Often it is not difficult to get within a factor of 2 for the magnitude of the balance weight. Put that in the correct location and you have no worse vibration than that with which you started. Put that in the wrong place and the vibration can be 3 times as high.

Both selections are important. If I have high vibration to start with (I guess that is why I am balancing.) then get the weight in the correct location. Even if you have the exact magnitude, if you put it in the wrong location the vibration can double - 2 times bad is worse or broken.


Economically, if the rotor was running with somewhat high vibration and you hit the correct angle, the machine can probably run or limp along until it can be taken care of. Hit the wrong angle, and the machine is not running. This is very important for critical machines on which the operation depends.

Furthermore, there are not 359 wrong places to put the weight just as there are not an infinite number of wrong magnitudes of balance weight and only one correct one. One is never going to get a perfectly balanced rotor; one needs to get it to acceptable vibration levels.

The consultant who comes in with the dart board approach to selecting the angle (or the weight) may be escorted out of some plants - at least those that care about their operations.

Jim's quote:

quote:

The AMOUNT seems to me to be far more important. I would prefer to make three runs with an insufficient amount of TW rather than one with a TW that is destructively large, regardless of where I put it.This is your opinion and I respect it.

Sorry for editting this here after originally posting this reply, but...... to get "3 runs" might take 3 days on some motors/systems which have a cool down time required before a restart is allowed or it takes several hours even to coastdown.

I think some have lost the true meaning of the original question. No one here on this post, from what I have read and understood, is saying the trial weight's weight is not important, just that it is irrelevant to estimating the trial weight's angular first position. These are 2 different, unrelated calculations.

I am sure everyone agrees that having a weight "10 times" more than what the original imbalance was, is ridiculous and I am sure anyone trying to balance something is not going to go that far intentionally. But again, the trial weight's weight is irrelevant to the original question, "where to put the first trial weight?".

Anyone who doesn't try to estimate a weight's weight and not be ridiculously too heavy, but use one which will give reasonable amount of change in amplitude and phase, doesn't need to be attempting to balance anything, in my opinion. But again, this estimation of the weight is not at all related to the angular position it will be placed in.

But let me ask this question..... How many times does a person go on a balance job, especially a contractor, and has the weight of the rotor in question given him/her so as to perform the calculation to give a reasonable weight? Not very many times is my guess and from my experience. One must learn to visually and mentally analyze the size of the rotor and estimate the trial weight if one does not have the rotor weight provided. Doing this via long distance is even more difficult. One can only use the data: width, diameter, type metal, thickness of the metal, etc., told him by someone from the sight.

After hundreds of balance jobs without any rotor weight provided, we, or at least some of us, develops the skill to do just this, estimate the trial weight size to give enough change without being overweight and without the known weight of the rotor.

The simplicity of determining this position makes it such a small time consumer that it is unreal.

Pete quote:

quote:

These may be relatively small benefits associated with carefully selecting the tw angle. And it is right to ask whether it's worth the extra complexity which might make us overlook something much more fundamental (the K.I.S.S. principle mentioned above).

As mentioned, the time it takes to calculate this position is so small, usually less than 1 minute, after obtaining the reference data and coastdown data, this leaves time to perform any calculations for the weight's weight, which should have done been calculated prior to the start of the job if one has the rotor weight, but which is not always the case. Only my opinion and I could be totally wrong.

Since the "voting" seems to have stopped and probably some "viewers" have become bored with the long winded responses on this the original thread, I will post, in a new thread, the actual "results and choices" from the case used in this thread .

Many thanks to all who have contributed their "views" and "procedures" and "votes".

This message has been edited. Last edited by: Ralph Stewart,

Ralph - fwiw I agree with you. I was trying to outline both sides and I think there are good comments on both sides. But I don't think anyone was long-winded. Certainly not me?

quote:

But I don't think anyone was long-winded. Certainly not me?

Certainly not you Pete.

"Long Winded" might not have been the best choice.

