Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
Dave Reynolds

Member
Registered:
Posts: 53
Reply with quote  #1 
I am inquiring about vibration data collection above 10khz.
Example: centrifugal compressor, 3rd stage is 36000 rpm with 35 teeth, gmf is 1,260,000 or 21,000 hz...
What technique can be used to identify useable vibration data at this frequency or higher, like 3.25 gmf?

Thank you in advance
Dave

GaryVibe

Member
Registered:
Posts: 16
Reply with quote  #2 
Hello Dave. I have collected some high frequency data on gearboxes in a similar range that you inquire about. I purchased B&K Types 4394 accels with a 25Khz response. I measured the time data and then processed it to help identify 1st and 2nd orders to about 22-23Khz. The data was repeatable when I stud mounted the sensor to the the bearing housing. We had the shop remove the gearbox cover and then stud mount the sensors insides due to access to the internal shaft bearing locations. Then ran the cables through a conduit fitting to a box on the outside of the gearbox.  This was not an easy or inexpensive measurement, however the data was clear and repeatable. I tried stud mounting to the outside of the gearbox and never had good, repeatable plots from those locations and never could clearly identify the gearmesh. 
These were newly manufactured gearboxes and this was for acceptance testing. The sensors were left installed with a junction box mounted to the outside with the sensor connections. 

The data was useful in identifying the overall vibration levels at those high frequencies and the plotted data was provided to the end users for reference use in the field. 

Hope this helps. 

Gary K
RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,810
Reply with quote  #3 
Dave, I'm monitored 20 or so Centac 2- and 3-stage compressors for about 15 years now, and as far as I know, we've never had a gear problem on any of them.  They spin fast, but the rotors are actually pretty light.  I've never gone to a lot of trouble to monitor the compressors - the prox probes seem to be adequate.  If they see a significant change in the probe vibration, they're just going to have the air-end rebuilt anyway.  There's really nothing else you can do with these units.  So is it possible to reliably monitor/diagnose these units? Yes.  Necessary? I'm not so sure.
__________________
"The trend is your friend"
Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,402
Reply with quote  #4 
We do a bunch of three-stage Centacs and I don't think we have had any gear issues, either.  I mostly focus on the motor and the roller bearings on the bull gear. On the compressors, I look at running speed and VPF but don't find much.  A couple of times we found excessive running speed harmonics and found a little chip in the rotor and a bearing wear.

But if I wanted to monitor the gmf, the setup Gary suggests is the way to go if someone will pay for it.
Dave Reynolds

Member
Registered:
Posts: 53
Reply with quote  #5 
Thank you all for the quick answers, I agree with all comments, always good to network with people who have been there and done that....
Thanks again
Dave
Alex

Avatar / Picture

Sr. Member
Registered:
Posts: 204
Reply with quote  #6 
Here comes the question if it is worth to buy special high frequency accelerometer, let's say B&K for 25 kHz limit. Just to measure 1x GMF. We all know that 2x GMF is more often reliable than 1x GMF. And than we have the same problem again.
John from PA

Sr. Member
Registered:
Posts: 963
Reply with quote  #7 
As others have pointed out, you will find these machines, at least with respect to gearing failures (as root cause) very reliable.  In my experience with this type of machine, integrally geared compressor as so-called by API, most of the issues will be related to rotor related frequencies; unbalance, bearing instabilities, etc.

Now, on to what you desire to accomplish:

Quote:
Originally Posted by Dave Reynolds
What technique can be used to identify usable vibration data at this frequency or higher, like 3.25 gmf?


From a practical standpoint reliably monitoring out to beyond 3X of gear mesh is so difficult (in this type of gearing) it is an exercise in futility and just not worth the effort.  I have been involved with numerous such attempts, even at the gear manufacturing level, and it isn't practical nor does it deliver worthwhile information that isn't available with monitoring to slightly over 2X of mesh, and for that I recommended the Metrix line, specifically the 736 or 736T.  It will have to be a permanent installation and be aware that mounting is critical (API 670 mount preferred) and the cable will need to be carefully anchored based on the accelerator manufacturer recommendations.  Even then you should only trend this information, a qualitative as opposed to quantitative approach.
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.