Windshield Cracks Hold Secrets of Impact
physics.aps.org
physics.aps.org
Of course, after the initial maiden impact nothing else will ever happen for the duration of the ownership of the car.
This is probably somehow related to Murphy's law. The secret to these impacts seems to be that I'm behind the wheel and on any highway at all. I've contemplated not driving the car for 2 weeks after buying but I'd hate to admit that I'm getting superstitious about this.
The thing that struck me (no pun intended) is that Plexiglas is a polymer and thus has a more-or-less regular repeating internal structure, while glass is amorphous, with little internal structure. So comparing the crack structure between them after similar impacts probably wouldn't give you the same information about the energy input. Which might be the point, but it appears they're comparing apples & oranges, so lessons learned in glass vs. lessons learned with polymers wouldn't be transferable.
Also - automobile windshield glass is laminated (to prevent shards of broken glass flying around the inside of your car), and there are different materials used for the laminating plastic & adhesive, depending on the "recipe" chosen by the manufacturer. So the crack structure would be different between them, and again, I doubt the lessons are transferable.
Different specific windshield designs will, of course, vary. But this is a good starting point across designs.
Now, if they said that their formula was consistent within a certain class of materials & construction, I'd buy it. But even then with the Plexiglas, their formula predicted 7 cracks, while the experiment produced 8. They need to explain that, and I'm saying it's because of the near-crystalline internal structure of Plexiglas. If they had altered their hypothesis & and prediction, and then retested, and did some analysis to explain it (perhaps I'm wrong - maybe its because of some other factor and not the internal structure) I'd buy it.
Getting 7 cracks versus 8 in the experiment is actually a pretty good fit. Particularly since repeating the experiment multiple times would be unlikely to give the exact same number of cracks. (The cracking pattern does have a lot of randomness in it.)
What is more important is that if the sqrt scaling rule proves reasonably consistent across a wide variety of types of materials, then testing the cracking pattern for different kinds of impacts across different kinds of windshields becomes a much more tractable problem.
The obvious followup becomes how angle of impact matters. My hope would be that it is something reasonably simple - such as very little is deposited by friction, so it is only the component of the velocity normal to the surface that matters.
Establish those two things reasonably well, and now you're in a good position to measure every kind of windshield a handful of times, and you have some useful information for forensics on all kinds of windshield cracks.
However, given a glass which produced n cracks at an unknown speed, they can take an – ideally – identical glass, hit it at a known speed, count the number of cracks and infer the previously-unknown speed. If the square root law is indeed as simple as I imagine (I didn’t read the paper), they will only need one data point to estimate the unknown speed, though more are obviously helpful. Over time, one could easily build up a database of materials, since there are only so many different cars, that shouldn’t be too difficult.
Another fun fact is that any day you have a small chip or crack, it can instantly turn into a huge crack that is unrepairable. So if you have a small chip or crack get it repaired right away. Check with your insurance company - it will probably cost you nothing.