During the field portion of a recent training class vibration readings were taken before and after the precision shaft alignment of a pump and motor. This was done with a hand held vibration meter with a single plane accelerometer. The readings were approximately as follows:
Motor NDE V-.07, H-.07, A-.11. Motor DE V-.07 H-.07
Pump DE V-.12, H-.12, A-.15. Pump NDE V-.15 H-.15
These numbers are not bad necessarily but would indicate some misalignment and some possible bearing issues in the pump. After recording the “as found” alignment readings using the Fixturlaser EVO, it was found to be out of alignment 100 mils vertically and 30 mils horizontally (1.0 mil = .001″). The corrections were made and the precision shaft alignment completed to the tolerance specifications for an 1800 rpm motor. There was virtually no looseness in the coupling prior to alignment and significant looseness post alignment.
The pump and motor were restarted and the noise level had greatly increased. It goes without saying that the vibration had also increased! The readings were approximately as follows:
Motor NDE V-.09, H-.09, A-.09. Motor DE V-.11, H-.11
Pump DE V-.18, H-.18, A-.18. Pump NDE V-.23, H-.23
So what happened?!
Here’s my take. First of all looseness in the bearings and the coupling became apparent, once the coupling strain was relieved after the alignment was completed. It is not known how long these machine components operated in this condition, but it was obviously long enough to cause significant wear in the bearings and the coupling. Therefore significant looseness in the rotating components after the shafts were collinear.
Note: looseness is not the cause of vibration, but the amplifier of vibration which explains the increase in vibration.
I was curious to know how different coupling types might react to looseness and found this interesting study. This particular coupling had 6 rubber pin inserts and the pump shaft was visibly oscillating back and forth in the pump.
Detecting misalignment using vibration analysis
A large majority of people who actively participate in CBM and Pro-Active maintenance programs will tell you that shaft misalignment will be indicated by higher running speed and/or twice running speed vibration frequency components with high axial vibration and a 180 degree phase shift across the coupling.
This is true only some of the time. Figure 1 shows several vibration spectral patterns on rotating machinery operating under misalignment conditions.
Figure 1. Various misalignment vibration ‘signatures’ with different types of flexible couplings.
Notice that the patterns are different and don’t always show running speed and/or twice running speed vibration frequency components (1x and 2x). Several controlled tests by several individuals over the past ten years have indicated that vibration spectral patterns can be different under similar misalignment conditions depending on the type of flexible coupling installed on the machinery and under certain conditions, virtually no vibration can be detected even under moderate to severe misalignment. There are several points that you should be aware of concerning the use of vibration analysis for misalignment detection :
– there typically is not a linear relationship between the overall vibration amplitude and the amount of misalignment (i.e. it is possible that the vibration could decrease as misalignment increases)
– it is possible for the vibration levels to increase after re-aligning a piece of rotating machinery.
– if the majority of vibration is occurring at multiples of running speed, ‘phase angle’ data is somewhat meaningless.
8 Comments
[…] This was done with a hand held vibration meter with a single plane accelerometer. The […] Read More Source: The Alignment […]
Excellent article!
Mac
With identical readings in both the horizontal and vertical readings I am suspect of the readings. Typically the horizontal and vertical differ because of the different stiffness levels between the two plans.
Motor NDE V-.09, H-.09, A-.09. Motor DE V-.11, H-.11
The relationship (identical) stayed the same after the alignment. Some more information is needed.
Like HP and RPM. As a PDM professional I would be after more information before making a judgment call on this equipment.
Thanks for your response, Dick! While stiffness does change often depending on the make-up of the machine, it isn’t an absolute. Correcting misalignment doesn’t always lessen the vibration, as I’m sure you know. It too depends on many factors, stiffness, shaft lengths, bearing types, shaft deflection, coupling type, and so on.
I think Mac’s observations are simply his, since Mac is a very intelligent guy, and a great mechanic, but not vibration analyst. Many mechanics expect that alignment will reduce the vibration, but sometimes it can get worse. If vibration levels do increase, it is easy to assume that the alignment made no impact on the reliability of the machine. I have seen many instances where precision alignment made a minimal change, but bearing frequency vibration could increase (due to reducing radial loads), Coupling-related vibration might increase (due to wear on the element), and seals start leaking (because the seal had been working as a third bearing).
So, high vibration doesn’t always equal misalignment, and precision alignment doesn’t always equal reduced vibration. Experienced vibration people like yourself know this, but not all mechanics, supervisors, or engineers do.
Dear Stan,
Interesting topic for me as elect Engr. I worried about my motor failures, let me know what u concluded relation between vibration & alignment being not true always. such incidents will be in what percentile out no. of incident a CBM guy encounter misalgnment issues generally finds thru axial vibration?
From Stan,
The correlation between motor vibration and misalignment is much more complex than can be answered here. It can be affected by the coupling type, speed, base, load, as well as motor quality. Magnetic vs. physical centers can also have an effect. It is our intention to set up a study on this over the next several months, and begin compiling data. Once we have some knowledgeable conclusions, we will publish them. Stay tuned!
Hi Mac,
in figure 1 I can see sin, cos and tan definition, but there aren’t various misalignment vibration ‘signatures’ with different types of flexible couplings.
Roberto, sorry for the confusion as the wrong image was in the blog. The image has been corrected and should make sense now. Thanks for the heads up!