un balance category

dear all

as you know for measuring relative vibration by proximity sensor on journal bearing use two sensor located by two channel x and y

45 degree from each others. to evaluating un balance (static or coupled ) we measure phase difference between de and nde of shaft.if we have  no difference phase between x channel of sensors and have a 180 degree difference between y channel of two sensors.if i evaluate by x channel we have static unbalance but by y channel we have couple unbalance.

which one is correct ?

best regards       

Original Post

When I read this, I think there may be some discrepancy between the words and intended meaning. 

Are the sensors one each plane (end) 45 degrees apart, or does the x sensor on one end vary from the x sensor on the other end by 45 degrees, or is 45 degrees correct?  Besides, I have no clue as to which is x and which is y.

x in phase end-to-end - what is the y out of phase end-to-end.  If these two conditions occur at the same time (which they can or at least approximate), I would try to understand the dynamics better.

I am having some issues with the intended meanings here. 

William_C._Foiles posted:

When I read this, I think there may be some discrepancy between the words and intended meaning. 

Are the sensors one each plane (end) 45 degrees apart, or does the x sensor on one end vary from the x sensor on the other end by 45 degrees, or is 45 degrees correct?  Besides, I have no clue as to which is x and which is y.

x in phase end-to-end - what is the y out of phase end-to-end.  If these two conditions occur at the same time (which they can or at least approximate), I would try to understand the dynamics better.

I am having some issues with the intended meanings here. 

i have two proximity sensor on each bearing that have two channel out put.( X channel & Y channel that located 45 degree of each others on  shaft bearing.

to identify static or couple un balance, i measure phase between X channel of  NDE bearing and X channel of DE bearing.( i see no difference phase).

but for Y channels between NDE and DE of bearing i see 180 degree phase difference.

my question is , which kind of un balance i have on the shaft? static our couple

 

best regards 

 

aziz58 posted:

dear all

as you know for measuring relative vibration by proximity sensor on journal bearing use two sensor located by two channel x and y

45 degree from each others. to evaluating un balance (static or coupled ) we measure phase difference between de and nde of shaft.if we have  no difference phase between x channel of sensors and have a 180 degree difference between y channel of two sensors.if i evaluate by x channel we have static unbalance but by y channel we have couple unbalance.

which one is correct ?

best regards       

Can you make an isometric sketch showing all four transducers and their mounting angles?

 

Attachments

Photos (1)
I would assume I understand what you meant. Lets number 1 at one end and 2 at the other end. You are saying the phase X1-X2= 0 degree while Y1-Y2= 180 degree right? In my notes I would say you might have either misalignment or dynamic imbalance. Check your X1-X2 & Y1-Y2, is it really 0 & 180 or in-between. How about X1-Y1 & X2-Y2? Take as well casing measurement for Horizontal and vertical & axial and see the difference in amplitudes.
Ralph Stewart posted:
aziz58 posted:

but for Y channels between NDE and DE of bearing i see 180 degree phase difference.

my question is , which kind of un balance i have on the shaft? static our couple

 

best regards 

 

Is this a center hung rotor or overhung rotor?

Thanks,

Ralph

 dear ralph

rotor is over hung . third stage of three stage centrifugal compressor .

spherical posted:
I would assume I understand what you meant. Lets number 1 at one end and 2 at the other end. You are saying the phase X1-X2= 0 degree while Y1-Y2= 180 degree right? In my notes I would say you might have either misalignment or dynamic imbalance. Check your X1-X2 & Y1-Y2, is it really 0 & 180 or in-between. How about X1-Y1 & X2-Y2? Take as well casing measurement for Horizontal and vertical & axial and see the difference in amplitudes.

dear spherical.

i have X1-X2=0 degree and Y2-Y1=180

X1-Y1=90 AND X2-Y2=90

BEST REGARDS

John from PA posted:
aziz58 posted:

dear all

as you know for measuring relative vibration by proximity sensor on journal bearing use two sensor located by two channel x and y

45 degree from each others. to evaluating un balance (static or coupled ) we measure phase difference between de and nde of shaft.if we have  no difference phase between x channel of sensors and have a 180 degree difference between y channel of two sensors.if i evaluate by x channel we have static unbalance but by y channel we have couple unbalance.

which one is correct ?

best regards       

Can you make an isometric sketch showing all four transducers and their mounting angles?

 

 

Attachments

Photos (1)
John from PA posted:

There are some dynamic reasons why this could occur but one of the first things I would do is a "loop" test to insure the data you are getting positively comes from the physical location of the probe.

Additionally, make sure your data is 1X compensated.

 

dear john

i should say that this shaft is flexible .(42000 rpm) third stage of tree stage centrifugal air compressor . we have trip on high vibration .

can you explain more about loop test   .you mean my data maybe unreal?

and how about 1x compensated? you mean i have to do slow roll compensated?

we have vibration value increasing by  increasing rotational speed  . 

aziz58 posted:
John from PA posted:

There are some dynamic reasons why this could occur but one of the first things I would do is a "loop" test to insure the data you are getting positively comes from the physical location of the probe.

Additionally, make sure your data is 1X compensated.