I know sometimes people have a tendency to not read the complete text of something if they don't understand what is going on or the words are to longgggggggggggggggg and they have to look up the meaning or the abbreviations are not clear, etc.

I probably am the long winded one if there is indeed a long winded one in this thread.

I enjoy your in-depth replies Pete. You put a lot of thought into what you say. Good to see that.

Everyone who replied to this did a good job and I appreciate it greatly.

Sorry Ralph for interjecting the issue of appropriate trial weight amount in balancing an industrial fan. As I believe this forum serves an educational purpose to those that maybe have never balanced a fan, I thought it was OK to include the topic in the discussion. I apologize for the disruption of commentary.

Nevertheless, with due respect to all that have forgotten more than I know about balancing, I am comfortable in believing that calculating the correct amount of trial weight is far more important than where it is one attaches the weight to the part to be balanced.

I was taught (at/by IRD folks) that the purpose of the ‘trial weight’ in a balancing procedure is to achieve either a 30% change in amplitude or phase. I utilized the IRD Model 245 with Model 544 velocity pickups equipped with 80 lb. pull magnetic attachment(s) and never had a problem with location of trial weight placement but was always careful in calculating the amount of trial weight based upon machine configuration of what it was I was balancing. I suppose I balanced between 90-110+ fans in varied industrial environments over about 12 years performing field services for IRD and never had one ‘explode’ doing it the way I was taught, dumb or not. I was never finished with a job before achieving at least .5 mils or less residual remaining unbalance (considering unbalance the overriding system problem). I balanced a fan in a foundry outside Chicago doing it the way I was taught that had vibration exceeding 100 mils to begin and managed to get it down to 9 mils in one shot (with a 6 pound correction weight and 3 pound trail on a 15 ft. diameter fan impeller about 6 to 7 ft. in length) the exact amount of residual vibration present on the third floor containing 12 other fans. It did require, however, the assistance of two maintenance associates to secure the pickups in place so they would not ‘walk-off’ the sides of the supporting pillow-block bearings. One could feel the vibration upon entering the building. I figured that with this motor-belt-sheave driven unit, the existing heavy spot on the impeller would take the motor for a walk and find its way to the bottom at shutdown, therefore my trail weight placement was at the top. I’m not sure if I was lucky or just employing common sense.

I trusted the instrument I used developed (firmware/balancing program) by one of, if not the brightest engineer I have ever had the privilege with which to work. John was perceptive enough to realize all those that used the 245 wouldn’t necessarily be the experts Mr. Foiles and Mr. Castleman are with respect to balancing techniques. I suppose if I employed ‘dumb’ methods/techniques to perform tasks I was assigned to do, in retrospect, I feel fortunate to have been so lucky during that 90-110+ balancing jobs I accomplished while a consultant for IRD Mechanalysis.

Mr. Foiles: I respect your posts submitted here on this forum. Every time I embark reading a post or reply you put I feel as if I can learn something from your experience and wisdom. If I owned a power plant, nuclear or fossil fuel, and needed a turbine generator of any size balanced (as to your reference to Pete) I would certainly ask for your quote to do the work then do what all power plants do: bring in the turbine generator OEM to do the job. As you so accurately put it ‘too much risk!’

Markoleo: I have been reading posts and replies to posts by Rusty since Mr. Petersen created this venue. Rusty is a very knowledgeable and experienced engineer and I, for one, appreciate his candidness. His ‘dumb’ reference was with respect to a concept I was taught by IRD which he no doubt has a fair amount of passion. I personally, took no offense to his statement. You are no doubt a quality individual and caring person with sensitivity towards others feelings, knowledgeable in our industry and obviously an asset to your organization.

