This is a Centac Air compressor( Mod 2CV25M3E) driven by a 600 HP motor. Any guesses as to what the gear mesh frequency might be? Reading was taken with a 10mV high frequency accelerometer at 12800 lines resolution. I am still trying to find some info on these compressors. tooth counts and bearing numbers.

Attachments

Original Post

A bit of a WAG here because you haven't given us much information.

 

The URL http://www.ecompressedair.com/...parts/2cv-model.aspx has a parts kit for the 2CV Centac.  What I see in examining the parts list is a 1st, 2nd, and 3rd stage pinion and of course a gear.  Assuming two of the stages are mounted on one pinion that breaks down to one gear and two pinions.  Do you know accurate speeds of the pinons?  That would help in calculating probable tooth counts.  Lacking speeds, and back to the WAG my hunch would be with 1106977 cpm as mesh but only because I can divide by the speed of 3570 RPM and come very close to an integer (310.07).  That is not very scientific but again, we need some more information.

 

Another thought comes to mind...if you can do a time synchronous average do that using the gear keyphasor for the trigger.  Non-synchronous components will drop out of the spectrum.  Might help...never know until you try it.

 

How is your accelerometer (make and model would also be helpful) mounted?  Some of that garbage out at high frequency (1.2 to 1.5 million cpm) may be the result of the mounting resonance. 

Last edited by John from PA

   I was hoping that if I could get a guess on the gear mesh I could back into the pinion speeds. All three of the pinions appear to be driven directly from the bull gear. The little info available suggests that there are rolling element bearings on the bull gear shaft which is at motor speed. All the stages have plain bearings? I am seeing some noise in the 3rd stage that I am wondering about. I should have included those also, was trying to keep it uncluttered. Here are some more readings on the first, second and third stage. The accelerometer is stud mounted and is a CTC AC220-1D. Your tooth count wag is the same one that I was looking at, seemed the most likely. If I can get the compressor down, I will get some tape on the bull gear shaft and try some time synchronous readings.

  I have ben playing with the autocorrelation function and the circular format. It appears to show peaks related to the tooth count/ impeller vanes in some machines. One specific gear train that I am confident of the tooth counts shows them very clearly. Any thoughts on that process?

Attachments

Just to confirm on your comment "All three of the pinions appear to be driven directly from the bull gear" there are three distinct pinions as opposed to two, one having a compressor wheel on each end.  In the three pinion arrangement there are commonly three rotors (gear and two pinions) at the horizontal splitline and the third pinion is above the gear, commonly at the 12 o'clock position.

 

I guess this machine does not have proximity probes?  If it does can you supply spectrums so we can estimate a 1X of each pinion?

Last edited by John from PA

John,

Your description of the arrangement is correct. The machine does have only 1 prox probe on each stage. It reads out on the panel display in mils. That is another area I would like to be able to collect with the 2130. Need to do a bit more research on that. 

It just amazes me how difficult it is to get some of this info from the makers of this equipment.

Originally Posted by Rusty Cas:
If there is a 'service' company that works on this compressor, they will have that info, and likely be more forthcoming with it. There should be a label or magnetic business card on the panel. One of the larger non-OEMs is Air-Relief (airrelief.com). They'd probably help you, especially if they thought there was a chance to get their foot in the door.

I just record the panel readings (mils) as  keypad entry points for each machine. You could read the prox probes, but there's a real danger of tripping the compressor. (Has been discussed here before - search for 'trip Centac')

As Rusty has pointed out, one has to be careful when connecting to a Centac machine.  If the proximity probes are Bently then the probes likely interface with a DC 990 transmitter.  Unfortunately because of economic and design constraints placed on the DC 990 by OEM and end users the output is not buffered as they would be on a normal Bently rack.  

 

Any pinion spectrums from the past?

 

If you desire to know more about the DC 990 (make sure there is a space) then do a search on this forum.  There have been quite a few posts in the last year or two.

 

 

 

 

Last edited by John from PA

I looked up that CTC AC220-1D accelerometer and it has a resonant frequency of 34 kHz.  That of course is with a stud mount so anything that deviates from that will cause a lowering of the resonant frequency.

 

If the gear tooth count is 310, then mesh frequency is 18845 Hz, slightly more than 1/2 the published resonant frequency for the accelerometer.  That could be a concern; most likely you are aking at measurement at a frequency on the rising portion of the resonance curve.  I suggest you primarily use the numbers obtained for trending ( a qualitative measurement) as opposed to a quantitative measurement.

Last edited by John from PA
Originally Posted by Rusty Cas:
When considering the "amplitude" at these high frequencies, keep in mind the relationship between acceleration and frequency. What actually is a "high" amplitude (g's) at 18kHz?

Extracting from ISO/DIS 10439-3 Petroleum, petrochemical and natural gas industries — Axial and centrifugal compressors and expander-compressors —

 

Part 3:
Integrally geared centrifugal compressors

6.3.1.3.2 Casing vibration shall not exceed an overall level of 4 Gs peak at gear mesh frequency.

 

Note that the "DIS" in ISO/DIS 10439-3 is for "Draft International Standard" so there was a later version not in my possession.  But I know this particular wording is unchanged.

 

Just as a point of interest if we go back to the Dudley Gear Handbook limits were put forth that would support a limit of about 20 g's peak at 20 kHz.  

Rusty:

I was able to contact someone from Air Relief in KY. He informed me that the 2CV series compressors have no rolling element bearings. This came as a surprise. I am going to provide him with some serial# and nameplate info and hopefully he can get me some tooth counts.

John:

Thanks for the ISO information. These compressors seemed really high on amplitudes the first time I looked at them but it seems to be the norm. I will get back to everyone with what I find out.

