2. Fundamentals of vehicle crashworthiness

Crashworthiness is the ability of a vehicle structure and its restraint systems to protect occupants and other road users during a collision. In LS-DYNA correlation work, crashworthiness is not an abstract property — it is expressed through specific crash modes, standardized test protocols, and measurable injury criteria that the simulation must reproduce. Before any model can be validated, the analyst must understand what physical event is being simulated, which regulatory or consumer test defines it, and which metrics determine pass or fail.

2.1 Crash Modes and Load Paths

Vehicle crashes are classified by impact direction and structural engagement. The principal modes are:

• Full frontal rigid barrier — the entire front end contacts a fixed wall; the primary load path runs through the front rails, bumper beam, and cradle into the rocker and floor.
• Offset deformable barrier (ODB) — only part of the front end engages a deformable honeycomb barrier, producing asymmetric loading and significant yaw and cabin intrusion.
• Side impact — a moving deformable barrier strikes the door and B-pillar; the load path is short and stiff, with limited crush space.
• Rear impact — a moving barrier strikes the rear structure, primarily a seat and head-restraint problem governed by whiplash metrics.
• Pole and small-overlap — a narrow rigid object engages a small portion of the front structure, bypassing the main rails and loading the wheel and A-pillar directly.
• Rollover — quasi-static roof crush and dynamic rollover events test roof strength and occupant retention.

Each mode produces a distinct deformation pattern and deceleration pulse. In LS-DYNA, the analyst must confirm that the model reproduces the correct collapse sequence: which members buckle first, where plastic hinges form, and how the load transfers to the occupant compartment. A model that matches the peak deceleration but deforms in the wrong order is not correlated.

2.2 Regulatory and Consumer Test Protocols

Two families of protocols dominate vehicle development. Regulatory tests (FMVSS in the United States, ECE/EU regulations in Europe, and equivalent national standards) define legal minimum requirements. Consumer tests (NHTSA NCAP, Euro NCAP, IIHS, C-NCAP, and others) are more severe and drive star ratings and market perception.

For correlation, the analyst must reproduce the exact boundary conditions of the chosen protocol: barrier geometry, honeycomb properties, impact velocity, and the mass of the moving deformable barrier. A common error is to run a generic frontal impact and then compare it against an ODB test — the load paths and intrusion levels are fundamentally different.

2.3 Injury Metrics and Their Physical Meaning

Injury criteria convert measured accelerations, forces, and deflections into a normalized risk value. The most important are:

• HIC (Head Injury Criterion) — integrates resultant head acceleration over the most critical time window (typically 15 ms or 36 ms). HIC₁₅ = 700 is a common regulatory limit; Euro NCAP uses HIC₁₅ and HIC₃₆.
• Chest deflection and Viscous Criterion (VC) — chest compression and the rate of compression; VC combines deflection and velocity to capture soft-tissue injury.
• Femur force — axial load transmitted through the femur, typically limited to 10 kN.
• Nij (Neck Injury Criterion) — combines axial force and bending moment in the neck.
• Pelvis and rib deflection — measured in side impact using the ES-2re or SID-IIs dummy.

These metrics are computed from the dummy response, not from the vehicle structure directly. In LS-DYNA, the analyst extracts accelerometer and load-cell time histories from the dummy model and post-processes them with the same filtering and windowing used in the physical test. Filtering class (CFC 600 for HIC, CFC 180 for chest) must match the test laboratory's practice, otherwise the comparison is invalid.

2.4 From Test Protocol to Simulation Setup

Translating a protocol into an LS-DYNA deck requires four inputs: the barrier model, the impact velocity, the dummy model, and the instrumentation. For a Euro NCAP ODB test, the barrier is a 1400 kg trolley with a deformable honeycomb face; the vehicle is towed at 64 km/h with 40% overlap. The simulation must include the honeycomb material model (typically *MAT_HONEYCOMB or *MAT_CRUSHABLE_FOAM) calibrated to the barrier certification data.

Consider a practical example. A sedan model is prepared for a 64 km/h ODB run. The analyst sets the initial velocity using *INITIAL_VELOCITY_GENERATION on the vehicle part set, defines the barrier as a rigid body with prescribed motion, and activates contact between the barrier face and the vehicle front structure. The dummy is positioned per the protocol using *INITIAL_VELOCITY and *BOUNDARY_PRESCRIBED_MOTION for the seat and belt. After the run, the analyst plots the B-pillar acceleration, the firewall intrusion, and the dummy HIC₁₅.

Typical failure symptoms and their causes:

• HIC too high, structure matches — airbag firing time or belt force limiter setting is wrong.
• Intrusion too low — contact stiffness or material strain-rate parameters are too stiff.
• Barrier penetrates vehicle — contact definition missing or penalty stiffness too low.
• Energy balance error — hourglassing or unphysical element deletion.

Each symptom points to a specific modeling decision, and resolving it requires understanding both the physical test and the numerical method. This is the essence of correlation: not matching a single curve, but reproducing the correct physical mechanism under the correct protocol with the correct metrics.

The chain shown above is the backbone of every correlation study. A mismatch at any link — wrong crash mode, wrong protocol, or wrong metric extraction — invalidates the comparison regardless of how well the mesh or material model is built. In the next stage of the workflow, these fundamentals become the acceptance criteria against which the LS-DYNA model is judged.