I could be absolutely wrong, but I don't expect it to fail immediately. I believe it would likely be something that happens over time so I'm trying to see how durable this is and I like to know where it would tend to fail first so I know what area would benefit from additional reinforcement.
Yes, I completely agree. That was precisely my point regarding the definition of work/fail. The question would be better phrased by taking a pool on the number of duty cycles it takes before a caster fails.
PLA is surprisingly strong with a tensile strength of ~7000psi. It just has a very short transition from elastic deformation, to plastic, to failure. IE, its brittle. I'd take a wild guess and say its in the 100s of cycles before something might happen. But, that's just a wild guess without knowing the actual load on each part, in order to get an idea of what region you're operating in.
From your rendering I can't tell how thin the material is in the caster mounting pocket. But, my money is on a failure around those lower 3 M6 bolts. The plastic material is taking all the tensile load there through the counterbores....with no support from any of the surrounding aluminum structures (either as compression or friction).
Still, I think it will take more than 100 cycles to see one fail. Not 1, 2, 5 or 10.