then build the channel list. Reversing this order — building the channel list first and fitting an ATD to it — is the most common cause of a test that runs cleanly but answers the wrong question.