Gentlemen,
I am getting ready to help balance a centrifugal pump impeller at the shop tomorrow. I have never balanced one on a stand before, so I have to ask, Where do you put weight? Doesn't look like any place I can see. If you put them on the top or bottom of the impeller, I'm afraid it would rub the volute, maybe?
I'm brain dead at the present time, so be gentle.
Anyway, turn an old man on to some new tricks.

Dave

Hi Dave,

I guess that you'll have to remove material opposite where your balancer indicates to add some.
It's usually done by grinding some material off the tips of the blade and blending the contour as best you can.

If you're lucky, you'll hit the needed balancing correction smack on one blade and if not, then you will have to use polar graph paper and plot the required amount of material to be removed over the adjacent blades.

Now a problem will arise !
How much material should you remove ?
Your balancing m/c will tell you how much to add or remove. Adding is easy since you can weight it, removing material by grinding off will have to be done a little at a time, else you're likely to overshoot your correction.

I do not know of other ways. Other members of this forum may suggest adding fast curing epoxy or other glued on material.

If you do that and there is a lot of material to add, you may interfere with the flow of whatever the impeller is pumping.
That may create pulsations and you will be faced with another type of problem later.

Also that material can be eroded away with time.

So... get your hand grinder ready and good luck !

MarkoLeo

P.S. Use wax for trial weights if so.
Aircraft engine impellers have a small rim on either side near the hub where one can add small rivets for corrections, but I suppose that it is not the case with your impeller.

I have stand balanced pump impellers, turbine wheels and blower impellers (Hoffman and Lamson). All weight corrections were made by "removing" mass by grinding. A word of caution though, I always checked the computed "correction weight" by adding mass using balancing putty. If the results proved satisfactory, then I proceeded with "removal" 180 deg. opposite. However, nice and easy does it, remove less than you think is required and spin the rotor again to ensure you are on the right path. Your phase angle should remain close to the same (within say 10 deg.)amplitude should drop. Now, you also do not want to concentrate too much mass removal from a small area as to thin the impeller wall excessively, you may want to spread it out a bit. Measure twice and cut once so to speak.

Your final mass change will be achieved by grinding..but as already pointed out check twice, grind once.

When the balance machine gives up its results, check them by adding the required mass of blu-tack (or similar) held in place with duct tape (its not going too fast so this should be just fine) - when certain that you have the correct result (principally location) - remove blue tac and grindoff in steps 180 degrees opposite

Dave,
Some excellent replies already, especially about confirm with trial weight before grinding. You can also remove weight by grinding in the area between the blades and the wear rings. This is the area I have been taught to use because it is outside of the flow and less likely to cause problems. Usually the wall is thick in this area and as suggested earlier, spread the ground area out to avoid creating a thin spot.

Ronnie

I would suggest not to grind on the vanes or blades! Grind on the side removing small amounts at a time. You can center-punch points to a certain depth and keep track of how much you're taking off. But don't grind where it'll affect hydraulic balance.

Gentlemen,

Thank you all very much. I was afraid I would have to remove weight. After sandblasting, this thing don't look any too healthy. It's pumped a lolt liquid in it's day.
I'll figure something out.

THanks again!

Dave

How bad worn is the impeller?

May consider using belzona material to resurface impeller if a new one cannot be purchased. We have done this on river pump impellers before and belzona has material for this that works well.
http://www.belzona.com/

From what you have said, I might possibly try adding weight to balance, if you can lightly tack weld or tig weld the weight in final location.
Then coat entire surface with belzona material suited for resurfacing impellers.

Then test run again, could likely lightly grind belzona material in needed location to produce balance results for running.

my 2 cents.
Mike

Mike makes a good point. This enhances the thread the other day on balance weights... The belzona material comes in bronze, and a number of metals - pretty good glue.

Yes Sam & Mike,

Belzona makes some good materials. We used to patch holes in conveyor blets with their rubber. Have also used their metals. Hadn't thought about that. I have to wait on the customer now to see what they want to do. They had the impeller bored out and another shaft made somewhere. ANyway, there is 0.004" clearance between bore and shaft. ALso the keyway is 0.003" to big and key is sloppy. I told them I couldn't balance anything with this much looseness. Even if I put it on another shaft and balanced it, it would run poorly on the setup they have now. He wasn't pleased with what we had to show him. Now we're waiting on what they want to do.
Dave

Dave

Gentlemen - I have a large number of customers in the pump balancing business and no one balances the impeller twice (once with mud and once by grinding) unless they are having a problem with the machine (rare). If your machine is working properly, put it in the "remove" mode and go to grinding - but do not do it on the blades, please. Grinding on the outside (or backside in the case of an open face impeller) is much easier, both from a plane separation standpoint as well as obtaining some "meat" to grind on.

Earl, I'm sorry but I'll get right to the point - going directly to grinding IS BAD ADVISE, I don't care what kind of balance machine you sell. It takes no more that a couple of extra minutes, if that to check your run by adding weight and then grinding 180 degrees off. Remember, you can't put back what you took off if there was an error either human or machine error.

If the rotor is a 'fist tme' deal then the influence coff' method may be best w/attached trail wtgs. But on rotors in a balancing machine where stored in memory and recalled a trim and going directly to metal removal may be perferred. The trail wgt step is not necessary.

Yvan - I do not want to get into a contest, but I really do not think anyone who is experienced in shop balancing thinks my advice is "bad" about not mudding in a pump impeller before grinding. First of all, I did not say I was selling machines to these people - I consult, train, calibrate etc. for over 200 of these type of shops. The machines range from IRD, Schenck, BalanceMaster, Dynamics Research, Hofmann, etc., etc. And, yes, I do sell occasionally. But back to adding mud - you only are proving that the machine will not lie to you by balancing with mud. With a grinder in your hand you are also proving the same thing. Grind carefully in a controlled area on the part and re-spin it - if the angle moves considerably or the amount does not decrease, THEN, perhaps you should dig out the balancing wax and see if you have a problem. But please, after 35 years of doing this, don't tell me I give my customers "bad advice". Thanks.

