Sorry, i'm very new to this field, and i can't comprehend why misalignment can cause the 2x freq goes up dominantly?? i try to make a picture of a simple parallel misalignment and can't make out how it can make an "M" or "W" pattern in time domain and thus making the 2x spectra higher. Anyone can describe it to me??

Thx in advance!

Opiq,
If you have a very well defined "M" or "W" in your TWF, chances are you have an imbalance condition. The 2X turning speed would typically be a different issue unless the imbalance is extreme enough to bow the rotor shaft. But as I said earlier, a well defined M or W most of the time (a very high percentage) will indicate imbalance. I could be wrong...this is such an artistic science..
There is a good chance you have several things going on at the same time, especially if your starting a new program where one never existed. A well defined pattern in the TWF as you described coupled with a dominant 2X turning speed might suggest both misalignment and imbalance. Check and correct the misalignment first. If the 2X diminishes in the spectra and you have a persistent well defined M or W in the TWF look towards balancing.
Good Luck!

First of all pure angular misalignbment creates a 1X predominantly in the axial direction and pure offset misalignment creats a 2X predominantly in the radial direction. In the real world typically there exist both types of misalignment so hence charts that somewhat lead you astray. 2X doesn't always mean misalignment and 1X can mean misalignment. Don't always believe the charts!

Years ago I extracted the attached from a source I know longer remember.

John from PA

misalignment.pdf (110 Kb, 251 downloads)

I agree it may not be as simple as the charts. I believe misalignment might show as 1x, 1x & 2x, or might not show at all.

One mechanism we can visualize for production of 2x would be if the stiffness varies twice per revolution. In particular for a shaft with a keyway: Vertical stiffness is lower when keyway is at 12:00 and 6:00, higher when keyway is at 3:00, 9:00. Stiffness has a component varying twice per revolution.

hoo,, i see... it's somehow not exact isn't it? thx for the reply all..

i really interested in the pdf John from PA attached,, i understand the angular misalingment, but can't tell what's going on in the offset,, err,, can someone "narrate" it for me?

i understand if the stiffness is varying. But why it must varying twice a rev, not three or for times?

I disagree with the premise. The effects depend upon the machine (coupling included). ElPete'ws statement regarding stiffness variations is one leading cause.

2X is not common on some machines.


John,

I don't see the point of your pictures.

I agree with John From PA . This document was earlier available on SKF website . Document no-CM5003(Vibration Diagnostic guide) but now it is not available. I have still got a hard copy of the same.

Yes, that explanation showing a pin in place of a coupling floats around in a lot of training materials such as attached (starting at the middle of the page).

Similar to Bill, I don't see the logic of the explanation. If the coupling were symmetrical, it does not act like a pin rotating around at some radius from the shaft centerling as shown in the figure.

Perhaps if there were some non-uniformity in the rotating coupling or its attachment to the shaft, then it could make more sense. Some clarification is needed - what type of non-uniformity or why do we assume that a coupling is going to act like shown in the figure.

This message has been edited. Last edited by: electricpete,

MisalignmentPin.pdf (63 Kb, 111 downloads)

PETE, I will try to explain the 2x vibration as described by John From PA. Let us suppose a case of offset misalignment. When the two couplings are at 12 O' clock position , one coupling will be higher than the other.One coupling will try to pull the other downwards and the other will try to push it upwards causing maximum vibration in adjacent bearings. Same event will occur after 180 degree rotation. But , this time the scenario will be reverse. So, once every revolution, both couplings will be pushed/pulled by maximum force causing maximum vibration in adjacent bearings. Hence 2X vibration. Hope, my explanation is plausible. Would like to get responses.
Regards.

This message has been edited. Last edited by: AKHTAR,

quote:

I will try to explain the 2x vibration as described by John From PA. Let us suppose a case of offset misalignment. When the two couplings are at 12 O' clock position , one coupling will be higher than the other.One coupling will try to pull the other downwards and the other will try to push it upwards causing maximum vibration in adjacent bearings. Same event will occur after 180 degree rotation.

Misalignment has constant forces. Coupling offsets could have reversing forces. Misalignment of machine centerlines, other than thermal effects, is in one direction.

I have to agree with Bill and EPete. And, my reality experiences do not agree with the suppositions eluded to in the drawing and/or statements. Experimental setups of misalignment (real machines; not a rotor kit) also do not agree with what I've seen. I think one could generate a situation to mimic and some lending to this idea exist from mechanical readings but I think mostly vaporize like saying imbalance occurs at 2X.

Excuse me!
what is M or W timeform pattern?

Ricardo Góz from Brazil

quote:

Misalignment has constant forces. Coupling offsets could have reversing forces. Misalignment of machine centerlines, other than thermal effects, is in one direction.

