A wild guess. I'm assuming that the tail shaft was originally designed adequately for the application.
If the break occurs at the inner edge of the pulley, you might consider fretting corrosion. If the pully is not tight enough on the shaft, micro movement of the pulley on the shaft can cause fretting corrosion which creates stress risers and finally a complete failure of the shaft from fatigue. This can happen unbelievable quickly.
metalguy.
yes its where the bending load would be highest. another thing the shaft is supported by bronze bushing fitted to the slide plate.this slide plate is for tensionig the belt.the shaft is also supported by bearing on each side thy are vertical mounted and is also fitted on the side plate.when ever the shaft fails the bushings also fails.the thing is the bushings is always worn at the top.
Stupid question but you say this has happened before, are we sure that the shaft in use now is the same chemistry/heat treatment as the original and hasn't been substituted by a maintenence mechanic somewhere along the way? Clearly the mild steel is not up to the task at hand. I'm having a hard time visualizing the mechanism from your description so I'm unclear as to the significance of the bushing wear you mention. I'd try substituting 4130 or 4140 heat treated for the mild steel shaft and see if that improves things.
