3. Automotive materials under crash loading

Crashworthiness simulation in LS-DYNA is only as credible as the material models that drive it. A frontal rail that folds correctly in a physical sled test but tears prematurely in the model usually points to a mis-calibrated failure criterion rather than a meshing error. This chapter surveys the five material families that dominate modern body structures — steel, aluminum, magnesium, composites, and plastics — and explains how each one fails, which LS-DYNA material cards are typically used, and where calibration most often goes wrong.

Steel: Ductile Necking, Strain-Rate Sensitivity, and Spot-Weld Tearing

Advanced high-strength steels (AHSS) such as DP780, DP980, and press-hardened 22MnB5 remain the backbone of crash-critical structures. Their failure sequence under crash loading is well understood: uniform elongation, diffuse necking, localized necking, then ductile fracture. The calibration challenge is not the onset of plasticity but the point of fracture, which depends strongly on stress triaxiality and the Lode angle parameter.

In LS-DYNA, the workhorse combination is *MAT_024 (piecewise linear plasticity) or *MAT_037 for the constitutive response, paired with *MAT_ADD_EROSION or, preferably, *MAT_224 (GISSMO) for damage accumulation. GISSMO is preferred because it tracks damage non-linearly and supports element deletion based on a critical damage threshold rather than an instantaneous strain limit.

Strain-rate sensitivity is non-negotiable for steel. DP780 typically shows a 15–25% increase in flow stress at 100 s⁻¹ compared to quasi-static conditions. This is captured with *DEFINE_CURVE tables referenced by the LCSS field in *MAT_024. A common calibration error is using tensile test data at 0.001 s⁻¹ to represent a 10 s⁻¹ crash event, which under-predicts energy absorption by 10–20%.

Spot welds deserve separate attention. They are modeled with *MAT_100 (spot weld) or solid element clusters with *MAT_SPOTWELD. Failure is governed by a force-based or stress-based criterion, and the calibration must account for the weld nugget diameter, sheet thickness combination, and loading mode (peel vs. shear). A weld that holds in a coupon test may fail prematurely in a full vehicle model if the surrounding mesh is too coarse to resolve the stress gradient.

Aluminum: Lower Ductility, Anisotropy, and the Forming History Problem

Aluminum alloys — 5182, 6111, 7075 — offer weight savings but fail with far less ductility than steel. 7075-T6, for example, can exhibit less than 10% elongation at room temperature, making it sensitive to stress concentrations and manufacturing defects. The failure mode is typically shear-dominated fracture with limited necking, which means the calibration must capture the transition from ductile to shear fracture at low stress triaxiality.

Anisotropy is significant in rolled aluminum sheet. The Lankford coefficients (r₀, r₄₅, r₉₀) can differ by 30–50%, and the yield surface is better described by Barlat's Yld2000-2D than by von Mises. In LS-DYNA, this is implemented through *MAT_036 (Barlat) or *MAT_133 (Barlat with kinematic hardening). Calibration requires tensile tests at 0°, 45°, and 90° to the rolling direction, plus a balanced biaxial test if available.

The most under-appreciated challenge is forming history. An aluminum component that has been stamped and hemmed carries residual plastic strain and damage that alters its crash response. Ignoring this in the crash model — starting from a virgin material state — can shift the fracture initiation location and timing. The practical fix is a one-step or incremental forming simulation followed by *INITIAL_STRESS_SHELL mapping into the crash model.

Magnesium: Low Ductility, Twinning, and Temperature Sensitivity

Magnesium alloys such as AZ31 and AM60 are attractive for weight reduction but are the most difficult to calibrate for crash. Their hexagonal close-packed (HCP) crystal structure activates deformation twinning at room temperature, producing asymmetric yield in tension versus compression and a strong tension-compression asymmetry that von Mises plasticity cannot capture.

LS-DYNA offers *MAT_124 (plasticity with damage) and *MAT_233 (Barlat with asymmetric yield) for magnesium. The calibration dataset must include tension, compression, and shear tests, ideally at multiple strain rates. Temperature is also critical: at 150–200°C, magnesium becomes significantly more ductile, which matters for components near the exhaust or engine bay.

Fracture in magnesium is often brittle with little necking, and the failure strain is sensitive to stress state. A calibrated GISSMO model for AZ31 might show a fracture strain of 0.12 in uniaxial tension but only 0.04 in shear. This spread is larger than for steel or aluminum, making the calibration more demanding and the validation more critical.

Composites: Ply-Level Failure, Delamination, and the Mesh-Size Problem

Carbon-fiber reinforced polymer (CFRP) and glass-fiber reinforced polymer (GFRP) are increasingly used in crash structures, particularly in electric vehicle battery enclosures and front-end modules. Their failure is a sequence of interacting mechanisms: matrix cracking, fiber-matrix debonding, fiber pull-out, fiber fracture, and delamination between plies.

In LS-DYNA, the standard approach is *MAT_054 (enhanced composite damage) or *MAT_058 (Ladevèze) for the ply-level response, combined with cohesive elements or *MAT_COHESIVE_MIXED_MODE for delamination. The calibration challenge is that the failure parameters are mesh-size dependent. A coarse mesh will over-predict strength because it cannot resolve the stress concentration at the crack tip.

Practical calibration uses a combination of coupon tests: 0° tension, 90° tension, ±45° shear, and short-beam shear for interlaminar strength. The resulting parameters are then validated against a coupon-level crush test before being applied to a full structure. A common mistake is to calibrate on a single coupon type and assume the parameters transfer to a different layup or loading mode.

Plastics: Rate Dependence, Large Deformation, and Failure at High Strain Rates

Thermoplastics such as polypropylene (PP), polyamide (PA), and polycarbonate (PC) are used in bumper fascia, interior trim, and under-hood components. Their mechanical response is strongly rate-dependent: a PP copolymer might have a yield stress of 20 MPa at 0.001 s⁻¹ but 45 MPa at 100 s⁻¹. They also exhibit large deformation before failure, often exceeding 50% strain.

LS-DYNA uses *MAT_089 (plasticity polymer) or *MAT_187 (SAMP-1) for plastics. SAMP-1 is preferred for crash because it supports pressure-dependent yield, rate dependence, and damage. Calibration requires tensile tests at multiple rates, plus compression and shear tests if the component sees multi-axial loading.

Failure in plastics is often defined by a strain-to-failure criterion, but this is misleading because the failure strain depends on stress state and rate. A more robust approach is to use a damage-based model with a critical damage parameter, calibrated from notched tensile tests that span a range of stress triaxialities.

Cross-Material Calibration Challenges

Across all five families, three challenges recur. First, strain-rate data is often missing or extrapolated; a calibration based on quasi-static data alone will mispredict crash response. Second, mesh-size sensitivity means that failure parameters calibrated on a 1 mm mesh may not work on a 5 mm mesh; regularization techniques such as *MAT_ADD_EROSION with a characteristic length or GISSMO's mesh regularization must be used. Third, validation must go beyond a single coupon test; a component-level test that exercises the failure mode in the same stress state as the application is essential.

The next step in the calibration workflow is to translate these material-specific mechanisms into a structured test matrix and parameter identification procedure. The following chapters will walk through the coupon-level testing, curve fitting, and validation sequence for each family, starting with the steel grades that dominate current body-in-white designs.