Mr. Kilbey: I am quite distressed to hear of your colleagues’ misfortune. I would imagine his error was in the amount, not placement, of the trial weight. He no doubt failed to attend your CSI instructional session pertaining to ‘Machinery Safety’ (chapter pertaining to introducing trial weights to mid-high speed rotating equipment) that I am confident you must have taught while acting Training Director. IRD Mechanalysis conducted this training for all new Field Service Engineers joining the company, making them aware of hazards that may present themselves in the field. If IRD developed it, CSI eventually had it too.

Ralph, for your problem I would place a trial weight at 315 degrees and KNOW approximately how much based upon information I would obtain with respect to rotor configuration PRIOR to moving forward.

Rusty: Of all the ‘dumb’ concepts and Machine Condition Monitoring’ philosophy’s IRD Mechanalysis developed and indeed pioneered, at least you have a job doing things you are excellent at and enjoy and only have IRD to thank for creating the industry within most all of us now work. Had it not been for IRD and the ‘dumb’ things we did, you could still be milking cows on your grandfather’s farm in the heartland of Arkansas, working for Don’s former employer somewhere quite obscure rewinding electric motors or worse yet, as a VP of Engineering for a major fortune 500 company, dieing more quickly every day from the stress involved with the job! On second thought, scratch the VP role! I agree with Sam-stay away from wood and go with machined aluminum.

Ralph, one last question after your quote:

Sorry for editting this here after originally posting this reply, but...... to get "3 runs" might take 3 days on some motors/systems which have a cool down time required before a restart is allowed or it takes several hours even to coastdown.

Please read what you have stated above. I submit to you that even the ‘esteemed’ Mr. Foiles requires at least ‘3 runs’ to balance something he has never seen before (predicated on the technology I know and was taught)! An ‘original’ run, a ‘trial weight’ run, and a final balance run securing the correction weight. Has balancing technology advanced that much since I’ve been away? Can a balance job now be performed in two iterations? One (with no historic records)? I am always interested in learning new tricks (or old if I haven’t kept abreast with current events.

And talking about taking ‘risks’! Balancing something you have never seen long distance? Ralph, I hope you weren’t providing assistance to Mr. Kilbey’s colleague…And please let us know how many ‘runs’ in which the job was accomplished. Seeing as how 3 runs are unacceptable to you I would be curious to know.

Anybody balancing anything has a 50% chance of reducing or increasing the amount of original unbalance by virtue of trial weight placement. On the other hand, too much trial weight could be disastrous. and too little weight insufficient to achieve the 30% or so shift required for whatever program you use to make calculations (again, as far as I know).

Gee, I’m sorry Ralph, there I go again…


Gary Forsythe

I did not take time to calculate this but I would put my weight at 225 Degrees

I guess when you post in RED it gives additional creditability. The initial post, by you Mr. Stewart, contains insufficient information to answer the question, that is what is ridiculous. If you feel otherwise you are wrong. Just my . . . .

Gary,
Nice reply.

Sorry Gary, but I must have misunderstood what the "3 runs" were referring to. I thought the meaning was::
having a reference run,
a trial run that was no good,
a second trial run that was no good
and a third trial run that was finally usable
and then a "correction" test run
and possibly a "trim" run
making a total of 6 runs.

I read the rest of your post and find that there is no way, with my limited education, that I could give it justice by trying to responding to it.

But:

Due to circumstances way beyond my control, I am sorry to say I shall not now be able to post the "results" from this balance job since it appears that I don't have a clue as to what I am doing.

Jim, RED does not mean anything but seperating the quote from the reply making it easier to read.

Man!! You guys are tough on me.

BYE

This is becoming more and more like every other internet chat room. People spurt out their opinions, represent them as fact and say that anybody who says otherwise is an idiot.

Jim Richardson,

If you are not interested, just don't participate and keep your opinions to yourself. The fact that so many people have participated shows that your opinion is not shared by many others, so either you or about 30 others are wrong. I'll stick with the 30 others.

Ralph,

I am very sorry for both my posts. I was out of line and sincerely apologize to all.

Ralph, if you are uneducated in what you do, I am totally ignorant! Please never stop speaking your mind. That is how I learn.


