Practical Solutions to Machinery and Maintenance Vibration Problems
Chapter 7, Misalignment
Section 19, Phase and Vibration Due to Coupling Misalignment (as Compared to Unbalance as a Source)
Measured at the bearings at each end of the same rotor shaft, horizontal and vertical vibration modes as revealed by phase due to rotor unbalance will show the vertical vibration mode shaking in a similar manner as the rotor's horizontal vibration mode. The vertical phase angle of a rotor's left bearing can show a similar phase angle at the right bearing. (Vertical mode shake is in-phase.) For this part of analysis, a horizontal phase reading at one of the above specific bearings is not to be compared with the vertical phase reading at the same bearing. Instead, the horizontal phase angle measured at the rotor's left bearing should be compared with the horizontal phase angle at the same rotor's right bearing. In this example, the horizontal phase angle at the left bearing should be relatively in-phase with the horizontal reading at the right bearing. In other words, the vertical shaking mode will be similar to the horizontal shaking mode. In order to work up to an understanding phase and vibration due to misalignment, further consideration will be given to phase and vibration due to unbalance.
With unbalance vibration, the vertical readings at the left and right bearings could be, for example, 180° out-of-phase. Caused by unbalance, the horizontal left and right readings would be approximately 180° out-of-phase with each other. The vertical shaking mode would be interpreted as being similar to the horizontal shaking mode.
Unbalance as the major vibration source could also, for example, result in vertical left and right phase readings that are approximately 70° out-of-phase. The horizontal left and right readings would also be approximately 70° out-of-phase (within one "clock position" or 30°). If the vertical comparisons of left and right phase readings were, for example, 100° out-of-phase, the horizontal phase relationships of 70° would be close enough to the 100° out-of-phase that it would be within one "clock position." And this is close enough to be considered shaking in a similar manner, both vertically and horizontally.
Now that unbalance vibration modes are outlined, it will be easier to see what is different or similar about coupling misalignment vibration. Some of the most important points about coupling misalignment's vertical or horizontal shaking modes can be shown best by assuming a coupled drive and driven units that are perfectly balanced, with perfectly straight shafts, perfectly concentric armature and so on. The only error is coupling/shaft misalignment, and the symptoms are as follows.
To make the differences between 1 x rpm unbalance symptoms and misalignment symptoms more clear, consider the phase relationships of readings obtained at one end of a rotor and at the other end of the rotor -- measured in the same direction. Assume that the 1 x rpm vibration measured in one direction is due only to unbalance. Consider the possibilities.
For relatively pure static unbalance, the phase relationships will be approximately in-phase (within 30°). For relatively pure couple unbalance the phase relationships will be approximately 180° out-of-phase. Obviously, unbalance could result in phase relationships that are either approximately in-phase or approximately 180° out-of-phase. However, depending on the unbalance for that specific rotor, the phase relationships from one end of the rotor to the other can be any possible combination, such as 45°, 92°, 125°, 71°, 310°, 183°, 60° and so on. Whatever phase relationships appeared in one direction (such as vertical) would approximately be repeated in the other direction (such as horizontal).
The 1 x rpm forces originating from pure misalignment are different. They can also cause relationships that are 0°out-of-phase or 180° out-of-phase. For misalignment that is relatively pure, but not perfectly so, the only relationships would be approximately 0° or approximately 180° (within one "clock position" of 30°). Unlike unbalance symptoms, a machine that has very few sources for 1 x rpm, other than relatively pure shaft-to-shaft misalignment, will not have phase relationships other than approximately 0° or 180°. Therefore, phase relationships of for example, 70°, 45°, 125°, etc., will usually be due to unbalance. Phase relationships of approximately 0° or 180° can be due to the special situations of mostly static or mostly couple unbalance. Or, they are the typical result of misalignment.
Fortunately, for almost all machines with 1 x rpm sources primarily due to unbalance, the vertical shaking modes are very similar to the horizontal shaking modes. This refers to amplitudes as well as phase relationships. For misalignment, however, about 80 percent of the time would be one shaking mode in one direction (such as approximately 180° out-of-phase) and an entirely different shaking mode in the other direction (such as approximately 0° out-of-phase). This refers to amplitude as well as phase relationships. But what about the other 20 percent? For those, misalignment can cause similar phase relationships in both directions, making it difficult to separate the possibilities of unbalance from misalignment. However, in such situations, the comparisons of amplitudes in both directions will usually separate the symptoms. For example, unbalance as a source, will produce, at one bearing, relatively similar amplitudes in both directions. Although the amplitudes for the bearing at the other end of the rotor will likely be very different, they too will be relatively similar when measured in both directions.
In addition, unbalance as the main source (uncomplicated by resonance in one direction or the other) most often produces higher amplitudes in the horizontal direction as compared to the vertical directions. Most machines are more flexible horizontally than vertically. Unbalance can be diagnosed as the source when the horizontal amplitudes are larger than the vertical amplitudes by anywhere from only slightly to as much as twice as large. When the horizontal amplitudes are more than twice as large, there is reason to suspect a resonance in one direction and not the other, or that the source is due to misalignment.
This textbook contains only part of the information in our Practical Solutions seminar.