stress calculations on an impeller

[stopsunset]

Hello everyone,
First time I've posted an inquiry here although I've used it as a reference in the past. I have a double suction impeller with a shrink fitted wear ring that experienced a fatigue failure.I want to calculate the stresses on both the impeller hub and ring(both tangential and radial). I have the formula to calculate the static stresses induced by the shrink fit however I wanted to determine a way to find the stress as a function of speed on the rotating rings.Does anyone know what those formulas would be?
TIA

 

[Bjegovic]

what is the nature of dynamic loading, if any? it seems to me that dynamic component of the load on a wearing ring is small comparing to static load, in normal conditions, so fatigue may not be the main cause of failure.

 

[vanstoja]

Rotational stresses in a shrink-fitted wear ring on a motor driven pump impeller with a maximum rotational speed of 3600 RPM will be generally neglible and unlikely to cause a fatigue failure of the wear ring. A more likely cause of wear ring failure is surface rubbing contact which can result in very high "flash" temperatures which can cause thermal stresses or induce vibrational stresses by rubbing contact such as "stick-slip" action. The surface profile of the contacting surfaces are important, eg. labyrinth vs plain annular seal. Was the failure initiated in a notch in the labyrinth toothed surface of the ring? What are the possible radial thrust overload conditions and forces on the impeller that would cause wear ring contact? These are some of the questions that need to be answered to diagnose impeller wear ring failures.

 

[stopsunset]

thanks for the replies...I calculated the rotational stresses and they were indeed neglible compared to the yield stress of the material.There was no rubbing or galling of the pump internals.Let me clarify the failure..the wear ring did not fail rather it was the impeller hub where the ring fits onto.The hub, which is part of the impeller casting, simpy fractured near the vane the vane profile.Further inspection of the components revealed that the wear ring was shrunk on .012" onto the hub.The hub that didn't fail(remember it's a double suction impeller) had a metal to metal fit.I calculated the tangential stress due to the .012" shrink fit and it yielded a stress above the materials yield strength.
thanks again for the input

 

[Bjegovic]

vanstoja,
thanks for a thorough explanation. When you mention local thermal stresses due to rubbing, what is their contribution given that the ring is constantly cooled by pumped liquid? More precisely, is this only a surface effect (somehow simillar to pitting due to high concentrated surface loads) , or it affects entire profile of the wearing ring?

I can imagine that a very small layer of metal is cooled and heated with each pump revolution. Can this create an initial crack?