impeller upper shroud contact

[CED Engineer]

First off, we are changing the design the remove the contact, but are looking at ones already out and about that could have this issue.

I am trying to calculate the worst case scenario of the impeller upper shroud on a closed impeller impacting the pump casing ring. The drawing clearance is 0.010” between impeller and casing wear ring, but based on a tolerance study calc recently completed, the impeller has a worst case droop of 0.020” at that location. We haven’t had issues with the design as the fit is tightly restricted by the assembly procedure, but recently we took one pump apart for an unrelated mechanical seal leak and found that the impeller shroud right at the suction side had been chewed up pretty heavily. As we are going through the root cause process, one of the things I wanted to try and calculate was, based on the known interference due to the calculated worst case droop 0.020”, the force at which the shroud could impact the casing wear ring. First time I have had to approach this topic and am looking for a little bit of guidance. The way I have pictured this is an off center installation causing the impeller shroud to hit the casing ring at the 6 o'clock location (horizontal centrifugal pump). Absent of shaft stiffness, is this more a function of pump speed and impeller weight?

 

[JJPellin]

I don’t know of any good way to estimate the contact force from a wear ring rub as you describe. But, first I should clarify some of your terminology. You use the term shroud contact but seem to be describing a wear ring rub. I am going to assume you are describing a radial rub on the eye-side wear ring of a horizontal centrifugal pump with a fully enclosed impeller.

There are at least four items that could be relevant in understanding this rub. You mention the issue of concentricity. If the stack-up of assembly fits allows the impeller to be installed off-center within the casing, this can contribute to a rub. This is associated with the concentricity of the fits between the bearing housing and the head and between the head and the case. A rub based entirely on poor concentricity would tend to show damage all the way around the impeller ring but at only on location on the case ring.

The second issue is related to run-out on the impeller ring. This is associated with how straight the shaft is and the fits between the impeller and the shaft and between the bearings and the bearing housing. A rub associated with high run-out would tend to show damage at one point on the diameter of the impeller ring but all the way around the case ring.

There are two forces that are important to understand relative to the potential for a rub. One is the hydraulic force pushing the impeller off center. This is associated with the design of the pump (single volute or double volute) and where it is operating on the curve. There are tools that can be used to estimate this force. The deflection resulting from this force can be estimated based on the design and materials of construction of the shaft.

The second force that is relevant is the Lomakin effect which is always pushing the impeller back toward the center of the wear ring clearance. This force is dependent on the wear ring geometry, the fluid properties and the differential pressure across the wear ring.

I am not sure what you mean by a worst case droop. It sounds like this is a combination of non-concentricity and run-out. We normally don’t estimate worst cases for these issues. The run-out on the impeller side ring can be directly measured on the assembled pump (before installation into the casing). This run-out should be less than 0.005” for most pumps. The concentricity from the impeller ring to the casing pilot fit can also be directly measured. Knowing these, plus the clearance from the head pilot fit diameter to the case pilot fit bore can provide good assurance that you won’t have a rub. But, the final proof is on assembly. If the assembled pump rolls smoothly with no detectable rub, it will probably not rub in operation given the Lomakin effect helping to keep the rings centered relative to one another. If the Lomakin effect is small (low differential pressure) and the hydraulic force is high (single volute operating far from best efficiency point), then a rub could still occur. The chance of a rub is also greatly increased if the pump could be run dry. This basically removes the Lomakin effect entirely.

If you could provide more details about your pump and the service it is operating in, we would be able to provide better answers to your questions.


Johnny Pellin

 

[Artisi]

A few drawing or photographs wouldn't hurt the discussion.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[CED Engineer]

Thanks for the response, yes I was referring to the wear ring region...
The impeller and casing ring both have run out requirements below .005. I would imagine that a leading cause would be during startup and shutdown of the unit when flow through the pump is limited thus removing any water centering assistance. We haven't had the same issue on prior units, and this one had a decent amount of run time, then had a routine maintenance evolution where I believe the impeller wear ring hub had a skim cut get to a machined finish and the casing ring was replaced with one that maintains the drawing clearances... after the maintenance was completed, it was a short amount of time until the wear ring hub was in pieces.