Dear all


Due to some reasons the load over one of the STG increased from 15MW to 25MW, no major change observed in Generator & Turbine online & offline vibrations


But the single stage reduction GB (6300 rpm to 3000 rpm) showed high increase in offline vibrations i.e from 8 mm/s rms to 13 mm/s rms, but the shaft vibrations were not having any change i.e 30 microns



Turbine rpm:- 6300

Generator rpm:- 3000

 No. of teeth on high speed shaft:- 37

No. of teeth on low speed shaft :- 78

No. of teeth verified from GB data sheet & name plate on GB


Sidebands for 3000 rpm also present in FFT


The FFT shows harmonics of 72242 which i am not able to correlate with anything in the GB ( i might be missing something), also there is no GMF peak in FFT. (MY Fmax is 8lac)


Note:- there are 3 similar STGs all having the same FFT spectrum in their GB, however they are running at low vibration amplitudes.


Is it possible that a GB does not show GMF??

What is this 72242??







Files (1)
Original Post

In cases where there is an unidentifiable frequency in a gearbox used for power generation I'm always a bit suspicious of the machine being a double helical and the axial spacing being incorrect.  Is the gearbox a single helical or double?  Does this machine have a limited end float coupling between the steam turbine and gearbox that has a pre-stretch value?  By limited end float I mean diaphragm or disk type.  Can you supply a cross section drawing of the gearbox?


Not that it is a major discrepancy, but I calculate mesh frequency as 233100 cpm as opposed to your 234,000 cpm.  Are you certain of the tooth counts and shaft speeds? 

I think the origin of this vibration is not at 72242 but half of it. Place your harmonic cursor on 36121cpm and you will see. There is each time a higher and lower amplitude of the harmonics of this component.


It then is a 5.7X running of turbine. It does not seem like it a lot, but the frequency can be of an inner race defect of a roller bearing. Probibly not present here but you could check it. (maybe not main bearing but something attached like an oilpum or so)


Do you have the ability to accomplish a time synchronous average on the signal using both (separately of course) the pinion and gear speeds as the trigger?  A good alternative would be increasing the resolution to better identify the frequency.  I see that the 72242 is very close to a 24X of the gear.  What would be of interest is if the frequency drops out with a pinion speed triggered TSA but remains with a gear triggered TSA.  This would support that a potential error from the manufacturing process might be generating a "ghost frequency".


Can you also elaborate on the transducer being used and how it is mounted when these readings are being acquired? 


2 things

1. what i can do is i can remove the magnet from accelerometer & use it as hand held..(if that helps)

2. To verify the frequencies, I used another make analyzer with separate sensor, however sensitivity & mounting method was same..similar frequencies were observed..


I personally doubt the teeth count..but i verified it twice from the name plate on the gearbox also from the data sheet provided from OEM.


now If i go with what Edwin said that its not 72k but half of that 36121,,he is right ..some points I get harmonics of the said frequency. but still cannot point the cause..

Originally Posted by Akhil rathore:

OOPS its John, I misspell


let me try for time synchronous averaging..I'll get back on this..

If you have doubts on the tooth count see if you can get a good feel for turbine speed.  I'm assuming you are connecting to the grid and it is reasonably stable and hence can trust the 3000 RPM.  But with the tooth counts you provided the turbine speed would be about 6324.  If I just use your 6300/3000 speeds and run that through my tooth count calculator there are many choices, most of which are not 37/78 tooth count.



Yes one Gas turbine GB was examined, & found "tiger stripes" on the gear, other than this nothing.

Let me refresh the details & question

3 no.s of Steam turbines

Gearboxes details

teeth= 37/78, rpm =6300/300

2 no.s of Gas turbines

Gearboxes details

teeth= 43/74, rpm =5300/300


All the above mentioned 5 machines show similar spectrum in their Load gearboxes (Herringbone gears) with harmonics of "72,000" as their problem causing frequency.

Also this "72,000" is not constant, it varies with power frequency. let me elaborate 

when power / grid frequency drops, lets say 49.60 Hz the value goes to 71,888

When power / grid frequency is high, say 50.15 Hz the value reaches to 72,014

that is it varies with power frequency.



Add Reply

Likes (0)