3. Test facilities and equipment

FMVSS 214 and UN R135 both evaluate occupant protection in lateral collisions, but they do so with different hardware, different kinematics, and different measurement chains. Before any test can be judged valid, the facility itself must be understood as a measurement system: the sled or barrier that delivers the pulse, the deformable face that shapes the loading, the rigid pole that concentrates it, and the instrumentation that converts motion into defensible numbers. This section walks through each element as it is actually used on a test floor, with the practical parameters and acceptance checks that determine whether a run is usable.

Sled systems and pulse generation

A crash sled is a guided mass accelerated or decelerated by a propulsion system — typically a hydraulic ram, a bungee or pneumatic catapult, or a servo-hydraulic actuator with closed-loop control. For side impact work the sled is most often used in deceleration mode: the carriage is brought to a target velocity and then stopped by programmable energy-absorbing elements (crush tubes, hydraulic dampers, or a servo-controlled brake) so that the resulting acceleration-time history matches a prescribed corridor.

The key concept is the pulse corridor. Regulations and internal test specifications define an envelope for the sled acceleration pulse, usually as a band around a nominal curve with tolerances on peak acceleration, pulse duration, and velocity change (Δv). A typical side-impact sled pulse might be specified as a peak of roughly 20–30 g with a duration of 50–80 ms and a Δv in the range of 25–40 km/h, but the exact values depend on the program. The sled must reproduce the corridor within tolerance across repeated runs; this repeatability is verified by integrating the measured acceleration to confirm Δv and by comparing the pulse shape against the corridor at multiple points.

Practical checks on a sled run include:

• Velocity at impact — measured independently (e.g., by a light gate or a flag system) and compared with the integrated accelerometer signal; a mismatch indicates accelerometer drift or triggering error.
• Pulse shape fidelity — the measured curve must stay inside the corridor; excursions outside indicate the energy absorber was mis-sized or the sled mass was wrong.
• Carriage alignment and rail condition — lateral or vertical acceleration on the carriage should be small; excessive off-axis motion contaminates the test and can damage the rail.

Sled testing is used for component-level and occupant-level work — seat and restraint evaluation, door trim interaction, and dummy response studies — where the full vehicle is not required. It is faster, cheaper, and more repeatable than a full-scale barrier test, which is why it is the workhorse for development iterations.

Moving deformable barriers (MDB)

The moving deformable barrier is the hardware that delivers the side impact in a full-scale test. In FMVSS 214 the barrier is a rigid, honeycomb-faced structure mounted on a moving carriage; in UN R135 (and the related UN R95 regime) the barrier is a deformable honeycomb block with a defined geometry and crush characteristic. The barrier is propelled along a track and strikes the stationary vehicle at a specified angle and velocity.

The barrier's deformable face is not a generic bumper — it is a calibrated energy absorber. Its honeycomb cells are specified by density, cell size, and depth, and the face is designed so that the force-deflection characteristic is repeatable. Before use, barriers are inspected for face flatness, honeycomb damage, and correct mounting; a barrier that has been used repeatedly will show crushed cells and must be refurbished or replaced. The barrier mass, the impact velocity, and the impact angle are the three parameters that most strongly determine the severity of the test, and each is recorded and verified on every run.

For UN R135 the barrier is typically a 1,300 kg-class moving deformable barrier striking the vehicle at 50 km/h with a 90° or offset configuration depending on the test variant. The barrier's honeycomb face is the element that spreads the load over the door and B-pillar region, and its condition directly affects the measured intrusion and dummy response. A common failure mode in barrier testing is a barrier that has drifted out of its calibrated crush characteristic; the symptom is a pulse that is too soft or too harsh relative to the reference, and the remedy is barrier replacement or re-certification, not adjustment of the vehicle.

The rigid pole

The rigid pole test is the second full-scale configuration in both FMVSS 214 and UN R135. Here the vehicle is propelled sideways into a fixed, rigid cylindrical pole — typically 254 mm in diameter — at a specified velocity and orientation. The pole is deliberately non-deformable: it concentrates the impact over a narrow contact patch, producing high local intrusion and a severe occupant loading that the broad MDB face does not reproduce.

The pole is mounted on a rigid base and its position is set relative to the vehicle's longitudinal axis so that the impact occurs at the intended location (for example, aligned with the occupant's seating position). The pole's diameter, height, and mounting stiffness are fixed by the regulation; the facility's job is to verify that the pole is vertical, that the base is rigid, and that the vehicle's path and velocity are correct at contact. Because the pole does not deform, all of the energy absorption must come from the vehicle structure, which is why pole tests are used to evaluate the strength of the door, B-pillar, and roof rail as a load path.

Instrumentation and the measurement chain

The instrumentation package is what turns a crash into data. In a side impact test the essential channels are:

All channels must be time-aligned. The standard practice is to use a common trigger — typically a contact switch or a light gate at the barrier — and to record a pre-trigger buffer so that the moment of contact is captured. Sampling rates for accelerometers are usually 10 kHz or higher; displacement channels may be lower but must be sufficient to resolve the intrusion event. The data acquisition system must be calibrated before the test, and the calibration records are part of the test documentation.

Putting the facility together

The figure below shows how the four elements — sled, MDB, rigid pole, and instrumentation — relate as a measurement system, with the signal and data paths that connect them.

The diagram makes the facility's logic explicit: the sled is the propulsion source, the MDB and the rigid pole are two alternative impact faces that shape the loading differently, the vehicle is the stationary target, and every element feeds a common instrumentation chain. The dashed path from the barrier and pole to the data acquisition block is a reminder that the impact face itself carries sensors — barrier carriage accelerometers and pole-base accelerometers — and that these signals must be time-aligned with the vehicle and dummy channels.

Facility acceptance and run validity

Before a facility is used for certification or development testing, it must demonstrate repeatability. A practical acceptance sequence is to run the same configuration three times and compare the pulse corridors, the Δv, and the dummy response channels. If the run-to-run variation in peak acceleration exceeds a few percent, or if the Δv varies by more than about 1 km/h, the facility is not yet stable enough for judgment-level testing. Common causes are inconsistent energy-absorber condition, rail wear, or trigger timing drift.

On each individual run, the validity checklist is short but non-negotiable: impact velocity matches the integrated acceleration; the pulse stays inside the corridor; the barrier or pole is in its specified condition; all channels are time-aligned and within calibration; and the high-speed video confirms the event sequence. Only when these are satisfied should the dummy response data be interpreted against the injury criteria. Treating the facility as a measurement system — rather than as a collection of machines — is what separates a defensible test result from a number that merely looks plausible.