Bill, I don't believe that this is true for all coupling types and situations. The misalignment force can change with load (torque) in couplings with friction elements such as those with elastomers and gear teeth. I don't have any data handy to back this up, but it is based on my experience. In most of my shaft misalignment cases, the vibration frequency was at 1xSS and not at 2xSS. I typically measure vibrations before and after I do a laser alignment, so I have good feedback on both the cause and the result of correction.

Walt

quote:

Excuse me!
what is M or W timeform pattern?

A pattern in the waveform that someone somewhere sometime has assigned to being a misalignment signal. Which may be real and then again may not be real, so I wouldn't trust it as being chilsed in stone as being misalignment.

I certainly agree that the existence of 2x in a spectra influenced by misalignment is not predictable. Neither, I believe, is any relationship between vibration and misalignment...as vibration amplitude may decrease under its effects. When...occasionally...there appears to be a relationship, I assume it is the result of how each radial transducer interprets the casing or shaft orbit at the bearing under the influence of misalignment. The orbit rules that help us diagnose misalignment explain that a narrowing of the orbit ellipse, or even a peanut shape or external loop, are indicators of misalignment. I have always assumed that the axis of misalignment coincides with the minor axis of the resultant ellipse. Why there's ever a peanut shape or external loop...I dunno. But I've seen it, and normally assume it to be evidence of misalignment.

Given all that...if a transducer happens to be oriented such that it sees a flattening of the elipse, it may view this as truncation of the sinusoid, causing the FFT to interpret a 2x. Or...if it views a peanut shaped orbit or external loop, it would likely display a 2x as well.

A recent gradual change in our drip pump motor vibration revealed an increase in overall amplitude due to 1x in the horiz direction, and a "sharing" of 1x to 2x in the vertical direction without any overall increase. At a subsequent maintenance opportunity, the coupling was cleaned and some misalignment corrected...vibration returned to baseline values.

My discussion of the orbit rules and resultant effects on radial vibration is strictly my opinion and tested by only my personal experience. I would be interested if others agree or have similar experience.

Thanks in advance,
George

quote:

Bill, I don't believe that this is true for all coupling types and situations.

My first remark mentions couplings. Gear couplings could be expected to show 2X. Luckily, many of these have disappeared over the years, magic or for good reason. Axial vibration is also seen in gear type couplings as well as load/torque effects.

Pulsating torque and on and on, as mentioned earlier one must know their machine. Often in a forum such as this people present a spectrum or 2, perhaps with 2X, and mis-alignment is diagnosed without much knowledge of the machine. It is not that simple.

Take your 2-pole electrical beasty (genus: synchronous), shaft to shaft mis-alignment should and probable does affect the 2X vibration, but how can you tell with the natural 2X present from gravity.

Take couplings fitted with keys and keyways, this should and does produce 2X. Why? Because of the shaft asymmetry, machine design. Fit that machine with a hydraulically dilated diaphragm coupling, I wouldn't bet on 2X for that because of mis-alignment.

OK, one can get very large mis-alignment to the point of being harmful or destructive. Non-linearities in the bearings could cause 2X -- or wreck the bearing first.

When I first saw pictures explaining axial vibration from mis-alignment, I was perplexed. It never made sense to me and still doesn't make sense. Often pictures mislead,even photographs.

quote:

Originally posted by Ralph Stewart:

quote:

Excuse me!
what is M or W timeform pattern?

A pattern in the waveform that someone somewhere sometime has assigned to being a misalignment signal. Which may be real and then again may not be real, so I wouldn't trust it as being chilsed in stone as being misalignment.

To add to the confusion... A "well defined" M or W pattern in the TWF would certainly infer imbalance more so than miss-alignment. I do agree the pattern can not always be trusted in it's purity but when considered/analyzed with an absence of other know forcing frequencies relative to that machine, does provide a valid visual starting point. The attachment is a case in point.

Irvin Redmond has an interesting article at
http://www.saudiaramco.com/irj/go/km/docs/SaudiAramcoPu...haftMisalignment.pdf and a scary picture .

Would the scarey picture be that of himself or all the algabra? Way over my head.....

Bill, isn't that your old unit in Daharan? Interesting; but, once I completely digest (if ever) and hash out some things I saw - I may have odds in a spot.

I've seen on real-world models --- motor and pump setup on its skid in the shop where you align perfectly - no askewed couplings, and then introduce misalignments of known increments for both angularity and parallel offsets. Move 0.005" and take vibration readings. In the grossly misaligned machine you can see 2X & 3X disappear leaving only 1X at higher magnitudes of vibration.

This is not disputing his paper or his findings and who can do it all or has time. It appears his model was based on a particular coupling.