3. Lighting for high-speed imaging

At 1,000–10,000 fps, the exposure window per frame collapses to 1 ms–100 µs. A sled test at 64 km/h moves the dummy head roughly 18 mm during a 1 ms exposure — enough to smear fine detail such as belt-webbing contact patches or airbag seam deployment. Lighting is therefore not a "support" subsystem; it is the parameter that determines whether the recorded image is measurable. This chapter covers source selection, synchronization, thermal limits, and how to make illumination uniform across dummies and vehicle interiors.

Continuous vs. Pulsed Sources: The Photometric Trade-Off

A continuous source (tungsten-halogen, LED panel, HMI) emits a constant luminous flux. Exposure is set by shutter angle and frame rate, so the required illuminance scales inversely with exposure time. Doubling frame rate halves exposure and demands double the illuminance for the same image brightness. This is the fundamental reason continuous lighting becomes impractical above a few thousand frames per second: the fixture count and electrical load grow linearly with frame rate.

Pulsed sources invert the relationship. A xenon or LED strobe delivers a very high flux for a short pulse — typically 10–100 µs — and the pulse itself acts as the shutter. Because the pulse is shorter than the frame interval, motion blur is bounded by pulse duration, not by frame rate. A 20 µs pulse at 5,000 fps freezes motion as effectively as a 20 µs exposure at any frame rate, while the peak intensity can be 10–100× the continuous rating of the same lamp.

In practice, most full-scale sled and crash facilities use pulsed xenon for the high-speed cameras and a small amount of continuous LED for alignment, focus, and reference. On-board cameras inside the vehicle often use continuous LED because strobe cabling and synchronization through a moving vehicle is difficult; the trade-off is a lower achievable frame rate or a wider aperture.

Strobe Synchronization

Synchronization means the strobe fires during the camera's exposure window, not before or after it. Three signals matter: the camera's frame sync (or "strobe out"), the strobe trigger, and the event trigger (the sled release or impact switch). The standard topology is a master clock or the fastest camera acting as master, distributing a frame-sync pulse to all cameras and to a strobe controller. The controller adds a programmable delay so the light pulse lands inside the exposure window.

Two parameters must be set explicitly. First, pulse delay: the time from the frame-sync edge to the start of the light pulse. Second, pulse width: the duration of the light pulse, which must be shorter than the exposure time to avoid light spilling into the next frame. A common starting point is a 20 µs pulse with a 5–10 µs delay after the sync edge, then trimmed by inspecting a single frame for even brightness.

Heat Management

Xenon strobes dissipate most of their input energy as heat in the lamp envelope and reflector. At 5,000 fps with a 20 µs pulse, the duty cycle is 10%, so average power is one-tenth of peak — but the peak is high enough that the lamp envelope can reach several hundred degrees Celsius. Practical measures include:

• Forced-air cooling across the reflector and lamp housing, with airflow directed away from the test article to avoid disturbing lightweight components.
• Duty-cycle limiting in the controller: cap the number of pulses per second and the maximum burst duration (e.g., 2 s at full power) to stay within the lamp's thermal rating.
• Thermal interlocks that inhibit firing if the housing exceeds a set temperature, preventing envelope failure and glass shrapnel near the test article.
• Heat filters (dichroic or IR-absorbing) between lamp and subject when the test article is heat-sensitive, such as a dummy with instrumented skin or a fabric airbag.

Continuous LED panels are cooler per lumen but still require heatsinking; a 500 W LED panel in a closed enclosure will derate within minutes. Mount drivers remotely and keep airflow across the heatsink fins.

Uniform Illumination of Dummies and Vehicle Interiors

Uniformity is judged by the variation in luminance across the region of interest (ROI). A useful target is within ±10% of the mean on the dummy's face and chest, and within ±15% inside the footwell or door cavity. Three techniques achieve this:

1. Large, diffuse sources close to the subject. A 1 m × 1 m softbox at 1.5 m produces far more even illumination than a bare reflector at the same distance, because the source subtends a larger solid angle.
2. Cross-lighting. Two strobes at 45° to the camera axis on opposite sides fill shadows from each other. For interior shots, a third strobe through the opposite window fills the far side.
3. Bounce and fill. A white bounce card or a low-power fill strobe inside the cabin lifts the shadow side without flattening the image.

Interior illumination is harder because the cabin is a closed box with reflective glass and dark trim. Practical steps: remove or open the doors on the camera side, place strobes outside the windows aimed inward, and use a matte white card on the far side to bounce light back. Avoid aiming directly at the windshield from outside — the reflection will dominate the frame. For dummy faces, a small on-axis fill at low power reduces the "raccoon" shadow under the brow without creating a specular hotspot on the vinyl skin.

Practical Setup and Verification

A reproducible starting configuration for a frontal sled test with a 50th-percentile male dummy and a 5,000 fps camera:

Camera: global shutter, 5000 fps, exposure 40 µs, f/2.8, ISO 800
Strobe A: xenon, 20 µs pulse, 45° left of camera axis, 2.0 m from dummy
Strobe B: xenon, 20 µs pulse, 45° right of camera axis, 2.0 m from dummy
Fill:     LED panel, continuous, 1.0 m from dummy, 30% power
Controller: delay 8 µs after frame sync, width 20 µs, burst limit 2 s
Cooling:  forced air 5 m/s across each strobe housing

Verification is done before the live test. Fire a single frame with the dummy stationary and inspect the histogram: the face ROI should sit between 60% and 80% of full scale, with no clipped highlights on the vinyl skin. Then run a dry burst at the target frame rate and check for flicker by comparing the mean luminance of consecutive frames — variation above 5% indicates pulse jitter or thermal derating. Finally, check the shadow side: if the fill is too strong, the image flattens and depth cues are lost; if too weak, the far side of the dummy falls below 40% and detail is unrecoverable.

Common failure symptoms and their causes: a bright horizontal band across the frame indicates a rolling-shutter camera with a pulse shorter than readout; uneven brightness left-to-right usually means one strobe fired late or not at all; a dark frame in an otherwise correct sequence points to a missed trigger, often from a loose sync cable or a controller that dropped a pulse under thermal load. Each of these is diagnosed by inspecting a single frame and the timing log, not by adjusting exposure blindly.