2. Hybrid III architecture

The Hybrid III is not a single monolithic device but a carefully engineered anthropomorphic test device (ATD) assembled from distinct sub-assemblies, each with its own material specification, structural role, and instrumentation interface. Understanding this architecture is a prerequisite for correct sensor installation, credible data interpretation, and meaningful comparison with the THOR dummy. This section walks through the physical build of the 50th-percentile male Hybrid III from the head down, identifying what each part is made of, how adjacent parts articulate, and where transducers are mounted.

Head and Neck Assembly

The head is a cast aluminum skull covered by a vinyl skin. The aluminum skull provides a rigid, repeatable load path, while the vinyl skin supplies the compliant outer surface that contacts the steering wheel or airbag. The skin is a replaceable wear item; its stiffness changes with temperature, which is why the entire dummy must be soaked at 20.6–22.2 °C for at least four hours before a test. The head is instrumented with three linear accelerometers mounted orthogonally at the head center of gravity, plus, in many configurations, three angular rate sensors. The accelerometer block is bolted to the skull at a machined mounting pad, and the exact CG location is documented so that measured accelerations can be transformed to the head CG.

The neck is a molded butyl rubber cylinder with a central aluminum load cell. The rubber provides the bending and shear compliance that mimics a human cervical spine, while the load cell measures the three forces and three moments transmitted between head and torso. The neck attaches to the head via a lower neck bracket and to the upper torso via a neck mounting plate. Critically, the neck is a calibrated, serialized component: its stiffness degrades with use, and it must be replaced or recertified on a defined cycle count. A common installation error is over-torquing the neck mounting bolts, which pre-loads the rubber and shifts the measured bending moment.

Upper Torso, Chest, and Spine

The upper torso is a welded aluminum space frame covered by a urethane chest flesh. The frame carries the shoulder assemblies, the neck mounting plate, and the chest instrumentation. The chest deflection measurement is the defining Hybrid III feature: a chest deflection potentiometer (or, in newer builds, an IR-TRACC-style rotary unit) is mounted on the sternum and connected by a rod to the spine box. As the sternum is pushed rearward during belt loading, the rod slides and the transducer reports deflection in millimeters. The sternum plate is a separate aluminum component bonded to the chest flesh, and its position must be verified against the drawing before each test series.

The spine consists of a segmented aluminum column with rubber elements between segments, allowing controlled flexion. The lumbar spine is a rubber-and-steel assembly that permits forward bending while resisting compression. The thoracic spine box houses the chest deflection transducer's reference end and provides the rigid datum against which deflection is measured. Six-axis load cells are sometimes mounted at the base of the neck and at the lumbar interface to capture spinal load transfer.

Pelvis and Lower Extremities

The pelvis is a cast aluminum structure with a vinyl-covered flesh form. It carries the iliac crest mounting points for the lap belt and houses the pelvis accelerometers (three linear, sometimes three angular). The hip joints are ball-and-socket articulations with adjustable friction; the friction setting is specified in the maintenance manual and affects leg kinematics during knee-bolster and footwell interactions. The femur is a steel tube with a molded flesh covering and a six-axis load cell at the knee joint to measure femoral force and moment. The knee slider is a replaceable aluminum-and-rubber element that controls knee-thigh-hip load transfer. The tibia is similarly a steel tube with a load cell at the ankle and knee, and the foot is a rigid aluminum plate with a vinyl covering and a heel accelerometer.

Articulations and Their Adjustment

The Hybrid III's articulations are not free joints; each has a specified torque or friction setting that must be maintained. The neck is a compliant element rather than a joint. The shoulders are ball joints with adjustable friction. The hips are ball joints with a specified breakaway torque. The knees and ankles are pin joints with adjustable friction and, at the knee, a defined stop angle. These settings matter because they determine how the dummy interacts with the restraint system: a hip joint set too loose allows the pelvis to rotate prematurely, changing lap-belt loading and chest deflection. A practical check is to measure breakaway torque with a calibrated torque wrench at the specified adjustment points and record the values in the pre-test log.

Sensor Mounting Map

The following table summarizes the principal transducer locations and the physical mounting interface for each. All locations refer to the dummy's own coordinate system, with the origin at the pelvis and the positive X axis forward.

Visualizing the Load Path and Instrumentation Chain

The figure below traces the primary structural load path from the head through the neck, chest, spine, and pelvis, and shows where each transducer intercepts that path. It is intended as a mounting reference, not a replacement for the assembly drawing.

From Architecture to Test Readiness

Treating the Hybrid III as an assembly of calibrated sub-systems changes how a test engineer prepares for a run. The practical sequence is: confirm the dummy serial number and build configuration; verify that each transducer is mounted at its specified interface and that its axis orientation matches the SAE J211 convention; measure and record articulation friction and breakaway torques; confirm the temperature soak; and only then proceed to positioning and belt routing. A useful pre-test record captures, for each channel, the transducer type, mounting location, serial number, and last calibration date. When a result looks implausible—for example, a chest deflection that is far below the expected range for the belt load—the first diagnostic step is not to re-run the test but to check the sternum plate position, the deflection rod engagement, and the channel map, because these are the most frequent sources of mounting-related error. The architecture described here is also the reference against which THOR's differences—its improved shoulder and thorax design, additional instrumentation, and different articulation philosophy—become meaningful rather than abstract.