Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,810
Reply with quote  #1 
I have had the CSI Triax since I bought my 2140 but haven't used it much as the straight cable wasn't really practical for route use.  At the time there was no option for a coiled cable, but CTC is now making a coiled cable for it.

The built-in base has 2 rails, so when you mount it in a horizontal position on a curved motor endbell (for example) the poles are A = Horizontal, B = Axial , C = Vertical (H,A,V see diagram below). Only CSI would do it this way, as opposed to H, V, Ax which is the "standard" (correct?).  In the description they even reference "horizontal, vertical, axial" and show it that way in the top illustration, but then where they show it attached to a curved surface (as would be normal), it's oriented H,Ax,V.  Why would they do this!?!?  (Yes, I'm 'venting' a little)

CTC refers to the channels as X,Y,Z so when I ordered the cable I had it wired X,Z,Y which should correspond to A,C,B or H,V,Ax... correct?  Or am I missing something?

I can't think of a way to test to be sure the cable is wired correctly, except to find a machine where the 3 measurement positions are physically very close to each other, preferrably on a common solid housing, where the magnet has equal attachment at all 3 positions, and where vibration is substantially different in all 3 directions.  Right?

Is there a better way to do this, in the field?  I could come up with something back in the shop I guess, but not right now.

CSI Triax.png


__________________
"The trend is your friend"
Beatnik

Sr. Member / Supporter
Registered:
Posts: 183
Reply with quote  #2 
Go in manual analyze, make sure all channels are activated and take a long measurement. You will see the live data move as you move the triax in each direction with your hand.

What is worst is that MHM doesn't use A-B-C, H-Ax-V or X-Y-Z! it uses 1-2-3... You'll need to test that to make sure you assign the right channel.

Edit: On my CTC triax (AC230) : Z = A, Y = B X = C

http://www.ctconline.com/fileup/AC230%20Series.pdf look at the picture.
Walt Strong

Sr. Member
Registered:
Posts: 889
Reply with quote  #3 
Rusty,

Hold the accelerometer in your hand at tap it with knuckles (or small plastic tip hammer) in each orthogonal direction (one at a time) while monitoring spectrum/waveform. This should identify what the analyzer "sees", and you can then resolve the mounting issues and make appropriate measurement point labels. Most tri-axial accelerometers that I use have Z-axis perpendicular to mounting surface and use the "right-hand-rule" for X-Y-Z orientation. Unfortunately you can't always orient the unit on machine correctly. Some systems use permanent blocks/disks with key/notch to control orientation, but these are somewhat restricted to "ordinary" machines.

Walt
RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,810
Reply with quote  #4 
Well, that's brilliantly simple!  And it works!  But, it looks like the cable is wired as ABC, in spite of what I ordered. But maybe I inadvertently ordered what I go.  [confused]   But that may be better I guess.  I was going to relabel the poles on the accel, but perhaps it's less confusing as is.  I don't think I have my original cable with me right now.  I'll compare the two.

Thanks Beatnik. It's certainly helpful to hang out with folks who are smarter than I am.

And thank you, Walt.  I'm still not getting the "right-hand-rule" -- there seems to be some disagreement across the internet about it.  I don't really even understand the markings I'm seeing on pictures of triax accels.  Guess I need to experiment and figure it out.  I may be the only one who is confused.  Maybe everyone else (CSI, CTC, Walt) is following established conventions.  I don't have to "like" those conventions though. (always the contrarian)



__________________
"The trend is your friend"
Beatnik

Sr. Member / Supporter
Registered:
Posts: 183
Reply with quote  #5 
From MHM help file for future reference: We can see, with the ctc file I provided earlier, that CTC and CSI disagree on X-Y-Z and A-B-C
mhmtriax.jpg 
 

Beatnik

Sr. Member / Supporter
Registered:
Posts: 183
Reply with quote  #6 
ok so I did a little digging and tried to remember my experience with the A0643TX.

When I switched from CSI to CTC (both cables and accelerometers) I had to change all the channels assigned in MHM because H and V were inverted! So I had to change all the channel assignements in MHM.

My A0643TX didn't last long. You might want to switch to the CTC one now and prevent having to assign all channels twice. Of course you could prevent doing it by having the cables wired differently if you change later but switching accelerometer now and using normal cables might be more simple.

I switched for many reasons.

1st: Being able to put the magnet of my choice (the CSI magnet is weak, and I see this as a big problem, specially for the axes perpendicular to the magnet)
2nd: The garante on CTC products
3rd: Reliability 
4th: L shape are more convenient than straight in my environnement
fburgos

Sr. Member
Registered:
Posts: 670
Reply with quote  #7 
Now i see the stock triax has -+3db in the range 0.5-10000hz primary axis other different, but ctc ac230 only has one specification +-3db 0.6-10000hz does it mean all axis have this sensitivity, then i could take two sets of peakvue
RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,810
Reply with quote  #8 
Even if the CTC triax has the same sensitivity on all three axes (which I doubt), I think PeakVue data will only be valid for the primary axis. PeakVue is measuring “stress waves” according to CSI literature.
__________________
"The trend is your friend"
Beatnik

Sr. Member / Supporter
Registered:
Posts: 183
Reply with quote  #9 
Quote:
Originally Posted by fburgos
Now i see the stock triax has -+3db in the range 0.5-10000hz primary axis other different, but ctc ac230 only has one specification +-3db 0.6-10000hz does it mean all axis have this sensitivity, then i could take two sets of peakvue


Yes all axis have the same sensibility but the rigidity of the triax is not the same in all direction, rigidity affect the resonnance frequency of the mounting and therefore the quality of the measurement at high frequencies. I agree with Rusty that the main axe measurement should be better.

