While I need to research this further, I believe a chamfer in the gear leg hole would help reduce the problems. For those interested, I dug up this
paper.
The situation being tested is reasonably similar to our application, and their data showed an improvement in shock load cycles to failure with the addition of a chamfer on one or both of the shear plates. The basic theory is that more energy can be absorbed by the bolt during shear loads.
This may or may not apply to our gear bolts (the same design applies to most gear legs in the RV series). I will continue to look into whether or not the chamfer helps or hurts.
As I believe I've written in these forums before, a very minimal preload is a good idea on these bolts. The nut's only real function is to prevent the bolt from falling out. Pre-load on these bolts is an additive stress to those encountered during shear loads. IF the bolt's preload was adequate to cause friction between the gear leg and mount, it would be a different story. My experience is that these bolts, torqued to spec., do not apply enough squeeze to the mount to create any friction.
It could perhaps be argued that some sort of a split housing might be better, as the bolt's primary function would change from shear to clamping (which would create friction to keep relative movement at zero).