2. Head impact simulation workflow

The head impact simulation workflow is the backbone of any pedestrian protection assessment. It transforms a raw vehicle exterior surface into a validated set of impact scenarios, each defined by a precise impact point, a launch vector, a headform model, and a set of boundary conditions. This chapter walks through the complete chain: preparing vehicle geometry, constructing the headform, defining impact points according to the regulation, and assembling a runnable LS-DYNA model. The emphasis is on the decisions that determine whether the simulation is physically meaningful or merely a numerical exercise.

2.1 Vehicle Geometry Preparation

The starting point is the vehicle exterior surface, typically delivered as a CAD assembly in STEP or CATIA format. For pedestrian head impact, only the outer skin matters: hood outer panel, fenders, A-pillars, windshield, roof, and any exposed structural members. Interior components are irrelevant unless they influence the outer panel's deformation through contact.

The first task is to extract a clean outer surface. In practice, this means removing trim, badges, and under-hood components, then healing gaps and slivers that would cause shell element distortion. A common target is a watertight surface with no free edges along the impact path. The surface is then meshed with shell elements, typically 8–12 mm in size for the outer panel, refined to 5 mm near impact points where bending gradients are steep. The hood inner panel and any reinforcement ribs are meshed separately and connected to the outer panel through spot welds or adhesive constraints.

Material definition follows. For a typical steel hood outer panel, use MAT_PIECEWISE_LINEAR_PLASTICITY (MAT_024) with a density of 7.85e-9 tonne/mm³, Young's modulus of 210 GPa, and a true stress–plastic strain curve derived from tensile tests. For aluminum hoods, MAT_024 or MAT_036 (MAT_3-PARAMETER_BARLOW) is common. The key point is that the material card must include strain-rate sensitivity if the impact speed exceeds roughly 5 m/s, because head impact speeds of 9.7–11.1 m/s produce strain rates in the range of 100–1000 s⁻¹.

A practical check: after meshing, run a free-edge check and a normal-vector consistency check. Inverted normals on the outer panel will cause the headform to pass through the surface without contact, producing a physically impossible result that is easy to miss in the animation.

2.2 Headform Model Setup

The headform is a rigid sphere with a deformable skin, defined by the Global Technical Regulation No. 9 (GTR 9) and UN Regulation No. 127. The adult headform has a mass of 4.5 kg and a diameter of 165 mm; the child headform has a mass of 3.5 kg and a diameter of 165 mm. The skin is modeled as a layer of rubber-like material over a rigid core.

In LS-DYNA, the headform is typically represented by a rigid body (MAT_020) for the core and a layer of solid or shell elements for the skin. A common approach uses MAT_OGDEN_RUBBER (MAT_077) or MAT_MOONEY-RIVLIN_RUBBER (MAT_027) for the skin, with parameters calibrated to match the drop-test response specified in the regulation. The calibration target is a peak resultant acceleration of 225 g ± 15 g when dropped from 376 mm onto a flat rigid plate, with a specific pulse shape.

The headform model must include an accelerometer output at its center of gravity. In LS-DYNA, this is done with *DATABASE_HISTORY_NODE or *DATABASE_ASCII_OPTION, referencing the node at the headform's CG. The resultant acceleration is computed as the vector sum of the three translational components. The Head Injury Criterion (HIC) is then calculated from the acceleration–time history:

HIC = max over t1,t2 of [ (1/(t2-t1)) * ∫ a(t) dt ]^2.5 * (t2-t1)
where a(t) is in g and t2-t1 ≤ 15 ms

The headform is positioned so that its center is at the impact point, with the impact vector normal to the vehicle surface at that point. The initial velocity is applied to the entire headform as a prescribed initial velocity, typically 9.7 m/s for the adult headform and 11.1 m/s for the child headform, corresponding to a 40 km/h vehicle impact speed.

2.3 Impact Point Definition

Impact points are defined by the regulation's wrap-around distance (WAD) method. The WAD is the distance along the vehicle surface from a reference line, measured in the longitudinal direction. For the adult headform, the impact zone is the area where the WAD is between 1700 mm and 2100 mm. For the child headform, the zone is between 1000 mm and 1700 mm. The boundary between child and adult zones is the WAD 1700 mm line.

Within each zone, impact points are selected on a grid. The regulation specifies a grid spacing of 100 mm in the longitudinal direction and 100 mm in the lateral direction, but the actual number of points depends on the vehicle width. For a typical sedan, this yields 8–12 points per side for the adult zone and 6–10 points for the child zone. Each point is projected onto the vehicle surface, and the impact vector is defined as the surface normal at that point.

