3. Sled system architecture

A crash pulse replication sled is an integrated system in which every subsystem — propulsion, sled structure, braking, fixture, and instrumentation — contributes to how faithfully the target deceleration pulse is reproduced. Understanding the architecture means understanding how these subsystems interact, because a pulse that looks correct on a bench calibration can still be wrong once the sled is loaded with a body-in-white and a fixture stack. This section walks through each subsystem, then shows how they combine in a typical pulse-tuning workflow.

Sled types and their pulse envelopes

Deceleration sleds fall into three broad families, distinguished mainly by how the pulse is generated and how much energy the system can absorb.

• Bumper/fixture sleds (rigid barrier type): the sled is accelerated and then decelerated by a fixed or deformable barrier. Pulse shape is largely determined by the barrier's crush characteristic. These are common for component-level tests where a short, high-g pulse is acceptable.
• Servo-hydraulic deceleration sleds: a programmable hydraulic actuator controls the sled's deceleration profile directly. Pulse shape, duration, and peak g can be tuned independently, which makes them the workhorse for occupant-restraint development.
• Accelerator sleds (reverse configuration): the sled is accelerated to a target velocity and the pulse is generated during acceleration rather than deceleration. These are used when the test article must experience a specific acceleration history, for example in some seat and battery-pack evaluations.

The choice of sled type constrains the achievable pulse. A servo-hydraulic deceleration sled can typically produce pulses from roughly 10 g to over 60 g with durations from 30 ms to 150 ms, while a barrier sled is limited to whatever the barrier material delivers. Before any test, the engineer should confirm that the target pulse lies inside the sled's validated envelope; otherwise the fixture design will be fighting the machine.

Propulsion methods

Propulsion determines the pre-impact velocity, which is the single most important input to the pulse. Three methods dominate:

• Bungee/cord propulsion: stretched elastic cords store energy and release it to accelerate the sled. Simple and inexpensive, but velocity repeatability is sensitive to cord age, temperature, and stretch length. Typical velocity scatter is ±2–3 %.
• Pneumatic propulsion: compressed air drives the sled through a piston or a direct-acting cylinder. Faster setup and better repeatability than bungees, with velocity controlled by reservoir pressure and valve timing.
• Hydraulic propulsion: a servo-controlled hydraulic system accelerates the sled to a precise velocity. Highest repeatability and the ability to program the acceleration phase, at the cost of complexity and maintenance.

Velocity is set from the target pulse using the impulse–momentum relationship. For a simple triangular deceleration pulse with peak a_p and duration t_d, the velocity change is approximately Δv = ½ · a_p · t_d. For example, a 40 g peak over 80 ms gives Δv ≈ 0.5 × 40 × 9.81 × 0.08 ≈ 15.7 m/s. The propulsion system must therefore deliver that velocity to the sled plus fixture mass, and the braking system must absorb the corresponding kinetic energy.

Braking systems

Braking is where the pulse is actually shaped in a deceleration sled. The braking system must absorb the sled's kinetic energy while producing the desired deceleration–time history. Common approaches include:

• Deformable metal tubes or honeycomb: the sled pushes a probe through a crushable element. Energy absorption is nearly constant, producing a roughly square pulse. Cheap and repeatable, but the pulse shape is fixed by the element geometry.
• Hydraulic dampers with programmable orifices: the deceleration force is set by orifice area and fluid viscosity. By staging multiple orifice stages, a trapezoidal or stepped pulse can be produced.
• Servo-hydraulic braking: a closed-loop controller commands the brake force in real time based on the measured sled acceleration. This is the only method that can track an arbitrary target pulse, and it is the standard for occupant-safety sleds.

In a servo-hydraulic system, the controller compares the measured acceleration to the target and adjusts the brake command at a loop rate of typically 1–5 kHz. The tuning parameters are proportional gain, integral gain, and a feed-forward term derived from the target pulse. Too much proportional gain produces oscillation and a jagged pulse; too little produces a sluggish rise and a low peak.

Fixture design

The fixture is the mechanical interface between the sled and the test article. Its job is to hold the article in the correct attitude, transmit the pulse without introducing its own dynamics, and survive the test. Key design rules:

• Stiffness first: the fixture's first natural frequency should be at least three times the highest frequency content of the target pulse. For a pulse with significant energy up to 100 Hz, the fixture should be stiff above 300 Hz.
• Mass budget: fixture mass adds to the sled mass and changes the required propulsion energy. Keep it as low as stiffness allows, and record it in the test setup.
• Alignment: the test article's principal axis must align with the sled's direction of travel. A 2° misalignment can produce a 3–5 % error in the measured longitudinal acceleration.
• Fastener torque: loose fasteners are a leading cause of pulse anomalies. Specify torque values and verify them before every run.

Instrumentation for pulse measurement and control

Pulse measurement relies on accelerometers mounted on the sled structure, not on the test article, so that the control loop sees the sled's motion. Typical instrumentation includes:

• Piezoresistive accelerometers (e.g., 0–500 g range) for the main pulse measurement, chosen for their DC response and ruggedness.
• Piezoelectric accelerometers for high-frequency content and for verifying fixture ringing.
• Velocity sensors or differentiated displacement for cross-checking the integrated velocity.
• Data acquisition at 10–20 kHz per channel with anti-alias filtering at 1/3 of the sample rate.

The measured acceleration is integrated once to get velocity and twice to get displacement. A common check is that the final velocity should return to zero (or to the known residual) and the displacement should match the sled's physical travel. If the integrated velocity drifts, the accelerometer is likely mis-scaled or the signal is offset.

The architecture above shows the essential signal and energy paths. Propulsion delivers velocity to the sled and fixture; braking converts kinetic energy into a controlled deceleration; instrumentation measures the resulting acceleration and feeds it back to the controller and to the test engineer. The fixture sits between the sled and the test article, and its stiffness and mass determine whether the pulse measured on the sled is the pulse experienced by the article. In practice, the most common source of pulse error is not the sled or the brake but the fixture — a resonant fixture will ring at its natural frequency and superimpose that ringing on the pulse, producing a trace that looks acceptable in peak g but is wrong in shape. Verifying fixture stiffness with a modal tap test before the first run is a cheap way to avoid a costly re-test.