 

dear john

i should say that this shaft is flexible .(42000 rpm) third stage of tree stage centrifugal air compressor . we have trip on high vibration .

can you explain more about loop test   .you mean my data maybe unreal?

and how about 1x compensated? you mean i have to do slow roll compensated?

we have vibration value increasing by  increasing rotational speed  . 

I have to say that 42000 RPM is pertinent information that you should have provided when you started this thread.

What exactly are the amplitudes?  

Yes, you should be using compensated data with the compensation acquired from a shutdown and after the machine has run long enough to be thermally stable.

A loop test is done to verify that the data really comes from the proper probe.  If alarms are disabled, one way of doing that is to disconnect the probe at the proximitor (having traced the wiring from proximitor to probe) and make sure the appropriate channel on the rack responds.  This insures the probe wiring "in the loop" from the probe to the rack channel is correct.  I can tell you that with a new machine, or a newly rebuilt machine, an error in wiring is quite common.

If I read correctly,  one looks from an end of the machine at one x probe and the other x probe is on the same side of the machine, similar for the y probes.

If correct, then to have x1 and x2 in phase and y1 and y2 out of phase, one of the orbits has more (maybe totally) reverse orbiting.    Checking that the probes are connected to where you think they are is good advice before proceeding.  

What are the x1 to y1 and x2 to y2 phase relations?

William_C._Foiles posted:

If I read correctly,  one looks from an end of the machine at one x probe and the other x probe is on the same side of the machine, similar for the y probes.

If correct, then to have x1 and x2 in phase and y1 and y2 out of phase, one of the orbits has more (maybe totally) reverse orbiting.    Checking that the probes are connected to where you think they are is good advice before proceeding.  

What are the x1 to y1 and x2 to y2 phase relations?

we have 90 degree x1 to y1 and x2 to y2

John from PA posted:
aziz58 posted:
John from PA posted:

There are some dynamic reasons why this could occur but one of the first things I would do is a "loop" test to insure the data you are getting positively comes from the physical location of the probe.

Additionally, make sure your data is 1X compensated.

 

dear john

i should say that this shaft is flexible .(42000 rpm) third stage of tree stage centrifugal air compressor . we have trip on high vibration .

can you explain more about loop test   .you mean my data maybe unreal?

and how about 1x compensated? you mean i have to do slow roll compensated?

we have vibration value increasing by  increasing rotational speed  . 

I have to say that 42000 RPM is pertinent information that you should have provided when you started this thread.

What exactly are the amplitudes?  

Yes, you should be using compensated data with the compensation acquired from a shutdown and after the machine has run long enough to be thermally stable.

A loop test is done to verify that the data really comes from the proper probe.  If alarms are disabled, one way of doing that is to disconnect the probe at the proximitor (having traced the wiring from proximitor to probe) and make sure the appropriate channel on the rack responds.  This insures the probe wiring "in the loop" from the probe to the rack channel is correct.  I can tell you that with a new machine, or a newly rebuilt machine, an error in wiring is quite common.

dear john

does it need machine running when doing loop test? operation dont permit to alarm disable .BUT I TALK TO INSTRUMENT STAFF THEY SAID ALL DATA COME REAL 

we have amplitudes as following:

x1=1.9 micron(RMS)

Y1=2 micron(RMS)

X2=1.5micon(RMS)

Y2=0.8micron(RMS) 

aziz58 posted:
 
does it need machine running when doing loop test? operation dont permit to alarm disable .BUT I TALK TO INSTRUMENT STAFF THEY SAID ALL DATA COME REAL 

we have amplitudes as following:

x1=1.9 micron(RMS)

Y1=2 micron(RMS)

X2=1.5micon(RMS)

Y2=0.8micron(RMS) 

The machine can be stopped when doing the loop test because you are essentially looking at what happens to the DC gap voltage when the probe is disconnected.  Stopped is the preferred method, otherwise you have to worry about tripping the machine.

vibramac posted:

If you can display 1X orbits do so. Provided amplitudes are extremely low.  Why only amplitude values and no phase on your last posting!!!!!!!!

 

  

I agree amplitudes are low.  In fact I would be asking why they are so low and be doing checks to establish that they are valid.  For information, 2 µm (RMS) = 0.08 mil (RMS) = 0.22 mil p-p.

spherical posted:
John Can you please comment on my point below. When, at the moment keyphasor fires and the highspot is between the X & Y probes, it would reflect as the First Positive Peak comes first with Y before X.

Yes, providing the X and Y are at the correct orientations (ie Y=45L and X=45R) and rotation  is CCW

spherical posted:

That was my point. Actually I am asking why system 1 has not compensated this situation.

Are you referring to compensation of the data?  Since we haven't seen the data plots, only numbers, we don't know about what if any compensation has been applied.

we have 90 degree x1 to y1 and x2 to y2 - back to the reverse precession.  

If the x signals are in phase and the y phases are out of phase with x to phases being 90 degrees, one of the ends is showing reverse precession. 

If you check the physical location of the signals and they agree with what is labeled as the probe positions, then one of the ends has reverse precession.  This does not necessarily rule out balancing as an effective tool, but it is something to consider.

Add Reply

×
×
×
×