4. Human body models for pedestrian impact

Pedestrian protection simulation in PAM-CRASH relies on three distinct impactor and human body model families, each tied to a specific regulatory or consumer-test protocol. Selecting the wrong model, or positioning a correct model incorrectly, invalidates the entire load case before the solver even starts. This section covers what each model represents, when to use it, and how to position it precisely in the PAM-CRASH pre-processor.

Model Families and Their Regulatory Anchors

The three model types map directly onto the test procedures defined in UN Regulation No. 127 (UN-R127) and Euro NCAP pedestrian testing protocols:

The lower legform impactor is a rigid-body assembly with deformable tibia and knee elements. In PAM-CRASH it is typically delivered as a ready-to-run LS-DYNA-format or PAM-CRASH-native input deck containing shell and solid elements for the tibia, femur, knee ligaments, and skin. The upper legform impactor is a simpler beam-like device with a force transducer at the impact face and a bending moment measurement section. The full pedestrian FE model — most commonly the THUMS (Total Human Model for Safety) or GHBMC pedestrian variant — is a complete human body with articulated joints, deformable soft tissue, and a detailed skeleton, used when the interaction between the vehicle front-end and the whole body must be captured.

Selection Criteria: Which Model for Which Question

The choice is not arbitrary. It follows from the engineering question being asked:

• Bumper and grille region optimization → lower legform. This is the standard certification test. If the goal is to pass UN-R127 or Euro NCAP legform assessments, use the legform impactor.
• Bonnet leading edge and upper bumper → upper legform. This assesses thigh and pelvis loading where the vehicle front-end is tall enough to strike above the knee.
• Head impact on bonnet, A-pillar, and windscreen → full pedestrian model or a free-flight headform. The full model is required when body kinematics influence head trajectory.
• Multi-region interaction, wrap-around, or ground contact → full pedestrian FE model. Only a full body captures secondary impacts and the effect of body rotation on injury distribution.

Positioning the Legform Impactor in PAM-CRASH

Positioning is defined by the vehicle geometry, not by the model's local origin. The procedure in PAM-CRASH follows the protocol's geometric construction:

1. Establish the vehicle reference system. The ground plane is Z = 0. The vehicle centerline is Y = 0. The front bumper is at the X position defined by the vehicle CAD. All impactor positions are expressed in this global frame.
2. Determine the impact point. For the lower legform, the impact point is the point on the bumper surface at the vehicle centerline (Y = 0) that is 25 mm below the bumper reference line, or as specified by the protocol's corner and center impact locations.
3. Set the impactor orientation. The legform's longitudinal axis is aligned with the vehicle's X-axis (travel direction). The impactor is oriented so its front face is perpendicular to the vehicle surface at the impact point. In PAM-CRASH, this is done by defining a local coordinate system for the impactor and applying a rigid-body transformation.
4. Apply the initial velocity. The impactor is given an initial velocity of 40 km/h (11.11 m/s) in the negative X direction (toward the vehicle). In PAM-CRASH this is set via *INITIAL_VELOCITY_RIGID_BODY or the equivalent PAM-CRASH keyword, applied to the impactor's rigid body part set.
5. Define contact. A surface-to-surface contact is defined between the impactor's outer skin and the vehicle's front-end surfaces. The friction coefficient is typically 0.3, as specified in the protocol.

For the upper legform, the positioning is similar but the impact point is at the bonnet leading edge, and the impactor is oriented to strike the vehicle surface at the specified angle. The upper legform's impact height is determined by the vehicle's front-end geometry and the protocol's impact point calculation.

Positioning the Full Pedestrian FE Model

The full pedestrian model requires a different positioning philosophy. Instead of a single impact point, the model is placed in a standing or walking posture and given an initial velocity relative to the vehicle. The key steps are:

1. Posture selection. The model is posed in a standard standing position with arms at the sides or in a mid-gait position, depending on the protocol. Euro NCAP TB024 specifies a standing posture with the leading leg slightly forward.
2. Vehicle-relative positioning. The pedestrian is placed so that the vehicle's front centerline aligns with the pedestrian's mid-sagittal plane. The initial gap between the pedestrian's leg and the bumper is typically 25 mm.
3. Initial velocity. The vehicle is given a forward velocity (e.g., 40 km/h) and the pedestrian is given a lateral velocity (e.g., 5 km/h) to simulate walking motion, or both are set according to the specific test matrix.
4. Contact definition. A global contact is defined between the pedestrian's outer skin and all vehicle exterior surfaces. The pedestrian model's internal contacts (between body parts) are typically already defined in the model's input deck.

Model Selection and Positioning Workflow

Practical Verification and Common Pitfalls

After positioning, verify the setup before committing to a full run. A quick check is to run a short simulation (5–10 ms) and inspect the impactor's initial contact with the vehicle surface. The contact should occur at the intended impact point, and the impactor should not penetrate the vehicle geometry before the intended time. In PAM-CRASH, use the CONTACT output and the GLSTAT file to confirm that contact forces develop smoothly.

Common positioning errors include:

• Incorrect ground plane. If Z = 0 is not set at the tire contact patch, the impactor will strike too high or too low. Always verify the vehicle's ride height in the model.
• Misaligned impactor axis. A few degrees of yaw or pitch can shift the impact point by tens of millimeters. Use the protocol's geometric construction to define the impactor's local frame, not visual estimation.
• Wrong initial velocity direction. The impactor must travel toward the vehicle. A sign error in the velocity vector will send it away from the target.
• Missing contact definitions. If the impactor's outer skin is not included in the contact set, the simulation will show no interaction and the impactor will pass through the vehicle.

For the full pedestrian model, additional verification is needed: check that the model's internal contacts (e.g., between the femur and pelvis, or between the skull and brain) are active. These are typically defined in the model's input deck, but they can be inadvertently deactivated if the model is imported into a new PAM-CRASH database. Use the CONTACT entity list to confirm that all internal contact definitions are present and assigned to the correct part sets.

Finally, document the exact model version, posture, and positioning parameters in the simulation report. Pedestrian protection results are highly sensitive to these inputs, and reproducibility depends on recording them precisely. The model version matters: THUMS v4 and v5 have different mesh densities and material models, and results are not directly comparable between versions.