What I was referring to when I mentioned using the belzona, was a material used to repair a worn out impeller, not mud used to balance the impeller with.

My thoughts was if it was balanced before totally recoating the impeller with belzona it may be easier to balance as long as the nearly the same thickness coat of belzona was applied.

Just an opinion,sounds like Dave is dealing with a customer that cant spend the bucks to buy a new impeller and needs a way to get this one back in operation.

2 cents was all it was.

Mike

Earl, you state:

quote:

Grind carefully in a controlled area on the part and re-spin it - if the angle moves considerably or the amount does not decrease, THEN, perhaps you should dig out the balancing wax and see if you have a problem.

So, why not take the cautious approach and check your results before grinding. If you for some reason find that your grinding does not provide the required results, how do you correct this error? - by grinding an equal amount on the opposing side. What I propose merely takes maybe two minutes extra, small investment I think to avoid a wheel that looks like a piece of swiss cheese.

Oh!...and please read Dave's post, he qualifies his limited experience, would it not be "better" advise to have him take the cautious approach?

I'm with Yvan on this one.

I don't care how many years experience the balance shop has got (that's why we go there in the first place) but...anyone balancing my compressor impellers has to get it 100% right first time and everytime - hence the use of blu-tac to prove the results of the balancing machine and the operator.

These compressors generate huge amounts of money and cost a huge amount of money to overhaul - every step of the process is monitored and documented. A couple of minutes to prove the balancing procedure is nothing.

I really can't see the time saving between a 'test grind' and the addition of a temporary proving mass

Two comments:
Earl, let us face it, the whole world is here and we are all different and in very variable situations. Nobody thinks your decades in balancing was in vain. You know your trade, but anyone, including myself (since 1965 in balancing), having to build confidence now, MUST by necessity build it her- or himself. All must take small selfproving steps when risking clients investments. You just have to "muddfill" or use temporary weights until you are dead sure you can grind/drill. If you Earl feel confident in your machine to do a direct cut, fine, but I personally would not be boasting with it. However, I say, live and let live, I personally take pride in trying to be extra cautious. Experience is the sum of all not-yet-lethal accidents....

My other comment is that I must have missed that nobody speaks about the precision of the arbor shaft. When balancing a pump rotor alone on a "help shaft" I always turn the rotor 180 degrees to verify that the runout of the arbor is small enough (we call it index- or Umschlagstest). If it is present and I measure it, and it is expensive to remove, I can also use the vector results to balance the part of unbalance of the complete rotor that is only attributable to the pump impeller. Hence know that I will enter a perfect impeller in the pump, disregarding the arbor error.

Hi all,

When I proposed that material should be removed instead of being added, was for the simple reason that the original author of the post had a problem adding material to his impeller fan.

I have never, and this is from a bad personal experience, added some heat curing epoxy or other type of glued on material for a final correction on any rotating piece of equipment, unless balancing some rim accepting bolts and washers had been provided for that purpose.

Call me 'old fashion' if you'd like. You stick to your guns and I stick to mine.

My suggestion to add wax was at the onset of his balancing operation, thinking that he might be balancing on an IRD or any other balancing m/c that needed to be calibrated for plane separation prior to actual balancing.

Ok... plane separation calibration ( cross-over effect ) may not be necessary if one chooses to do a single plane on the back face of the fan or impeller. However, how sure does one know that the unbalance is concentrated on the back face ?

Who can one tell with certitude where the nodal point lies on a rotor ?

I've always performed a quick two plane separation test using wax as a trial weight medium just to ascertain that phenomenon.

And, to add on to what Arne Lindholm suggested on using a slave balancing arbor... he is quite correct about checking for run-out of the mounting diameter of the fan vs the journals of the arbor bearings. Personaly, I used to trim balanced the shaft first although it might have look absolutely symetrical, concentric and whatelse.

It takes very little time and it is a more professionnal way of doing things right the first time.

By the way, I hate it when someone discards with a swat of his hand... years of experience.

To just work with and put all your money on all the fancy equipment around that does all kinds of fancy tricks, wave form, ect. and gives you more data than you can deal with, in my opinion, will never but never replace years of on-hand experience and a common sense approach to a vibration problem.

The instruments are there to guide you... none will tell you exactly what you should do.
The final decision is yours and your degree of years of experience is your best allied.

MarkoLeo

Dave,

The only advice I didn't see mentioned is to blend the grind spot(s). Sharp edges can make some really odd noises when they install the pump.

Earl,

I respect your experience, and I do agree that most balance shop do NOT balance a rotor twice. I speak from the experience of my 7-1/2 years in a balance shop.

However, the situation presented is slightly different. You have a guy that does mostly "in situ" balancing performing shop balancing on a machine that is most likely used infrequently (at least by him). There is a large group of folks here that are primarily in the field, we've been in Dave's shoes and felt the confusion of trying to recall how to operate shop equipment. Many of us don't have the luxury of being a shop on a regular basis.

On the first rotor he balances, he should add a removable weight, so that he can be confident about the operation of the machine. Earl, I know your reputation well enough to know that you would do likewise if you were not familiar with the balance machine.

MarkoLeo,

"Technology will never replace common sense and good judgement" (good judgement comes from experience)

I'm curious, do you really see much multi-plane in pumps? I'm picturing single stage overhung pumps, since I've always had the convenience of seperating multi-stage pumps (boiler feedwaters) for balancing.