Correctly referred to, it is precession, not precision.

I would first of all runout compensate the filtered data.  Your amplitudes are relatively low so any runout present may have quite an impact on the data.

On your statement "Physical cross connection of probes between X and Y is unlikely to me."  Do a loop test, then you will know for sure.  Never assume anything unless you personally have witnessed a verification test.  

Your logic is not good for having x and y switched.  What John proposed should have been a first step, even before posting.  Generally, this is the most likely possibility.

as far as I know 1X filtered orbit should always be forward precision only

This can occur without a having a fault, but one would look for a reason.

precession

it will be always forward in normal working speed, only if there is rubbing.

but, the rotation direction and the ‘X,Y’ probe location must be correct, there is a easy way to distinguish whether the backward precession is real, the rubbing is Instantaneous, not-stable, the vibration 1X amplitude and phase will wave in short time when rubbing occur. if there is a backward precession and vibration is stable then it‘s the setup problem. 

I guess even some rubbing cases can lead to forward precession.

Regards- Ali M. Al-Shurafa

Shurafa posted:

I guess even some rubbing cases can lead to forward precession.

Regards- Ali M. Al-Shurafa

One thing I've learned over many years is never rule out some unusual phenomenon but in my experience a rub with forward precession is likely rare.  But a rub is usually a secondary malfunction and the primary malfunction may change the rotor behavior.  When the contact is made between the rotor and stationary object a tangential force is developed due to friction.  The tangential friction force acts opposite to the surface velocity of the rotor.  There will be induced a torque on the rotor opposite to the direction of rotation, thus acting to slow the rotor.  There will also be a movement of the shaft centerline in the reverse precession direction.  In a direct orbit exhibiting a partial rub, this will often be indicated by an inside loop.  In my experience a rub will usually produce reverse components in the orbit or full spectrum plot formats.

There is also a situation of a machine with a split critical.  The observed precession can flip back and forth in the region between the "split".  Anyone who has observed "live" orbits using diagnostic instrumentation or an oscilloscope (what is that?!?!) knows that they change shape and flip around extensively as the machine accelerates and decelerates. 

Shurafa posted:

I guess even some rubbing cases can lead to forward precession.

Regards- Ali M. Al-Shurafa

yes，backward precession must be rubbing, rubbing not must have backward precession, depending on the severity of rubbing, slight rubbing only cause  the difference of two phase decrease. 

chitrang.nanavati posted:

Dear All,

In our machine train we have online BN 3500 and system 1. If we go to orbit plot of few bearings then I have noticed that 1X filtered orbit shows reverse precision and in few it shows forward precision. But as far as I know 1X filtered orbit should always be forward precision only

I have checked the direction of rotation in system 1 and it is consistent with actual direction of rotation

I have attached the screen shots of orbit plots. What could be the reason ?

Physical cross connection of probes between X and Y is unlikely to me. Because can not be in 5-6 out of 15-16 bearings

 

Can I see the Fullspectrum(Waterfall Fullspectrum) with Synchronous?

Anent the dynamics and rubbing, very often rubbing is forward, and dynamics of a system can and does result in backward precession.    Know your machines.

Newkirk (1920's or 30's) described a rubbing with forward precession caused by heating the rotor.  This is a common from of rubs in some machines.

Operating between critical speeds can produce backwards precession with asymmetric supports, which every machine has to some degree.  Know your machines.

William_C._Foiles posted:

Anent the dynamics and rubbing, very often rubbing is forward, and dynamics of a system can and does result in backward precession.    Know your machines.

Newkirk (1920's or 30's) described a rubbing with forward precession caused by heating the rotor.  This is a common from of rubs in some machines.

Operating between critical speeds can produce backwards precession with asymmetric supports, which every machine has to some degree.  Know your machines.

 

Be careful Bill, anent is considered archaic.  We don't want to give away our ages.

I see the following in the NYT, but you may consider them archaic, too.

Regarding; concerning: "This question remains a vital consideration anent the debate over the possibility of limiting nuclear war to military objectives" (New York Times).

I don't believe that MW has this as archaic (or OED?).  It does derive from middle-English.  ... Or perhaps I am getting old even archaic, myself.

I had a similar issues just recently. The 1X filtered orbit on the pump in a motor-pump system showed reverse precessive 1X vibration. This is in a plant that is currently in the commissioning phase and this was the first time the machine was run. Upon further investigation I found that the X and Y probes were installed in reverse from the manufacturer. To correct the issue, we swapped the X and Y wires at the transducers in the junction box as well as swapped the tags on the probes themselves. 

cguid32 posted:

Upon further investigation I found that the X and Y probes were installed in reverse from the manufacturer. 

Verification of the probe wiring through a loop test was one of the initial recommendations and the OP assured us that "Physical cross connection of probes between X and Y is unlikely".  I wish I was as confident as the OP seems to be.

A relatively simple test is to look at precession at slow roll, something down around 200 to 400 RPM.  If data shows reverse precession then, the likelihood of incorrect wiring is highly likely.

@Registered Member posted:

Correctly referred to, it is precession, not precision.

I would first of all runout compensate the filtered data.  Your amplitudes are relatively low so any runout present may have quite an impact on the data.

