2. Sled test vs. full-scale crash testing
In automotive crashworthiness development, engineers have two principal physical test methods at their disposal: the full-scale vehicle crash test and the crash sled test. Both produce deceleration pulses, structural deformation, and occupant loading data, but they differ fundamentally in what they control, what they simplify, and what they can and cannot tell you. Understanding these differences is not an academic exercise — it directly determines test cost, schedule, instrumentation strategy, and the validity of the conclusions you draw from the data.
What Each Method Actually Controls
A full-scale crash test propels an entire vehicle into a barrier, pole, or moving deformable barrier. The vehicle's front structure, engine cradle, subframe, and body-in-white all participate in the crash. The deceleration pulse experienced by the occupant compartment is an output of the test — it emerges from the interaction between the vehicle's mass, stiffness distribution, and the barrier geometry. The test engineer controls the impact speed, impact angle, barrier type, and vehicle configuration, but cannot independently prescribe the pulse shape.
A sled test inverts this relationship. The vehicle body is replaced by a rigid or semi-rigid sled buck that carries the seats, restraints, instrument panel, steering column, and occupant dummies. The sled is accelerated or decelerated by a propulsion system — typically a pneumatic or hydraulic catapult, or a deceleration sled with crushable or hydraulic energy absorbers — to reproduce a target pulse that was previously measured or simulated. Here the pulse is an input: the engineer prescribes it, and the sled control system tracks it.
This distinction is the single most important trade-off. Full-scale testing captures the complete vehicle system but gives you only one pulse per test. Sled testing isolates the occupant-restraint system and lets you prescribe any pulse you want, but it removes the vehicle structure from the equation.
Fidelity: Where Each Method Excels and Where It Breaks Down
Full-scale testing is the reference method for regulatory compliance and consumer rating programs — FMVSS 208, FMVSS 214, Euro NCAP frontal offset, and similar protocols all require full vehicle tests. Its fidelity advantages include:
- Realistic structural crush and load paths, including engine intrusion, pedal displacement, and steering column kinematics.
- Authentic occupant compartment deformation and its effect on restraint geometry.
- Accurate mass distribution and inertia loading on the restraint system.
- Validation of CAE models at the full-system level.
Its limitations are equally real: each test destroys a vehicle, costs tens to hundreds of thousands of dollars, requires weeks of build and instrumentation time, and yields a single data point. Parametric studies — varying airbag timing, seat belt load limiter force, or seat position — are prohibitively expensive at full scale.
Sled testing trades structural fidelity for parametric control. Its strengths include:
- Repeatable, precisely controlled deceleration pulses with tight tolerances (often ±1 g or better in the controlled region).
- Rapid iteration: a sled test can be run in hours, and dozens of configurations can be evaluated in the time one full-scale test takes.
- Direct isolation of restraint system variables — airbag pressure, vent size, tether length, pretensioner timing, load limiter profile.
- Lower cost per test and non-destructive to the vehicle body (the buck is reusable).
The fidelity gap appears when the sled buck cannot reproduce the structural deformation that a real vehicle undergoes. If the vehicle's toe pan crushes and displaces the pedals, or if the A-pillar deforms and moves the steering column, the sled buck — being rigid — will not capture that. Similarly, the sled cannot reproduce the full vehicle's yaw, pitch, or roll response unless a sophisticated multi-axis sled is used. For side impacts, the sled must simulate the door intrusion and the struck vehicle's motion, which is a significant simplification.
When to Use Each Method
The selection logic is driven by the question you are trying to answer. Use the following decision framework:
| Question | Preferred Method | Reason |
|---|---|---|
| Does the vehicle meet a regulatory or rating requirement? | Full-scale | Regulations specify full vehicle tests; sled results are not accepted for compliance. |
| How does the restraint system perform across a matrix of pulse severities? | Sled | Pulse is an input; rapid parametric sweeps are feasible. |
| Does the vehicle structure intrude into the occupant compartment? | Full-scale | Sled bucks are rigid and cannot reproduce intrusion. |
| Is the airbag deployment timing optimal for a given pulse? | Sled | Timing can be varied independently and measured precisely. |
| How does the seat belt load limiter affect chest deflection? | Sled | Direct isolation of the restraint variable; low cost per configuration. |
| Does the vehicle's mass distribution affect occupant kinematics? | Full-scale | Sled bucks approximate but do not replicate full vehicle inertia. |
| Is the CAE model correlated at the system level? | Full-scale | Full vehicle validation is required before sled-level correlation. |
In practice, the two methods are complementary, not competing. A typical development program uses full-scale tests to establish the reference pulse and validate the CAE model, then uses sled tests to optimize the restraint system across a range of conditions, and finally returns to full-scale testing for final verification and compliance. The sled test is the workhorse for iteration; the full-scale test is the arbiter of truth.
Practical Example: Pulse Tuning for a Frontal Sled Test
Consider a frontal occupant protection program for a mid-size sedan. The full-scale test at 56 km/h into a rigid barrier produces a measured occupant compartment deceleration pulse with a peak of approximately 45 g at 60 ms and a duration of about 120 ms. The engineer wants to evaluate three airbag vent sizes on the sled.
The sled buck is configured with the production seat, steering column, instrument panel, and a Hybrid III 50th percentile male dummy. The target pulse is loaded into the sled controller. The first sled run uses the measured full-scale pulse directly. The resulting dummy chest deflection is 38 mm, while the full-scale test showed 32 mm. The discrepancy arises because the sled buck is stiffer than the vehicle and the seat mounting differs slightly.
The engineer adjusts the pulse by reducing the peak to 42 g and extending the duration to 130 ms, which softens the loading on the occupant. After two iterations, the sled chest deflection matches the full-scale value within 2 mm. Only then are the three vent sizes tested. This tuning process — typically 2 to 4 runs — is a necessary step before any parametric study, and it is a key skill in sled test engineering.
Key Takeaways for Test Engineers
The choice between sled and full-scale testing is not a matter of one being superior. It is a matter of matching the method to the question. Full-scale testing answers "does the complete vehicle system meet the requirement?" Sled testing answers "how does the restraint system respond to a controlled input?" A well-run program uses both, with sled tests providing the iteration speed and full-scale tests providing the ground truth.
When planning a sled test campaign, always begin by establishing the target pulse from a validated full-scale test or a correlated CAE model. Then tune the sled pulse to match occupant response, not just the deceleration trace. Document the tuning iterations and the final pulse parameters, because those become the basis for every subsequent parametric run. Without this discipline, sled results can be precise but wrong — repeatable numbers that do not reflect what happens in a real vehicle.