Michael255,

I think you are referring to clearance between the impeller and the rear cover plate (seal chamber).  Changes to this clearance will not result in significant changes to pump performance.  However, increasing clearance can result in increased pressure at the seal chamber if wear rings are not provided on the discharge side (seal chamber side) of the impeller and subsequent higher pressure differential across the seal.

Increasing clearance on the suction side, is much more of a performance problem.  In this case, recirculation flow from the impeller discharge back to the pump suction results in significant reduction in the performance.  In the case of the discharge side of the impeller, the pump discharge pressure will be observed across the discharge side of the impeller. This does not result in any recirculation and therefore no loss in performance.

If the impeller is shifted beyond a reasonable match between the impeller and diffusers or volute cutwater, performance could also be affected.

Difficult subject, few people have had the liberty to try variations and re-test pumps.  Another complication is that the importance or unimportance of this clearance will be highly variable depending on pump type due to different hydraulic characteristics.

I believe Kawartha is correct.  Actual tolerances between impeller and seal plate are not critical.  I am not as sure as Kawartha that clearances between the impeller and the front plate or suction side are critical, I think those clearances may be uncritical also.

Kawartha is exactly correct that cutwater clearance to the impeller is of more concern, and furthermore, is the cutwater in line with the impeller if the impeller postion in the volute has changed?

Richard Neff
Irrigation Craft

PUMPDESIGNER has raised a question which I think warrants a little additional explanation.  Any clearance between the impeller and suction casing, or between the impeller and casing wear rings, which is greater than the pump design clearance: results in excess recirculation.  Even with as new condition clearances, there is recirculation but this is critically magnified with increased clearance.  Keep in mind that the higher pressure liquid on discharge from the impeller will flow towards the lowest pressure.  If there is no restriction between the impeller discharge and the impeller suction, all flow will recirculate.  Therefore pump designers try to minimize the recirculation flow by various methods, such as minimal clearances between the rotating and stationary parts, while ensuring that contact does not occur.  

Restriction is created by one or two methods depending on the impeller design. With open impellers (no shroud on suction side of impeller), the clearance between the casing and impeller is important to pump performance.  With enclosed impellers (shrouded both sides of vanes), the clearance between the impeller wear ring and the suction casing wear rings is crucial for performance.

Wear of recirculation control is the main reason for falling off performance of centrifugal pumps.

Carrying the results of wear to the next level leads to the creation of recessed impeller designs which create a forced vortex to pump the liquid.  These pumps are far less efficient, with much reduced flows and heads, compared to equivalent pumps with recirculation control.

Thanks to Kawartha & PumpDesigner for your comments.

I am really after any specific quantitative studies that have measured the drop off in performance against increasing clearance.

I am interested in the 'suction side' ie short circuiting from high pressure to low pressure, but also in any effects on the 'discharge side' We have done some work on the effect of increasing suction side clearance in horiz. split double suction WORTHINGTON type pumps and have found a dramatic drop off in performance (initial clearance 0.024" increasing to 0.100"). We have no comparable testwork for WARMAN slurry pumps. As I said, WARMAN have undertaken to do the work at some time.

Mike

Michael255,

I think you'll have difficulty collecting the data which you seek because pump manufacturers obviously try to maximize their efficiency and curves.  They are not as concerned with pump performance when the wear rings are worn, other than to advise customers that correct clearances are necessary to provide design performance.  The other problem is that every pump would probably be different, so 1 pump test would not necessarily provide the answers that you seek.

I've not tried to calculate the recirculation flow with corresponding drop in head due to gap between the suction casing wear rings, but perhaps an order of magnitude could be informative, based on a calculation using an equivalent sized orifice.