My shop is an old building from the 30’s that was a service station. The front showroom has four windows 6 feet tall and 8 or so feet across. Over the years people driving too fast have kicked up rocks from the driveway. I have three of these spalls.
Drives me nuts, but not nuts enough to spend the hundreds it would take to replace them. Someday, I hope.
Dictionary says glazier.
The technical term for this phenomenon is “spallation”. The “edge of the disturbance” you speak of as the fracture point is actually where the rarefaction wave (the portion of the wave directly behind the shock* wave, a decompression wave) meets the reflected shock wave. The angle of the cone varies depending on parameters such as density, bulk sound speed, yield strength, and speed of the original wave. In concrete, for a projectile traveling just under Mach 1, the waves will form a maximum about 15% of the materials thickness from the back surface.
A couple other terms, while we’re at it. A projectile is said to “penetrate” a target if it comes to rest in the target, while it “perforates” the target if it goes clean through it. Pretty impressive to watch a projectile weighing a ton, with a foot and a half diameter, punching through three slabs of concrete, each two foot thick.
*Shock wave means it’s traveling faster than the speed of sound. No shock, no rarefaction. However, I believe the same phenomenon takes place at subsonic projectile speeds, although the cone will be more shallow.
That’s the speed of sound in the medium, right? The speed of sound in glass is going to be different (higher, I’m guessing) than the speed of sound in the air surrounding the glass.
when this type of scene is shot in a movie, would they produce the glass with hole by actually shooting through it?
CalMeacham, that…was…awesome. 
Right, sorry, I was thinking of detonation behavior. In a non energetic material, the sound pressure wave will propagate to the rear surface, reflect, and meet with the projectile coming from the other direction. If the projectile was traveling at near Mach (through the air), it will only have traveled 15% (I switched it in my above example!) of the way through the concrete before the sound wave (which, as you say, is faster) has already reflected off the back and come back to meet the projectile. Therefore, the sound wave traveled 185% of the thickness of the material in the same amount of time.
The speed of sound was generally around 350 m/s during testing (dependent on temperature, density, humidity), while the bulk speed of sound in concrete is about 3,100 m/s. In this case, the projectile would travel 2*350/3100 (see below), or about 22 percent of the way through the material before it meets the reflected wave. In glass, with a bulk speed of 5500, the number will be closer to 13%.
The reason it can be closely modeled as 2*projectile speed/bulk sound speed is because when the wave just hits the rear surface, the projectile has completed the first part of the ratio. It will then complete the same ratio. Some simple algebra will yield the exact answer, but as projectile speeds decline, they become closer to the gross simpilization.