centrifugal pump impeller

[wellman]

I work for a company that has copied a few different impellers. When testing these impellers we have had mixed results. Capacity and head we have good luck. Generally we produce more head at shutoff and less capacity at open discharge. In the middle of the curve at general operatong conditions for each model we have similar similar between both impellers. Efficiency we are consistently low 1-2 points and BEP is shifted slightly to the left. After reviewing the new impeller vs. the old I suspect that the vane geometry itself is very close. I think the problem is in the oreintation of the vane in the impeller. The vane on the outside of the impeller appears thicker. I think the when we machine the impellers to diameter we machine off a little of the "side" of the vane and not stricly the end of the vane. So basically the entering and exiting angle of the vane is wrong. I dont know if the problem is in the pattern equipmrnt or the foundry procedure. If in the pattern how might I correct it without modifing the pattern or is that the only option. Any suggestions or sources for impeller vane design. These are vertical turbine multistage pumps. Open and Closed impellers.

Thanks

 

[vanstoja]

With your description of a higher shutoff head and lower runout head plus efficiency reduction and bep shift to lower flow, I suspect that your new impellers may be seeing slightly different inlet flow conditions in the test loop. Moderate inlet coswirl or counterswirl instead of zero prerotation might account for such performance shifts. Are you using the same test loop and inlet piping arrangement for the new impellers? Machining the discharge diameter of the new impellers doesn't strike me as a contributing factor to performance variations though significant overfiling or underfiling of blade trailing edges can appreciably alter total dynamic head.

 

[EdStainless]

How many stages?
What overall diameter? (housing diameter)
I ask about size because in smaller pumps this can be a bear of an issue.

I think that the tip, both inlet and outlet, may be part of this. Have you put both the new and old parts on the CMM and checked out the differences?

What about diffusers? Same ones?

= = = = = = = = = = = = = = = = = = = =
Corrosion never sleeps, but it can be managed.

http://www.trenttube.com/Trent/tech_form.htm

 

[wellman]

The impellers in question are a 12in nominal bowl assmebly.
I have tested a 2 stage and a three stage with results about the same. I have tested the impellers in the same bowls and each performs differently. I have tested each assembly s0 that the only difference in the entire setup is the impellers so I feel confident that the problem is in the impellers. There are two models that I have trouble with. One model I do underfile and one model I don't but I have similar results in both models.

Thanks

 

[Artisi]

are you testing against an original unused impeller in the same test configuaration or comparing your pirated part to a certified performance curve.




Naresuan University
Phitsanulok
Thailand

 

[wellman]

We are testing the a new unused impeller against the pirated impeller in the same test configuration. The only difference is the impellers.

 

[Artisi]

Before looking into blade positioning,thicknesses, test equipment etc etc, you should be aware that there are many factors which can give you 1 - 2% shift in efficiency. Even OEM impellers, with the best manufacturing techniques, QA, etc will have a range of variations in performances of this magnitude - even greater in many cases, this is the main reason that you have a tolerance for performance in pump test codes.

A manufacturer I worked for some years back when the ISO test code was reworked slightly to become the Australian pump test code, compared 20 -30 individual prior test results to see how the "published curve" for a particular pump model compared / performed in accordance with the code. What was interesting other than the result of this study was the scatter of points and the variation in flow, head, effic, between individual pumps. All came within the test tolerance with some at the minimum and some at the maximum of the tolerance range.

What would be of interest is - how many OEM impellers have you tested and how many pirated ones have been tested under the same conditons.



Naresuan University
Phitsanulok
Thailand