Get a magnifying glass and look closely at the break itself. There will likely be two different surfaces visible. The part that looks like sandpaper (uniform rough crystalline texture) is the portion that failed suddenly and literally tore apart. The other surface (something less than half of the cross-section area, I will guess) ought to have "beach marks" - concentric lines, smoother than the big failure. These are the result of a crack working its way across the casting and point at the source of the entire failure process.
There may be a stress riser of some kind (nick, casting imperfection, whatever) on the outer surface of the casting where the smallest of the beach marks is. You might also be able to guess, from the edge where they start, which way the casting was overloaded (from the guardrail hit, for example).
With stock pieces, I don't know if it makes economic sense to Magnaflux uprights - particularly since they will bend in a hit before breaking. This should be a clue big enough to catch, right? Bending = exceeded yield strength of the material = the beginning of complete failure.
Dye penetrant testing should find them too but, again, if you factor in labor, it just might not be worth it. If you ARE going to go the the effort to get the parts clean enough to Zyglow, you may well be able to see the crack with a lens...
Remember too, when inspection shows a crack, you still have to decide whether it is sufficient to warrant replacement. We sent all of the uprights and suspension bits off of a Ralt RT5 to be tested once and EVERY SINGLE ONE came back tagged. The inspector told us that we didn't really need to worry about any of them, however, since we weren't flying them.
Kirk
Fortunately I did not make it to 12 before the failure. Is there a recommended replacement period for this part?