 

Thanks to all.

 

Phil

Last edited by P. Hine
Originally Posted by P. Hine:

John:

Thanks for the ISO information. These compressors seemed really high on amplitudes the first time I looked at them but it seems to be the norm.

I provided that information primarily for Rusty who asked "what is a high amplitude at 18 kHz."  The OEM should always be your first source for allowable operating parameters like vibration.  If the ISO or the equivalent API Specification was not invoked at the time of purchase, those levels may not apply.

 

It should be interesting what you learn.  Earlier I had provided a link to a 2CV parts/spares kit.  Part of the listing is below.  You have stated that each stage is a separate rotor yet each stage is shown as having only one "plain" bearing, which I infer to be a fluid film journal bearing.  The gear has two bearings and there are three thrust bearings.    

 

2 - 5X12568 - BEARING, BULLGEAR
1 - 5X16264 - BEARING, PLAIN, 1ST STAGE
1 - 5X16265 - BEARING, PLAIN, 2ND STAGE
1 - 5X16266 - BEARING, PLAIN, 3RD STAGE
3 - 5X12549 - BEARING, THRUST

Last edited by John from PA

Heard back from Air Relief. The 2CV compressors have 2 gearing arrangements . A standard pressure 100-125psig and a high pressure 150 psig.  The standard pressure tooth count is as follows:

 

Bull gear: 380

1st stage: 39

2nd and 3rd stage: 27

 

John,

Thanks for the parts list, I had overlooked the thrust bearings on the stages. I was focused on the single prox probe on each stage. The tooth count on the bull gear is much higher that I had thought. Looks like I will only be able to see the first gear mesh frequency. However now the pinion speeds can be calculated.

Rusty: Thanks for the info about Air Relief. Looks to be a valuable resource.

Originally Posted by P. Hine:

The standard pressure tooth count is as follows:

 

Bull gear: 380

1st stage: 39

2nd and 3rd stage: 27

Is there one pinion (27T) with the 2nd and 3rd stage wheel (as I first suspected) or two pinions each with 27T in addition to the 1st stage with 39T?

John,

The arrangement of the stages are looking from the driven end of the compressor 1st stage is at ~3:00, 2nd stage at 12:00 and 3rd stage at ~9:00. They all appear to be driven independently. There are 3 prox probes so I assumed 3 individual shafts? 

 

Phil

Originally Posted by P. Hine:

John,

The arrangement of the stages are looking from the driven end of the compressor 1st stage is at ~3:00, 2nd stage at 12:00 and 3rd stage at ~9:00. They all appear to be driven independently. There are 3 prox probes so I assumed 3 individual shafts? 

 

Phil

Well certainly the way you've described it there are three pinions in mesh with the 380T gear.  One logical reason they would do that, as opposed to mounting the 2nd and 3rd stage wheel on opposite ends of the same pinion, would be the load on the pinion with dual wheels.  With the 380T gear the train is likely a fine pitch.  As the pitch gets smaller the ability to transmit torque goes down.  This of course makes for a more expensive gearbox because it necessitates the third pinion, bearings, and instrumentation.

 

These pinions usually utilize thrust collars to transmit thrust into the gear which is fitted with a thrust bearing.  Since you have a vendor willing to supply you with some technical data you might pursue getting a cross sectional drawing.  Most OEM's will part with that since it won't provide sufficient information to manufacturer any particular part.  As OEM's, they have to watch the parts business!      

 

 

 

Thanks John. I will continue the learning process about these machines. Nice to have a resource like this discussion board to tap into. Always something new to learn!

 

Phil

Originally Posted by Rusty Cas:
John, have you ever seen a Centac up close? Next week I will be at a site where they have a number of these sitting around. I will get some pics, and possibly some drawings or other info (they just rebuilt one on site).

Actually yes, but generally larger units, the last (maybe 2009?) being a C800 or a C1000.  Beyond that, my first job was Philadelphia Gear in 1968 and I think that was when Ingersol Rand (Phillipsburg NJ) came out with the Centac line.  Centac by the way means Centrifugal Air Compressor.  Being just "up the road" from King of Prussia I spent a fair amount of time in their plant.

 

I do not think I have ever seen this "2CV" so pictures or a drawing would be great.  I have seen, and probably in the 1960's as well, the Citroen 2CV, often called the Ugly Duckling.  I'll stop there as I'm giving away my age...oh, I forgot my birthday will be posted next May anyway. 

 
Last edited by John from PA

Centac design is a bull gear that drives a series of single stage pinion/impeller configurations.  The pinion is supported by two radial bearings with a single thrust bearing.  Only the drive end bearing is monitored with a single proximity probe (old model) whereas the later models and larger units have X/Y arrangements with a thrust probe.  The thrust probe can be used for a Keyphasor as well.

Attached is some data on the CV2 for both low and high pressure units. Included is a section drawing on a pinion/impeller arrangement.  I believe it is pointless to capture casing measurements (frequency activity is too high to capture quality data (i.e. acc mount), too much noise and casing transmission too low).  Only casing measurement worth its salt is the bull gear DE bearing.  Depending on the age of your model the  prox is typically BN 3300 or 990.  Have the customer place the channel on 'bypass' to connect in and take readings on 3300.  Have no problem to connect to 990 ensuring the portable equipment is DC powered 

 

Attachments

Files (1)

John,

Just a little knowledge about Centac compressors, huh?!!  Sometimes it really pays to have ancient history.  Speaking from experience, of course.

Ron Brook

Add Reply

Post
×
×
×
×
Link copied to your clipboard.
×