Gary Forsythe

quote:

I balanced a fan in a foundry outside Chicago doing it the way I was taught that had vibration exceeding 100 mils to begin and managed to get it down to 9 mils in one shot (with a 6 pound correction weight

Gary, your statement above proves to me that you don't really think that trial weight placement is not important. Or are you saying you put a 6 lb. weight on a fan with 100 mils vibration and didn't carefully determine the trial weight location? If so, you were extremely lucky. Actually, I think you very carefully calculated the trial weight position before attaching your 6 lb. weight. (Or was that your 3rd run? If so, it wasn't "one shot", it was "two shots".... a "one shot" balance is 'reference run + trial run = done')

I always calculate trial weight placement very carefully, and usually calculate my trial weight amount carefully using a simple equation I learned from Art Crawford. But I always make my trial weight "too light" as opposed to "too heavy". If a fan is light-weight and running fast, I always use a small weight because you have no idea where those fans are on the resonance curve. I will calculate weight placement carefully, but will be wrong probably half the time on that type fan.

Large, heavy, slower fans are much easier. Not long after I figured this stuff out, I went to a power plant to balance a gas recirc fan I'd never seen before. While waiting for it to be shut down we looked up the foundation plan drawing which had the "reaction" weights for the bearing pedestals. This told me the weight of the rotating element plus the bearings. After taking reference data and recording the coastdown phase/amplitude, I figured the weight placement and calculated a balance weight of 900 grams. "That may be a little heavy" I though so I used half that much for the trial weight. I missed the weight placement by 5 degs. and the required balance weight was actually 902 grams. I never doubted the technique after that.

For 90% of the equipment you will balance, you can get your trial weight in at least the right 'quadrant' 90% of the time doing a quick 30 second calculation as outlined in Art Crawfords vibration book, Vol. 2. So why not take the 30 secs. it takes to do it?

It goes without saying that you don't just throw a weight on the fan without doing some calculations first. This is all much simpler than it sounds and after you do it a few times it is very easy.

One last note and then I'll post no more on this.... we all do dumb things sometimes. I went to balance a large baghouse fan that was running with 10 mils vibration (not quite "shaking the world", but close). I carefully calculated the position and amount of my trial weight (about 10 lbs.) and had it welded in place. The vibration doubled to 20 mils with about 1 deg. change in phase. I had managed to put exactly the correct amount of weight, exactly 179 degs. out. HOW ?!?!? I had marked the rotation backwards on the fan housing.... they were not impressed.

Happy balancing!

What is Art Crawford's equation for trial weight amount?

The following equation will give you an approximation of the trial weight:

U = (A * W)/2000

where U = imbalance = m * r
where m = balance weight mass & r = weight radius
A = amplitude, mils peak-peak
W = weight of the rotating element

Example: A = 10 mils, peak-peak
W = 10,000 lbs.
Then U = (10*10,000)/2000
U = 50 lb-inches
If R = 30 inches, then m = 1.67 lbs.

I would, or course, use 1/2 that amount for a trial weight.

This works better for large, heavy machines at slow to medium speeds where resonant amplification is often not a factor. I always work in 'grams' because I like whole numbers. I also always balance in 'mils' because the lag angle doesn't have to be corrected when doing my placement diagram, plus again, I like whole numbers or at least not really small numbers like when using velocity.

I am with Rusty, with good data there is no reason that you can't at least start in the right quadrant, and plus or minus 45 degrees is probably good enough for an initial angular placement.

Based on my experience, I tend to go a bit heavy on my cal weight because I don't like to waste a run and there have been plenty of times I was too light to get a good effect. But that also assumes the vibration is high, but not extremely high to start with, so there is a bit of margin available for the vibration to go up if I goof. As a result of having a good starting angle and an agressive cal weight I have driven the vector through the middle a few times however.

I take the extra time to do the calcs, cause I don't like to goof....