It's often linked here but here is "the" peakvue paper:

http://www2.emersonprocess.com/siteadmincenter/PM%20Central%20Web%20Documents/PeakVue_IntroductionPaper2001.pdf


Looking back at it I found something interesting: P.34

"4.2.7 Selecting the Optimum Sensor for PeakVue Measurements

Due to the dispersion of the stress wave packet, it is desirable to locate the

sensor as near to the initiation site as possible. This generally will be near or on the

bearing housing preferably in the load zone. Stress waves will propagate in all

directions. Hence the selection of axial, vertical, or radial is less (relative to normal

vibration monitoring) of an issue than is mounting the sensor in or near the load zone

with the caution that we are monitoring “waves” and hence, must always be cautious of

encountering nodal points which can occur due to multi-path transmission and in the

vicinity of sharp corners, etc."

RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,810
Reply with quote  #10 
Quote:
Hence the selection of axial, vertical, or radial is less (relative to normal

vibration monitoring) of an issue than is mounting the sensor in or near the load zone



But I would think that as you mount the sensor further from the bearing ("initiation site"), that the axis selected becomes more important.  For some bearings my sensor is literally 2" or less from the outer race with a solid metal path, so I can see how the selection of H/V/Ax doesn't matter.  But what if you're on the endbell of a motor and your sensor is 12-18" from the bearing?  Even though you may still have a fairly solid path, is it going to matter then?

__________________
"The trend is your friend"
fburgos

Sr. Member
Registered:
Posts: 670
Reply with quote  #11 

For you guys where is the load zone? and where do you measure peakview?

first question comes from; according to TA, schenck and pruftechnik, they all agree that we must select an envelope preferably on the load zone, TA and schenck select horizontal position for enveloping, I remember reading on a book by Victor Wowks and he stated the envelope is best on vertical (botom or top).

second question; I have 5 big reduction gearboxes, 4 shafts/steps , the horizontal measurement is out of the question for the middle shafts, the vertical is a solid metal contact but far on the load zone (6-8 inches) but all the axial are couple of inches from the load zone (bottom vertical), since for CSI the peakview location is less important than proximity to the load zone, it means should I preferably take peakview on axial (with main accelerometer axis.)

Quote:
Originally Posted by Beatnik



It's often linked here but here is "the" peakvue paper:

http://www2.emersonprocess.com/siteadmincenter/PM%20Central%20Web%20Documents/PeakVue_IntroductionPaper2001.pdf




thanks, some heavy reading is waiting for me, just let me finish this planned maintenance alive and i'll catch up.
DDane

Avatar / Picture

Member
Registered:
Posts: 18
Reply with quote  #12 
Rusty and others,

   Ive been an onsite data collector/analyst for 12 years (short timer) now. My training over the years has steered me away from route data collection with a tri-ax ACCELS so I always went X,Y,Z with single accel. The only time I have used a Tri-ax otherwise is for ODS. But one exception is I will say we have 2 Frick Ammonia Compressors that had IMI Tri-Ax Accels on them but they seemed to amplify amplitude tremendously so we converted to single axis accels with better results. We use SKF Aptitude and Microlog AX and GXs. Just curious your thoughts on route based collection with Tri-Ax's?

                                        Regards, Dave Dane
Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #13 
Quote:
Originally Posted by DDane
Rusty and others,

   Ive been an onsite data collector/analyst for 12 years (short timer) now. My training over the years has steered me away from route data collection with a tri-ax ACCELS so I always went X,Y,Z with single accel. The only time I have used a Tri-ax otherwise is for ODS. But one exception is I will say we have 2 Frick Ammonia Compressors that had IMI Tri-Ax Accels on them but they seemed to amplify amplitude tremendously so we converted to single axis accels with better results. We use SKF Aptitude and Microlog AX and GXs. Just curious your thoughts on route based collection with Tri-Ax's?

                                        Regards, Dave Dane


Dave,

You have a 'mounting natural frequency' that should usually be 3x your fmax or higher. if you have a crap connection between the accel and the machine, this can decrease, but generally stays higher than we need for single axis accels, even if we have a crappy handheld probe. However, the transverse directions of a triax will have a natural frequency muuuuch lower, especially with crap mounting. Most mounting is crap in the eyes of a triax, unfortunately! You've got to keep your magnets in really good shape and 'prepare' connection spots if you don't want to have a bunch of BS inflated noise at higher frequencies that will amplify your overall values. If you really want to trust your transverse channels, you need to mount with adhesive/stud. Otherwise only low frequency components are really trustworthy.
fburgos

Sr. Member
Registered:
Posts: 670
Reply with quote  #14 
Quote:
Originally Posted by Beatnik
ok so I did a little digging and tried to remember my experience with the A0643TX.

When I switched from CSI to CTC (both cables and accelerometers) I had to change all the channels assigned in MHM because H and V were inverted! So I had to change all the channel assignements in MHM.

My A0643TX didn't last long. You might want to switch to the CTC one now and prevent having to assign all channels twice. Of course you could prevent doing it by having the cables wired differently if you change later but switching accelerometer now and using normal cables might be more simple.


Hi beatnik, i´m about to do this transition.

May I ask what is the code for the cable you have?


I was quoted this

C561.png 


but the principal axis on the AC230-1D is Z, i believe I need this to not make any changes to my Database

C580.png 


Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.