Forensics: Brake-light bulbs at the scene of an accident

I was reading an article in the paper about dangerous railroad crossings. Here’s an interesting quotation regarding one Mr. Armstrong, whose car was hit by a train:

So, how could the light bulbs, after an accident, indicate whether they were on at the time? Surely they do not retain heat for very long after the car’s been totalled. Do bulbs break differently when they’re on vs. when they’re off?

It’s all explained here

      • To put it simply, tungsten filament is brittle when cold, but flexible when hot. So if the brake lights were on (because the brakes were being applied), the brake lights should still have their filaments intact.
  • This is also a clue investigators search for in light-aircraft wrecks. Of course now that more and more lighting is moving to LED, this will not be an option in the future–particularly for auto brake lights, which on new cars are already LED’s (due to their longer lifespans in use).
    ~

I once saw an electronic microscope picture of a broken filament from a broken brake light bulb, they knew it broke while it was on because the broken end had a little glob of metal instead of a sharp cut. I guess that when the filament breaks, it starts with a crack, as the crack grows larger the cross section of the filament at that spot decreases, thus the electric resistance and the temperature rises, and before breaking completely part of the filament melts and forms that glob of metal.

Fascinating, thanks all.

Forensics is cool. I’d totally be a forensic investigator if I wasn’t so afraid of zombies.

True, but now that every car has an ECU (i.e. an on-board computer) they’re starting to design them to also be the equivalent of a flight data recorder. Info like speed, brake & gas pedal position, acceleration/deceleration rates, even the stereo volume right before an accident will be able to be retrieved.