According to the ISO 1940 standard, the value of the permissible residual unbalance is used to give a suitable value for the trial weight. The permissible residual unbalance is obtained by multiplying the specific unbalance by the mass of the rotor:

U(per)= e(per) x M

e(per)= (G X 159.15)/F

U(per): permissible residual unbalance (gr.mm)

e(per): specific unbalance (gr.mm/kg)

M: rotor mass (Kg)

F: rotor speed (hz.)

G: balance quality grade

R: radius (mm)

The ISO 1940 standard recommends to choose a mass equivalent to 5 to 10 times the permissible residual unbalance as trial weight.

Taking the example of Rusty and assuming a rotation speed of 720 rpm (12 hz.) and a balance quality grade of G2.5:

U= (G x 159.15 x M)/(F x R)
U= (2.5 x 159.15 x 4545)/(12 x 762)
U= 198 gr.
Trial weight = 198 x 5= 990 gr. (2.18 lbs.)

ISO 1940 is of course for rigid rotor low speed balancing, but I like an approach like this.

For those in the US (with English? units) a balance quality of G = 0.665 would be equivalent (The same as) to the API criteria of 4W/N for permissible imbalance in ounce inches, W in pounds and N in rpm. 4W/N would most of the time be much to small for a trial weight, 16W/N or 32W/N would probably be better.

One may see 32W/N in the API for rotor response.

I can see where as Alec’s suggestion 5*Uper for G2.5 would be around TW=75*W/N (W in pounds, N in RPM, TW in inch-ounces) in the same format as Bill suggests. Does the ISO standard really suggest to use 5-10x the target tolerance? What if my target tolerance is G6.0? I can see why trial weight should change based on W and N but why should the trial weight go up just because I had a higher target residual unbalance?

I have also heard a suggestion to use a trial weight which creates a force no more than 5% of rotor weight. So TW <= 5% * W /(23*(N/3600)^2) . This TW suggestion would vary inversely with speed^2 while the last suggestion (based on ISO level) would vary inversely with speed.

I have never heard Rusty’s suggestion before. His doesn’t depend on speed (other than it only applies far below resonance) and his varies with initial unbalance unlike the previous two suggestions which don’t depend on initial unbalance. I guess the logic is that you want to be able to create an effect more than 30% of initial vib, then you will need more weight for a higher initial vibration.

Attached is a spreadsheet to compute and compare these three types of trial weight estimates:
TW1=0.5*(A*W/2000)*16 (based on vib - Rusty = 50%*Crawford’s)
TW2=32*W/N (Multiple of ISO Limit – like Alec’s and Bill’s )
TW3 = W /(23*(N/3600)^2) (Gives force equal to 5% of rotor weight)
(Units: W in lbm, N in rpm, A in mils pk/pk, TW in inch-ounce.)

In all cases I can see where it would be valuable to have coastdown data as others said to reduce the trial weight magnitude if the machine appears near resonance.

This message has been edited. Last edited by: electricpete,

TrialWeightMagnitude1.xls (28 Kb, 29 downloads)

Rules of thumb like this have a tendency to get people in trouble when followed blindly.

http://en.wikipedia.org/wiki/Rule_of_thumb
"A rule of thumb is a principle with broad application that is not intended to be strictly accurate or reliable for every situation."

Thumbrules are a good starting point for discussion if you don't have a lot of experience (like me).

What other factors might you think about in selecting a trial weight if you have not balanced a similar machine before?

"I can see why trial weight should change based on W and N but why should the trial weight go up just because I had a higher target residual unbalance?"

It's a good question. If a G6.3 quality grade was the requirement, the calculated trial weight would be 2.5 times bigger than G2.5 adding a risk to cause excessive vibration and damage to the machine if it is placed by lack of attention on the heavy spot. Maybe the users of ISO 1940/1 should always check if the calculated trial weight will not produce a dynamic force exceeding 10% of the static weight of the rotor? (F: m * r * w^2)