High Vibration in a Gas Turbine

Respected Members;

We have face a very critical situation regarding the vibration in a Gas Turbine CENTAUR 40 model by Solar Turbines.

Before explanation of the nature of the vibration, I share the machine design data for reach the root cause.

1. 11 Stage Axial compressor.

2. 15015 RPM of the Turbine engine.

3. Two shaft Gas Turbine.

4. There are 3 tilt Pad bearings on turbine engine side.

5. First critical speed about 28 % and second critical speed at 56%.

6. OR36 with NVGATE and Vibroport 41 is used to capture the FFT or Time wave forms. There is no option in the turbine for phase analysis.

Now i explain the fault for diagnostic.

 When the turbine start every thing is going fine. No High vibration found at any frequency.

When engine reaches at 87% the Peak at 2X is going to increase while 1 X is not increasing and 2X reaches at the alarm value as speed increase but 2 X goes down when engine achieved 92 % RPM.  It means the vibration found high between 87 to 92 % RPM band and just on the diffuser section  where bearing # 2 found. 

On the other hand the vibration on bearing 1 and bearing 3 is well with in limit.

Here i attached some data in picture format which helpful for discussion.

I describe the file name and their detail

1. GP-V and GP_H.JPG shows the spectrum during 87 to 92% band.

2. SKID_SPECTRUM.JPG  shows the FFT at skid base.

3. support_FWD.JPG shows the FFT at support FWD side of the engine.

File 2 and 3 FFT taken to check the soft footing and looseness.

The other is the view of the diffuse section where velocity probes are attached.

Here is also attach a Time wave form get by cross chanel.

One thing i observe this vibration is also occur at deceleration at the same band.

also if i stay the engine between this band for 6 to 10 minutes the 2 X peak starts to goes down but still high.

If more data required please feel free to ask and advise me how i fix it?

GP_H means Gas producer Horizontal, and GP_V means at vertical.

Regards 

GP_VERTCAL AND HORZ AT ENGINESKID AND SUPPORTSKID_SPECTRUMSUPPRT FWDTIME WAVE FORMVIEW_TURBINE

Attachments

Photos (6)
Original Post

When engine reaches at 87% the Peak at 2X is going to increase while 1 X is not increasing and 2X reaches at the alarm value as speed increase but 2 X goes down when engine achieved 92 % RPM.  It means the vibration found high between 87 to 92 % RPM band and just on the diffuser section  where bearing # 2 found. 

What do you mean by alarm value?  Is this a value that you have decided and not Solar?  Do you not have other vibration transducers on a Solar turbine, which are monitored with actual alarms?

You give but 1 turbine speed.  There are two speeds.  Any comment?

 

Hello, Good day

The alarm value is given by the OEM that set the value 0.17IPS FOR A SINGLE PEAK BETWEEN 15 hz to 2khz. 

The over all alarm value is 0.25 IPS.

I understand ur next question about two speeds, that 87 to ,92 percent the peak at 2X is going to high but before and after this speed band its value is well within limit

If u have any other meaning please explain little more

Regards

Interesting, the Solar turbines that I have seen use shaft relative probes for the bearings (forget old Saturns for this).  Are these Solar supplied?  On an epicyclic output gear?  Some other Solar 'custom (or selectable) feature?'  

I am surprised that I haven't run into these alarms from Solar and am interested in finding out more.

The meaningfulness of this frequency range on a gas turbine, I am more familiar.  Most manufacturers do not use that high a frequency, perhaps not that low either.  Many other frequencies (particularly with seismic measurements) will show up on a GT.  Blading, all that flow energy, especially in the exhaust, depending upon the location bleed valves (and these can have vibration faults as well as other faults).  They have flexible supports and a flexible casing.  Is there some p[lace to find this information on the Solar site; otherwise, I'll have to ask a Solar guy who is onsite.  

Bill, the Centaur 40 gas turbine generator arrangement is a single-shaft gt with tilt pad bearings. The monitoring system on the turbine is velocity based. The gearbox is an epicyclic, planetary arrangement as I recall and it has an accelerometer based system. The 4-pole generator can be equipped with proximity probes.

In the mechanical drive the Centaur 40 gt is a two-shaft, still with tilt pad bearings. The instrumentation system is velocity based. I do not recall ever seeing one with proximity probes, although that is not to say they might exist.

Early on the OP stated the turbine was a two-shaft, so I assume this is a mechanical drive arrangement although the driven machine hasn't been mentioned.

 

The original post indicated that his turbine was a two shafts GT.

1. 11 Stage Axial compressor.

2. 15015 RPM of the Turbine engine.

3. Two shaft Gas Turbine.

4. There are 3 tilt Pad bearings on turbine engine side.

It does matter what we are talking about.  We have both types of Centaurs.  The closest Solar turbine to me is in the next building, but it is one of those rarer Mercury 50's.  

William_C._Foiles posted:

The closest Solar turbine to me is in the next building, but it is one of those rarer Mercury 50's.  

LOL, at my former gear employer, they used to refer to the Mercury 50 as the "regurgitated" as opposed to Solar's name "Recuperated".  We were never fortunate enough to get the contract to supply the epicyclic gearboxes.

Hello Friends;

I explain all the detail, I want you all friends discuss about the cause of High vibration.I discuss only here driver that is gas turbine engine. The driven equipment is centrifugal compressor.

In driver section there are 5 tilt pad bearing. 3 in the Turbine and 2 are in the power turbine as i mentioned above it is two shaft turbine.

So in my view, there are some possibilities

1. Resonance (But it is not confirmed because if we maintain the speed when 2X is high after some time that is 20 to 40 seconds the peak is in the range but still high)

2. If i see the time waveform might be rotor rub indication appear but it is conform when engine disassemble. What is your opinion?

3. Soft footing and looseness not in my view because 1/2,1/3,1/4 harmonics not shown and the graph is not higher from its axis.

Please discuss what is the problem for high vibration in the specific speed band after and before it is good. Any suggestion?

High 2x can be a concern.  It may be indicative of a serious issue.  On GTs I don't suspect traditional type soft foot because of the support system.

Rubs can't be ruled out.  

Is the vibration changing?  Getting worse?  (Even better could be a bad symptom).  It would be great if  you had information regarding shaft bows from shaft relative measurements -- particularly if this is changing.  

I read that you have the issue on the compressor-turbine section and not the power turbine section.  

 

Thanks William_C , It is a serious issue, But please suggest me what type of vibration signature we get to verify the shaft bow.

In GT there is no option for proximity probe, Please suggest what

 1. Time wave form 

2. Waterfall

3.Cascade 

4. FFT with tracking constant or synchronous, Unfortunately orbit graph option in NVGATE 3.0 2006 version is not supported.

Regards 

There are fewer things you can do with only casing measurements on a GT.  You can look for a hysteresis between starts and stops on Bode and polar plots.  I think that polar plots are easier to see a trend for this kind of data.

If a bow is worsening the problem should eventually get worse with higher vibration.  A change in the 1X vibration is important.  

Of course since you have an issue with 2X vibration you can trend this.  In fact, you can generate Bode and polar plots for this and trend these data- startups, coastdowns, and during operation (like a polar plot at speed).

If you have noticeable changes like step changes between starts or more continuous changes while operating, this needs attention.  Because this is very unusual, you need to keep and eye on this and determine a likely cause.  Step changes during operation are not unheard of; they can be a rub initiating or ceasing or some thermal issue around the rotor structure (may not be too serious - but can be).

 

Add Reply

×
×
×
×