Originally Posted by
OnePerRev
......So it comes back to the gearbox. The first report showed two pinion fracture surfaces, one looks darker than the other. Also, the smearing to the integral race is discolored differently than the rest. The lab guys are no doubt evaluating relative amounts of corrosion, and discoloration. Don't be surprised if this box had corrosion internally prior to the crash.
It seems quite possible that it may have started in the gearbox epicyclic stage based on the public source information I have seen. I still don't understand why debris from the planet bearing outer race spall was not picked up by the lube oil chip detector. A race surface spall large enough to initiate a rim fracture should have generated a significant amount of ferrous debris. I took another look at the oil flow path from the output stage planet gear spherical roller bearing down to the sump. In the sketch attached I traced the oil flow path in green. One thing that caught my attention is the oil tray just below the first stage planetary carrier. It forms a ledge around the inner perimeter of the housing, and based on the inboard location of the drain opening in the tray it seems that. Due to the close proximity of the rotating carrier surface just above, windage will impart a significant swirling motion to the return oil draining onto the tray, causing it to pile up around the housing wall. The high velocity swirl motion in this return oil will cause denser ferrous debris particles to separate and collect on the inner housing wall and upper tray surfaces. This type of
dynamic oil debris separation is quite effective and is actually used in many aircraft lube oil systems. One thing that would be interesting to see is if there was significant amounts of ferrous debris trapped in this location around the upper perimeter of the oil tray.
As always, just pure speculation from someone (me) with no detailed knowledge of this gearbox design.