Practical Solutions to Machinery and Maintenance Vibration Problems
Chapter 7, Misalignment
Section 9, Comparing Misalignment with Unbalance Symptoms
Over the last 26 years or so, most vibration analysts have considered a large amplitude at 1 x rpm as almost always due to rotor unbalance. Shaft-to-shaft misalignment was supposed to produce a symptom of high amplitude at 2 x rpm. However, with the increased use of vibration spectra of amplitude vs frequency, it has become easier to analyze the display of amplitudes throughout the frequency range. It was then found that most often (about 90 percent of the time) the largest amplitude produced by misalignment was not at 2 x rpm but, instead, at 1 x rpm. This revealed that unbalance was often blamed for vibration due to misalignment.
Another reason for misdiagnosis was that most misaligned shaft/couplings do not produce a high amplitude at 2 x rpm, but rather an amplitude often much smaller than at 1 x rpm. Yet, the amplitude at 2 x rpm, although not as large as at 1 x rpm, would be "larger than usual." This is best explained by specific examples as shown in the following diagrams. In all the situations but one, the amplitude at 1 x rpm is kept the same. All suggest the possibilities for symptoms of unbalance compared to misalignment.
In this diagram, the 1 x rpm amplitude is high, but the 2 x rpm amplitude is very small in comparison. The other possible harmonics are not visible. Unless there are other symptoms for misalignment such as revealed by phases, this machine can be diagnosed as out-of-balance. (It can also be due to some other source for 1 x rpm.)
The 2 x rpm harmonic's amplitude is still much smaller than that at 1 x rpm. However, it is "higher than usual." For a reference, "usual" is shown in the above example. Symptoms for misalignment start to become significant as the 2 x rpm amplitude becomes greater than 1/3 that of the 1 x rpm amplitude. Note that the 3 x rpm harmonic is almost not visible. This machine has a fair chance of being properly diagnosed as misaligned. But though the 2 x rpm amplitude is higher than in the preceding situation, it is still not very much higher. Therefore, this machine can also be properly diagnosed as out-of-balance. Both possibilities have equal chance of being correct.
It is suggested that the possibility for misalignment doesn't get strong
until the 2 x rpm amplitude gets greater than 1/3 that of the 1 x rpm
amplitude -- or if the other harmonics (such as at 3 x rpm) are also
higher than usual.
When the 2 x rpm amplitude reaches or exceeds ½ that of the
1 x rpm amplitude, then the diagnosis of misalignment is probably correct.
Although the amplitudes at 3 x rpm and 4 x rpm are considerably smaller
than those at 1 and 2 x rpm, they are much "higher than usual"
and, therefore, contribute to a strong diagnosis for misalignment.
This machine is producing much "higher than usual" amplitudes at several harmonic frequencies. The amplitude at 2 x rpm is almost as large as that at 1 x rpm. Between unbalance and misalignment, this machine is probably misaligned. (However, when the 2 x rpm is considerably "higher than usual" always determine if there is a resonance to that frequency. The defect might not originate from misalignment, but rather something else that creates 2 x rpm.)
Here the amplitude at 2 x rpm is greater than that at 1 x rpm. Misalignment doesn't always create a large amplitude at 1 x rpm. Sometimes it creates "higher than usual" amplitudes only at the lower harmonic frequencies but not at 1 x rpm. In situations where the 2 x rpm amplitude is higher than the 1 x rpm amplitude, the diagnosis of misalignment has the highest chance of being correct. However, as with all vibration analyses, it should be crosschecked with other symptoms such as revealed by phases, etc. Another possibility is that the misalignment was actually acceptable. The extra high 2 x rpm amplitude can then be the result of some part resonating to 2 x rpm, thereby magnifying the vibration at that frequency (and not the others).
A coupling with two segments per coupling half, such as a coupling
with a rubber cross and each coupling half with two jaws, will result
in considerably higher amplitudes at a frequency of rpm x number of
jaws (on each coupling half). For example, a misaligned shaft/coupling
with three jaws or segments per half coupling will produce much higher
amplitudes at 3 x rpm than are usual for that amount of misalignment.
The same is true for four segments producing 4 x rpm and so on. This
is also true for couplings with many segments or slots, such as on grid-type
couplings. Segment-related frequencies also provide an ideal frequency
for monitoring misalignment.
This textbook contains only part of the information in our Practical Solutions seminar.