Hi all, when I discovered mine failed I brought it to the materials lab at work and got it analysed. It failed from stress corrosion cracking as a result of the preload hoop stress generated by the tapered spacer as James says. The alu they have used is very brittle so they evidently didn't want to clamp across the lugs, and instead used this rather crude expanding collet solution. To be honest, it is just a horrible design…! SCC causes tiny pits on the surface of the component where sustained stress is experienced (effect is vastly increased where permanent strain is induced - i.e. material is taken beyond yield – in this case if someone has done it up without a torque wrench because ‘they’ve been tightening bolts since you were in nappies’ and therefore know better!). This isn’t necessarily the same as the kind of corrosion that you’d imagine on steel etc, it can be barely visible to the naked eye and just cause miniscule surface pitting. Not saying the two aren’t linked, but they aren’t necessarily the same hence I agree with Simon, I don’t think it is necessarily the steel corroding that has caused the bracket to fail, I think the bracket has done that on its own. That being the case, it appears that time under load is more important in this case than the actual load seen by the bracket, that just polished off the job, and of 90% of the tails I’ve heard, the force required to finally ping it was small (I broke one of mine open with a screwdriver…). That would also correlate well with the fact that we are now seeing a raft of these failures (I’m now aware of double figures – I think someone was trying to keep a log somewhere, but can’t remember where).