Originally Posted by
OnePerRev
Yes Pablo, but more specific, the sun mesh drives one flank and the ring drives on the other side, whereas in a single mesh you only have loading typically on one side of a tooth. On top of that, as a tooth is loaded it causes the root behind it to load in tension. When the next tooth is loaded the same root location is reversed into compression. When this same root is exposed to mesh on the ring gear, the opposite occurs. The first stress is compression, then tension. Some argue that this t-c-c-t is really a unique application of the goodman correction, and could be treated like a single cycle. Further complicating is that the exact root location of peak stress for each mesh may not and probably is not, the same, nor is the t and c magnitude the same, depending on deflection and other factors....
.. but this digresses. The fatigue in this case is starting at the bearing race ID, so more like a rolling wheel fatigue than gear tooth problem I would venture to guess.
OPR,
That's a good explanation. Basically, a simple epicyclic planet gear is similar to an idler gear. With regards to tooth root tensile fatigue stress due to bending, AGMA recommends a material knock down factor of 0.70 for teeth subject to the type of alternating reverse load cycles experienced by planet or idler gears.
Another thing to consider is that aircraft gears normally have the tooth root fillet surfaces carefully shot peened. This creates a compressive pre-stress in the surface. Since fractures in the surface initiate/propagate from excessive tensile stress, any residual compressive stress in the surface must be relieved before the surface is stressed in tension. Shot peening, when properly applied, can be very effective at improving tooth bending fatigue performance.