Practical Solutions to Machinery and Maintenance Vibration Problems
Chapter 7, Misalignment
Section 5, Misalignment Vibration at 1 x rpm of Rotor
The most common major vibration amplitude due to misalignment is at a frequency of 1 x rpm. It sometimes is accompanied, or even exceeded, by vibrations at other frequencies such as at 2 x rpm. But for about 90 percent of misaligned machines, the major vibration symptom is at 1 x rpm. Unfortunately, 1 x rpm is also the vibration frequency of other common vibration sources, such as from unbalance, bent shaft, eccentric armature, eccentrically running gear and so on. To determine if the 1 x rpm vibration originates from misalignment, or for example, unbalance, phase as well as amplitude readings must be obtained so that vibration mode (how the part is shaking) can be determined. When the shaking mode is known, it is easier to determine why the part is shaking.
To better understand vibration due to shaft/coupling misalignment, it is best to first review some of the symptoms of vibration due to unbalance. Measured at the same bearing, unbalance usually results in relatively similar amplitudes at 1 x rpm. If the only machine defect is due to unbalance, and there are almost no forces at 1 x rpm due to misalignment, bent shaft, etc., then the vertical amplitude at a specific bearing is usually a bit smaller than the amplitude in the horizontal direction. Although most machines are much more flexible in the horizontal direction than they are in the vertical direction, the amplitude in the horizontal direction is usually not very much larger than in the vertical direction. At the same bearing, the vertical amplitude can, for example, be four units and the horizontal amplitude six units. There is question as to how much larger the horizontal amplitude can be relative to the vertical amplitude and still fit within the symptoms for unbalance. Update estimates that unbalance can still be the cause up until the point where the horizontal amplitude is double the vertical amplitude. (All such numbers are necessarily only rough guides.)
Rotating forces due to unbalance are the same in the vertical and horizontal directions. Shaft/coupling forces due to misalignment are almost never the same in both directions. For example, there can be large misalignment in one direction and relatively good alignment in the other direction. Or, there can be primarily angular misalignment in one direction and primarily parallel misalignment in the other direction. If misalignment causes relatively equal forces in all directions, it is an extremely rare coincidence. But for unbalance, equal forces in both directions are the norm.
For a machine with only unbalance as its vibration source (at the same bearing), it is very unusual for the vertical amplitude to be larger than that in the horizontal direction. If the vertical vibration amplitude is appreciably larger than that in the horizontal direction, the source can still originate with unbalance, but something else will be magnifying the amplitude in one direction as compared to the other. Most likely this will be due to the magnification of resonance. If resonance is not the reason, the appreciably larger vertical amplitude will likely be due to shaft/coupling misalignment.
Phase comparisons reveal shaking modes. Shaking modes give by far the best symptoms for coupling misalignment, as compared to other 1 x rpm vibration sources, and should receive the most weight for judgment. Other symptoms, such as axial vibration and comparisons of amplitudes at different harmonic frequencies, are also good symptoms but should not receive as much weight as shaking modes. For example, if shaking modes indicate coupling misalignment and axial vibration amplitudes do not, then give more credence to the shaking modes. The section on axial vibration will show that higher than usual axial vibration can also originate from other sources or may be magnified by resonance.
For misalignment, there is no relationship between the way a rotor vibrates vertically, as compared to the way it vibrates horizontally. In fact, with coupling misalignment, it seems that the vertical mode very rarely is similar to that of the horizontal mode. For example, the rotor can be "rocking" vertically (each end shaking approximately 180° out-of-phase) and translating horizontally (each end shaking approximately in-phase). This makes sense since there is no special reason for the vertical misalignment to resemble the horizontal misalignment. Note the two kinds of misalignment shown in Fig. 1. The same shaft can, in the vertical direction, be in angular misalignment and, in the horizontal direction, be in parallel misalignment. Or, there can be both angular misalignment and parallel misalignment at the same time.
In fact, not only can the vibration modes be very different, but the amplitudes can also differ considerably. For example, at a specific bearing, the horizontal vibration may be 6.0 mils, whereas its vertical vibration may be almost zero. All vibration data in the diagrams are at the frequency of 1 x rpm.
Looking at Fig. 2, the vertical amplitudes of 7.1 at the left and 2.2 at the right end may have originated from unbalance. If from unbalance, the left end of the rotor will have much greater unbalance than the right end. However, when observing the horizontal amplitudes, the right end shows considerably more amplitude than the left end. In other words, the vibratory motions of the rotor are very different between the vertical and horizontal directions. "Very different" points to misalignment as the source, rather than unbalance.
For Fig. 3, the same amplitude numbers were used as in Fig. 2, except that the horizontal amplitudes for the left and right ends were exchanged. In Fig. 3, the symptoms may easily have originated from unbalance. However, they may also originate from misalignment. To determine which one it is with a very high degree of accuracy, phase data must be used. Spectral data, as will be discussed later, is also very valuable. When amplitudes are very much larger in one direction compared to the other, do not jump to the conclusion that the source must be due to misalignment. Instead, the analyst should first determine if the vibrating machine has other reasons for large amplitudes in only one direction, such as:
• resonance in one direction, but not the other
• vibration due to an eccentric gear, or a gear running "eccentrically"
• machine's structure is very rigid in one direction and very flexible in the other direction (this is not too common)
Another precaution: ordinary reasoning can lead to the assumption that a high amplitude of horizontal vibration is caused by large horizontal misalignment. Actually, in many situations, the horizontal amplitude readings are primarily the result of the vertical misalignment. In like manner, the vertical amplitude readings are primarily the result of the horizontal misalignment. However, in many other situations, vertical misalignment does cause high vertical vibration, and horizontal misalignment does cause high horizontal vibration. As this confuses vibration experts as well as those new in vibration analysis, it is suggested that the analyst consider the most likely possibilities for misalignment error in one direction vs the other. For most ordinary machines, such as motor/pumps, fans/blowers, etc., there are usually more misalignment mistakes made that affect vertical misalignment. Some common errors that create vertical misalignment are:
• ignoring, or wrong figures for, thermal growth
• sagging of fixture parts due to gravity
• poor shimming/soft feet
• non-repeatability of initial indicator or laser readings when rotating fixtures full 360°
In fact, not only can the vibration modes be very different, but the amplitudes can differ considerably. For example, at a specific bearing the horizontal vibration may be 6.0 mils, whereas its vertical vibration may be almost perfect, resulting in only 0.2 mils.
This textbook contains only part of the information in our Practical Solutions seminar.