The selection of impact points is not purely mechanical. Points must be placed on the hood, fenders, and A-pillars, but not on the windshield glass unless the glass is part of the outer surface. Points on the hood must avoid the hood hinges and latch, which are structural discontinuities that can produce artificially high accelerations. In practice, the engineer adjusts the grid to avoid these features while maintaining coverage of the zone.

A common error is to define impact points on the undeformed CAD surface and then apply them to the meshed model without checking the mesh normal. If the mesh normal at a point deviates from the CAD normal by more than 5°, the impact vector will be wrong, and the resulting HIC will be inaccurate. Always verify the normal direction in the preprocessor.

2.4 Assembling the LS-DYNA Model

The complete model consists of the vehicle mesh, the headform mesh, contact definitions, and boundary conditions. The contact between the headform skin and the vehicle outer surface is defined with *CONTACT_AUTOMATIC_SURFACE_TO_SURFACE, with the headform as the slave surface and the vehicle as the master. The friction coefficient is typically set to 0.3, based on experimental data for rubber-on-steel contact.

The vehicle body is constrained at the suspension mounting points and at the rear structure, simulating the vehicle's inertia during impact. A common approach is to apply *BOUNDARY_SPC constraints at the suspension pickup points and to include a lumped mass at the rear to represent the vehicle's total mass. The simulation time is typically 30–50 ms, which is sufficient for the headform to rebound from the surface.

Output requests include the headform CG acceleration, the contact force, and the vehicle panel deformation. The acceleration history is filtered with a CFC 1000 filter, as specified in the regulation, before calculating HIC. The final result is a HIC value for each impact point, which is then compared against the regulatory limit of 1000 for the adult headform and 1000 for the child headform (with a lower limit of 700 in some jurisdictions).

Head Impact Simulation Workflow — Vehicle Geometry to Impact Point Definition Vehicle Geometry CAD outer surface Hood, fender, A-pillar Mesh 8–12 mm shells MAT_024 steel / Al Headform Model Adult 4.5 kg / Child 3.5 kg Ø165 mm rigid core MAT_077 Ogden skin CG accelerometer node Impact Point Definition WAD 1000–2100 mm Grid 100 × 100 mm Surface normal vector Avoid hinges / latch LS-DYNA Contact SPC / mass 30–50 ms run HIC output Impact Point Grid on Vehicle Plan View (Schematic) Hood Outer Panel Windshield / A-Pillar Zone WAD 1700 mm WAD 2100 mm Adult headform impact points (WAD 1700–2100 mm) Child headform impact points (WAD 1000–1700 mm) n (surface normal) HF

2.5 Practical Considerations and Common Pitfalls

The workflow described above is conceptually linear, but in practice it is iterative. A typical iteration loop involves running a subset of impact points, checking the HIC values, and then refining the mesh or material parameters if the results are inconsistent with physical expectations. For example, if a point on the hood produces a HIC of 400 while an adjacent point produces 1200, the difference is usually due to a local mesh distortion or an incorrect material assignment, not a real physical effect.

Another practical issue is the treatment of the windshield. In many vehicles, the windshield glass is laminated and bonded to the body. If the glass is included in the model, it must be represented with a layered shell or solid element formulation that captures the glass–PVB–glass sandwich. If it is excluded, the A-pillar and roof rail must be modeled accurately, because they become the primary load path. The choice depends on the regulatory zone: for the adult headform, the windshield is often outside the WAD 1700–2100 mm zone, but for the child headform, it may be inside.

Finally, the headform skin material model requires calibration. The Ogden parameters are not universal; they depend on the specific rubber compound and the manufacturing process. A common approach is to start with published parameters from the regulation's calibration procedure and then adjust the shear modulus to match the drop-test peak acceleration. This calibration step is essential because the skin's stiffness directly affects the contact force and the resulting HIC.

Key deliverables for this stage:

  • A meshed vehicle outer surface with verified normals and material cards.
  • A calibrated headform model with a CG accelerometer output.
  • A set of impact points defined by WAD and grid spacing, with surface normal vectors.
  • An assembled LS-DYNA deck with contact, boundary conditions, and output requests.

The next step in the workflow is to run the simulations and post-process the acceleration histories to compute HIC. That process, along with the interpretation of results and the iterative design changes that follow, is covered in the subsequent chapter.

← Previous 📖 Contents Next →