I haven't seen this failure mode of these servos previously... Is it possible the elevator halve in question was exposed to a shock load, i.e., possibly bumping the elevator into something otherwise experienced as hangar rash? It’s plausible the failure is simply an indication of flight loads exceeding the ability of the servo, but I find that hard to believe as these servos are well proven. However shock loads induced to the surface and transferred to the servo have been known to cause this failure mode, typically realized after crash damage.

We have had psoriatic notices of the m/p output gear failing of this servo, but the gear axle pin boss failure unknown to me.