Hi,

Please see the attachment. these were the vib levels with phase measured on the GT-Gearbox-Comp train on 2 consicutive days to check the repeatability in readings. vib levels and phase were very consistent.

it was decided to put the correction mass on the coupling between gearbox and 3rd stage compressor where the vib levels are high. casing vibs are less.

the trip levels are set at 90 microns. so we cannot attempt a normal balancing here with trial weights. since adding trial may increase the vib levsl and the machine will trip. the customer says i need to go for single plane balancing near 3rd stage compressor and mass provision is provided on the coupling adjacent to that.

my question is how to estimate the correction mass and angle without adding trial weights. and get the results in first shot....any idea...?

Madhu

comp_ampl_and_phase.xls (39 Kb, 32 downloads)

Just to add up the vib levels are taken from the bentley panel and angle from the key phasor

Madhu

quote:

since adding trial may increase the vib levsl and the machine will trip.

"May" is true but an experienced balancer with transient speed data will be able to determine where to add the trial weight so that most likely the amplitudes will go down, not up as you infer.

Do you have any transient data that would give knowledge of where the balance resonances or critical sspeds are located?

John

You do know that you don't always put the correction weight where the vibration is highest? John's suggestion of looking at transient data can help with the likely end of the machine to add weights as well as the angular position.

One must be careful of nodal locations. It is easy to have a node between the measurement point and a coupling where the weight is to be added. This will only throw the correction angle off by 180 degrees. Having a rotordynamic model will help with such analysis and decisions.

No matter what "system" you attempt in making a "one shot" deal, it is only going to be a S.W.A.G. (Scientific Wild A-- Guess).

And I doubt if you get it with "one shot" unless you have some history on the same procedure done on this same machine. If it is done in "one shot" it will be 99 44/100% luck.

If you know whether or not it is running absolutly below its critical frequency, then the percent might be about 50/50.

I see that there is a 135 degree phase between one direction and 125 on the other one across the coupling. Is it possible the vibration is from misalignment? Since there is also 114 degrees and 125 degrees diffrence from 3rd to 2nd stage, which sort of indicates when the 3rd stage is going in one direction the 2nd stage is sort of going in the other. So would this say that a single plane on one end might make things worse? True there is a hugh amplitude differnce between the gearbox and the compressor and even the other coupled compressor, but vibration is like electricity, it travels in the path of least resistance.

One other thing, why can't the trip system be disabled while this is going on so the system will not trip? I guess they are wanting to stop it only once, huh? But what if this one attempt fails? Then we have to stop it a second and maybe even a third time, so why not be prepared and have it where it can be done right to start with and disconnect the trip system?

Only my opinion and I might be --------------------

quote:

One other thing, why can't the trip system be disabled while this is going on so the system will not trip?

Because its unsafe!!

Dangerous to make these recommendations when dealing with someone who might not know what they are doing.

quote:

quote:
One other thing, why can't the trip system be disabled while this is going on so the system will not trip?


Because its unsafe!!

Dangerous to make these recommendations when dealing with someone who might not know what they are doing.

Are you saying Madhu doesn't know what he is doing?

True it is dangerous to a certain degree and is only as unsafe as one allows it to be. With this being a turbine operated system and more than likely having the ability to increase the speed at a very controlled rate, any approaching danger from high vibration could easily be seen well before there became a critical situtation encountered.

I am not recommending anything. I asked "why can't it----", I never recommended doing it, or did I? I do not think so.

I am sure there is or at the least should be an emergency stop system other than the automatic high vibration trip system, which is not always as reliable as some might think. I would hate to know that I had to trust only the high vibration shutdown switch on anything and not know where the emergency shutdown switch is even located. But then again, this being a controlled speed (I hope) rather than an electric motor that knows only stop and go, the danger is very slim of allowing the situtation to get out of control.

At 6000 rpm I would have someone with their hand on the turbine manual trip mechanism no matter what was being done.

OMOAICBTW

I doubt that one can say anything definitive regarding alignment from the phase difference across the machines with one point at running speed on high speed compressors. Transient plots may be of some use. These machines may (probably) operate above a critical speed.

What is the person to do with their hand on the E-Stop? The consequences for a shutdown have to be investigated. They could win a big yellow "chicken" shirt award if not worse.

quote:

What is the person to do with their hand on the E-Stop? The consequences for a shutdown have to be investigated. They could win a big yellow "chicken" shirt award if not worse.

The person would be instructed to shut it down if a critical situtation was being observed. As far as the "yellow chicken shirt award" goes, it might not say "chicken" it might be yellow but might say "I saved a million".

If the company can't deal with it (shutdown), maybe they should pull the compressor out and correct the balance, if that is the problem, in the correct fashion as opposed to placing a single plane balance on the coupling of just one end. Sounds like the client is accepting responsibilty of any malfunctions by restricting the procedure in doing the job.

Sometimes when a compressor has an emergency shutdown some of the gas inside has to end up outside (Usually, there are limits on flares and other emissions.). Whatever process uses the compressors may be stopped, which may trouble. Anyway, why would you lift the shutdown to replace it with someone’s reaction time?

I’ve had people push the ‘chicken button’ on jobs that resulted in problems.

The desire to balance insitu is understandable, but one doesn’t have much to work with in terms of balance planes. If one takes the compressor out one must find a high speed balance pit (preferably somewhere near the location, not in another part of the world), disassemble and component assembly balance, or some of both.

Without more evidence, it is not possible to state that this high vibration results from unbalance. Admittedly, unbalance is a distinct possibility, but has glitch been ruled out? What is the history of this problem - when did the vibration appear?

This problem needs transient data (especially 1X compensated polar plots from all 4 compressor probes) to figure out the operating mode along with the high/heavy spot relationship: with this info you can confidently add a small weight that will not make things worse and will likely make things better.

Hi,

Pls find the spectrums attached. sorry at present i do not have the transient data and mode shapes. i know that these are very much required to know the dynamics of the train.

Thanks
Madhu

Vibration_study_report.pdf (426 Kb, 17 downloads)