On your statement "Physical cross connection of probes between X and Y is unlikely to me."  Do a loop test, then you will know for sure.  Never assume anything unless you personally have witnessed a verification test.  

Dear Sir

I am working in inspection department and as the question above asked is same for me.

In our machine system 1. If we go to orbit plot of few bearings then I have noticed that 1X filtered orbit shows reverse precession and in few it shows forward precession. Our shaft center-line is at its desirable position orbit is consistent with slight oval shape displacements are well within ranges. As you prefer to have loop test, can you please explain how could i conduct it and weather it is possible while machine is running.

Thanks and regards

Doing a loop test while running should only be done by those very familiar with System 1.  First of all all trips have to be disabled so that you don’t inadvertently trip the machine.  Then you would trace a probe, say the “X” or horizontal probe to its respective proximitor and disconnect it.  You should lose the “X” probe reading within System 1.  Obviously if you lose something other than “X”, you have a wiring error.

An possible alternative to trace wiring is to measure the gap voltage at what you think is the proper proximitor based on tracing it from the probe, then compare that to the gap voltage indicated in System 1.

FYI, 1X filtered can appear reverse until properly compensated for slow roll.  This particularly true at low amplitudes where 1X runout is a substantial portion of the total amplitude.

Dear Sir Thanks for your valuable feedback we will definitely approach system 1 personal to get it rectified asap.🙏

@Registered Member posted:

Correctly referred to, it is precession, not precision.

I would first of all runout compensate the filtered data.  Your amplitudes are relatively low so any runout present may have quite an impact on the data.

On your statement "Physical cross connection of probes between X and Y is unlikely to me."  Do a loop test, then you will know for sure.  Never assume anything unless you personally have witnessed a verification test.  

Dear Sir

Is it possible that drive-end of compressor is showing forward precession in orbit and showing reverse precession in non-drive end. machine speed is 14500. please explain.

Thanks and regards

@Registered Member posted:

Correctly referred to, it is precession, not precision.

I would first of all runout compensate the filtered data.  Your amplitudes are relatively low so any runout present may have quite an impact on the data.

On your statement "Physical cross connection of probes between X and Y is unlikely to me."  Do a loop test, then you will know for sure.  Never assume anything unless you personally have witnessed a verification test.  

Respected Sir

Can you please let me know----If there is rubbing in any rotor within bearings which plots are helpful which may pin point this phenomenon.

Thanks and regards

Review the content at Rub Diagnostics based on Vibration Data

@Registered Member posted:

One thing I've learned over many years is never rule out some unusual phenomenon but in my experience a rub with forward precession is likely rare.  But a rub is usually a secondary malfunction and the primary malfunction may change the rotor behavior.  When the contact is made between the rotor and stationary object a tangential force is developed due to friction.  The tangential friction force acts opposite to the surface velocity of the rotor.  There will be induced a torque on the rotor opposite to the direction of rotation, thus acting to slow the rotor.  There will also be a movement of the shaft centerline in the reverse precession direction.  In a direct orbit exhibiting a partial rub, this will often be indicated by an inside loop.  In my experience a rub will usually produce reverse components in the orbit or full spectrum plot formats.

There is also a situation of a machine with a split critical.  The observed precession can flip back and forth in the region between the "split".  Anyone who has observed "live" orbits using diagnostic instrumentation or an oscilloscope (what is that?!?!) knows that they change shape and flip around extensively as the machine accelerates and decelerates.

Below attachment shows forward and reverse precession on same rotor but different ends. Is it possible? Or there is setup fault?

To your questions...Below attachment shows forward and reverse precession on same rotor but different ends.

Is it possible?  Yes

Or there is setup fault?  Yes, that can happen as well.

It is important that you do two things before establishing if the symptoms are a result of rotor dynamics.  One is to verify that the transducers are wired correctly and two make sure your data is properly compensated.

Thanks sir🙏

@Registered Member posted:

Review the content at Rub Diagnostics based on Vibration Data

I read this article sir

As it contains the statement that if orbit is not distorted much and even with mild rubbing, precession is always forward. In my case orbits are not distorted much , shaft centre line is in proper quadrant, not much changes in phase, polar plots and trend is almost uniform, so may I come to conclusion that there is setup problem in my probes?

From John:

To your questions...Below attachment shows forward and reverse precession on same rotor but different ends.

Is it possible?  Yes

Very true. Indeed rotor modes can be mixed as well with one end having forward precession and the other backwards.

Also, between modes one often finds reverse precession.

Far too often the probe configuration is not correct, and one does not have the precession that one thinks. Indeed one sees this on split criticals.

Very true thanks sir🙏

@Registered Member posted:

In my case orbits are not distorted much , shaft centre line is in proper quadrant, not much changes in phase, polar plots and trend is almost uniform, so may I come to conclusion that there is setup problem in my probes?

Don’t let these observations lead you to a false sense of security.  You’ve only presented us with 1X filtered orbits and to be honest I can’t read the amplitudes or determine if they are compensated.

@Registered Member posted:

Don’t let these observations lead you to a false sense of security.  You’ve only presented us with 1X filtered orbits and to be honest I can’t read the amplitudes or determine if they are compensated.

Ok sir I am sending you some plots for your kind consideration