3. Side impact load cases and barriers

Side impact simulation is only meaningful when the virtual setup reproduces a recognized physical test. In LS-DYNA side impact work, the load case defines almost everything downstream: barrier geometry and mass, impact velocity and angle, the struck vehicle's boundary conditions, and the injury criteria you will ultimately extract. This section reviews the principal side impact configurations — EEVC MDB, NHTSA moving deformable barrier (MDB), and pole impact — and explains how each is translated into a simulation model.

Why the Barrier Defines the Physics

A side impact differs fundamentally from a frontal crash because there is very little crush space between the struck vehicle's outer skin and the occupant. The barrier's face geometry and stiffness therefore dominate the loading path. A rigid pole concentrates force over a narrow contact patch and drives intrusion deep into the occupant compartment; a deformable barrier spreads load over a wider area and itself absorbs energy. Choosing the wrong barrier model changes intrusion by tens of millimeters and invalidates any occupant response you compute.

EEVC Moving Deformable Barrier (MDB)

The EEVC MDB, standardized in ECE R95 and used in Euro NCAP, is a 950 kg trolley carrying a deformable aluminium honeycomb face. The barrier is 1500 mm wide and 500 mm high, and it strikes the stationary vehicle perpendicular to its longitudinal axis at 50 km/h. The honeycomb is divided into six blocks with graded stiffness so that the lower blocks are softer than the upper blocks, approximating a car front end.

In LS-DYNA the barrier is typically represented with *MAT_HONEYCOMB or *MAT_CRUSHABLE_FOAM calibrated to the block-level force–deflection corridors. The trolley itself is modeled as rigid with a point mass. A common modeling error is to over-stiffen the honeycomb; the resulting intrusion will be too high and the barrier will not absorb its share of energy. Always verify the barrier's standalone crush response against the regulatory corridor before coupling it to the vehicle.

NHTSA Moving Deformable Barrier

The NHTSA MDB (FMVSS 214, also used in US NCAP and SINCAP) is heavier and faster than the EEVC barrier: 1361 kg at 62 km/h, with a 27° crabbed angle so the barrier's longitudinal axis is rotated relative to its velocity vector. The deformable face is a honeycomb structure with different block stiffnesses from the EEVC design. The crabbed motion means the barrier both translates and rotates into the vehicle, producing a combined lateral and longitudinal load that is more severe for the door and B-pillar.

Modeling the crab angle correctly is critical. In LS-DYNA, define the barrier's initial velocity vector along the global X-axis while orienting the barrier geometry 27° about the vertical axis. Do not simply rotate the velocity vector — that produces an incorrect contact sequence. The initial velocity card should read, for example:

*INITIAL_VELOCITY
$#   nsid       nsidex     boxid
         1         0         0
$#    vx        vy        vz       vxr       vyr       vzr
  17222.0       0.0       0.0       0.0       0.0       0.0

Here 17222 mm/s corresponds to 62 km/h along global X, while the barrier mesh is pre-rotated 27° in the keyword deck. The vehicle is stationary and constrained at the wheels with a ground plane and tire contact.

Pole Impact

Pole impact (FMVSS 201, Euro NCAP side pole) uses a rigid vertical cylinder, 254 mm diameter, and drives the vehicle sideways into the pole at 32 km/h (Euro NCAP) or 29 km/h (FMVSS 201) with a 75° angle between the vehicle's longitudinal axis and the pole. Because the pole is rigid and narrow, it produces localized door intrusion and is the most demanding case for side airbag and curtain deployment logic.

In simulation the pole is a rigid shell with *MAT_RIGID and a fixed boundary condition. The vehicle is given an initial velocity and yaw so that it contacts the pole at the specified angle. Contact between the pole and the vehicle exterior uses *CONTACT_AUTOMATIC_SURFACE_TO_SURFACE with a friction coefficient around 0.3. Because the pole is small, mesh density on the door and B-pillar must be fine enough — typically 5–8 mm — to capture the local bending and avoid contact penetration.

Comparing the Configurations

Translating a Load Case into an LS-DYNA Deck

A reproducible side impact setup follows a consistent sequence. First, import or build the barrier and position it relative to the vehicle using the regulatory geometry — for the EEVC MDB, the barrier centerline aligns with the vehicle's R-point or a specified offset. Second, assign initial velocity to the barrier (or vehicle for pole impact) using *INITIAL_VELOCITY. Third, define contact between barrier face and vehicle exterior, and between vehicle interior and occupant dummy if present. Fourth, apply gravity and ground contact to the vehicle. Fifth, set termination time to at least 150 ms for MDB cases and 120 ms for pole, with *DATABASE_BINARY_D3PLOT output every 1–2 ms.

After the run, check the barrier's kinetic energy history: it should decrease smoothly as the honeycomb crushes. A sudden drop or a negative sliding interface energy indicates contact problems. Then examine door intrusion at the occupant reference points and compare against the regulatory limit — for example, Euro NCAP requires that the door not intrude beyond a defined threshold at the abdomen and pelvis locations.

Common Failure Modes and Troubleshooting

• Barrier passes through the vehicle: contact thickness or segment size mismatch. Increase SSTHK or refine the vehicle exterior mesh.
• Honeycomb does not crush: material curve too stiff or *MAT_HONEYCOMB direction vectors misaligned. Verify the local axis orientation.
• Unrealistic B-pillar rotation: missing or incorrect spot weld definitions. Check *CONSTRAINED_SPOTWELD or *MAT_SPOTWELD properties.
• Energy balance error above 5%: hourglassing in under-integrated shells. Switch to *SECTION_SHELL with ELFORM 16 or add hourglass control.

The three configurations above cover the majority of regulatory and consumer-test side impact scenarios. The key discipline is to treat the barrier as a calibrated subsystem, not a placeholder: verify its standalone crush, confirm the velocity vector and orientation match the regulation, and check energy balance before interpreting occupant results. Once the load case is correct, the rest of the side impact model — door trim, airbag, dummy positioning — can be evaluated with confidence.