Originally Posted by
JohnDixson
Thanks, Riff-Raff. A previous poster cited an " ultimate " life for the gears at 5000 hours. I assume he meant the fatigue life or Component Replacement Time (CRT ). Can one assume that the gears are exposed to accelerated load testing, multiple samples, a three sigma curve reduction etc? Assuming this or a similarly rigorous procedure to support the 5000 hour life is in place, there must be some hard work going on at AH to discover what resulted in two fatigue failures well within the CRT for the part. Unless I missed it, there hasn't been any information at all regarding the failure surface striation count, i.e., how long the crack existed prior to failure. There are a couple of posts on this thread that may/may not point toward a factor affecting the gear loads, but I'd guess that by now, the AH transmission people are way beyond that sort of speculation.
John,
Don't know about the published CRT as others are posting about, but we can assume that the fatigue approach is NOT as you say for gears. It is close, however. The standards have not changed much in this area, in general it is a 140% overtorque test that is conducted in a rig. This equates to just under 3 sigma for steel. The load can be reduced further for multiple specimens. It is intended for tooth bending loads. Not sure what was done on this box, how much credit if any was taken for legacy assemblies. Nor do we know what other loads are considered. In the pinion, the cracks shown are not tooth bending, but initiate at the bearing integral race (our guess based on the released photos).
I concur that AH engineers are scrambling, regardless of what the press releases are saying.. in addition to experts from their transmission source, as I understand that it is completely outsourced, including design.
On the loads, Also agree as you suggest that there may be an anomaly, a previously not understood load source affecting the epicyclic, beyond that which it was certified to.