Online Course for Practical Solutions Vibration Seminar Schedule Online Vibration Book Request More Information

Chapter 5, Unbalance

The trial weight will be in the correct position when the phase angle is returned its original position (indicated before adding a trial weight). However, don't work to get the phasemark returned exactly, but instead work toward decreasing the vibration amplitude.

To shift the reference mark or phase in a desired direction, move the trial weight in the same desired direction. If the trial weight is lighter than the original unbalance, on the next spinup the reference mark or phase will shift in the desired direction.

• If the trial weight is too heavy (even if only slightly too heavy), the phase will shift in a direction opposite that intended. For example, the trial weight may be shifted clockwise. On the next spinup the phase will shift counterclockwise, indicating that the trial weight is too heavy. (This phenomenon can be used to determine if the trial weight is too heavy or too light--simply by shifting it and determining which way the phasemark moves.)

• Even when the trial weight is too heavy, the phase will still indicate the proper direction to shift the weight. If the trial weight remains too heavy, and the weight continues to be shifted in the direction indicated, the weight will finally reach the proper position (the true light spot); but the reference mark or phase will be 180° opposite to where it was during the original run with no trail weight.

• When the final trial weights are still attached to the rotor and the vibration level is within tolerance, the trial weights have to be converted into "permanent" weights (or permanent weight removal). There is always some error in the procedure that can result in a slightly better or slightly worse residual unbalance. Rather than aiming at the target weight, the above knowledge can be used to bias the final balance weight amount and angular position to insure that the permanent corrections result in even lower vibration than in the last trial run.

• The phase and amplitude at the final trial weight run (that balanced the rotor) can now be compared with the original phase and amplitude for each plane. By using the knowledge above, determine which way to bias the final corrections, to make sure the next run to check the results will produce even lower vibration levels than obtained before the permanent corrections were made.

• If the trial weight is added at a partially erroneous angular position, the next spinup will show a shift in phase. For a given amount of error in both amount and angular position, the degree of phase shift will be dependent upon how close the rotor is to true balance. For a rotor that is relatively out of balance, the phase shift will be slight as compared to the phase shift that results when the rotor is relatively close to true balance. Unfortunately, adjustment for angular corrections are based on phase shifts early in the balancing process, which usually produces relatively small phase shifts for a specific trial weight angular shift. However, if you don't adjust your thinking as you get closer to true balance, then you will be over shifting the trial weight, thereby making very fine balancing difficult and time consuming. Simply remember that phase shift, as a function of trial weight angular shift, becomes increasingly sensitive as you get closer to true balance. • Occasionally, accurately following a method of balancing simply does not work. Each test
run seems to produce results that are considerably off as compared to reasonably expected results. sometimes one run will indicate that the trial weight should be moved clockwise,
then the next run counterclockwise, and it is felt that this change is not due to heavy handed over correction.

In such situations, it is neither the fault of the system nor operator error, but rather some part of the rotor's mass shifting slightly from one angular and/or radial position to another. A heavy rotor's radial shift of just a few mils or microns is enough to drastically alter the results. The method for clamping the rotor to its shaft may be tight enough to go unnoticed when not rotating, but the shock of startup or braking may cause the insufficient clamping force to allow the rotor to shift radially a few mils. Some analysts assume that if the rotor is keyed to the shaft, the rotor cannot shift. While it cannot shift angularly, the shaft can still shift radially, creating a change in balance. For hollow rotors, also check for loose rust particles or scale. In other instances, a stray nut or bolt inside a space in the rotor may seek a different position with each spinup.

To summarize: a rotor can be mounted tightly on the shaft but not tightly enough for all conditions during operation. There may be stability for two or three spinups in a row and, suddenly, instability in the next spinup. In these situations, the rotor can appear to be mounted tightly, but is not tight enough for all conditions.

To test for the above unstable situation, remove all the trial weights, and test run the rotor again. If the phase and/or amplitude changes very much for the original readings, then one of the above conditions most likely exists.

• Sometimes in field balancing coupled rotors, the 1 x rpm troublesome vibration is not due primarily to rotor unbalance but instead to shaft/coupling misalignment. If balancing is attempted, usually the vibration phase and magnitude are measured at the place of highest reading. The usual procedures are then used to reduce the vibration amount. Assume that for balancing, the direction of the pickup is horizontal. But in such situations, as the horizontal vibration is decreased through balancing, the vertical vibration reaches a point whereby it actually increases. In other words, balancing "corrects" the vibration in one direction but increases the vibration measured at the same bearing, in the other direction (90° away). This indicates the source is not unbalance.

• Sometimes the 1 x rpm vibration phase oscillates back and forth, making it difficult to read the phase angle. This is most probably due to a "beat." (See "Effect Of A Beat On Phase.") However, the beat that can create the most balancing difficulty is one with such a low beat frequency that each cycle takes several minutes. If the phase is read too quickly, it will not appear to be oscillating. Therefore, the technicians may not notice that the phase changes even though no balance corrections were made.

Textbook Index

This textbook contains only part of the information in our Practical Solutions seminar.

Link to Seminar Schedule. Order a print version of this entire textbook.