menu 9

Online Course for Practical Solutions Vibration

Seminar Schedule

Online Vibration Book

Request More Information

Practical Solutions to Machinery and Maintenance Vibration Problems

Chapter 5, Unbalance

Section 11, Unbalance in Couplings

As most primary rotors, such as armatures, fanwheels and pump impellers are dynamically balanced, one of the worst remaining sources for creating larger than necessary amplitudes at 1 x rpm, is coupling unbalance. Mediocre to poor coupling balance is probably the weakest link of the various coupled rotors. For example, no one questions the need to dynamically precision balance the armature on one side of the coupling, or to balance the driven unit, whether it is a pump impeller or fan, on the other side of the coupling. Yet, almost all who require precision balancing of such rotors, fail to require the same balance accuracy for the coupling. Coupling manufacturers (mostly located in North America) have convinced themselves and their customers that most of their couplings do not usually require dynamic balancing at all. As one major coupling manufacturer recommends, unless certain speeds, bore size and hp are exceeded, there is no need for dynamic balancing. For example, at 1800 rpm this manufacturer does not recommend balancing at all unless bore size is over nine inches in diameter and the hp is over 5000! According to this coupling manufacturer, only the largest, highest speed special machines would require dynamically balanced couplings.

The reasoning seems to be based on the premise that the couplings are machined from homogeneous steel with no more than a few more mils of steel removed from one side of the periphery as compared to the other. Obviously, the weight of a few mils of steel couldn't possibly create enough unbalance to cause concern. True! However, that is not what creates unbalance in the couplings. Instead, unbalance is the result of the coupling bore not being perfectly centered or being off angularly. If the coupling mass is located only a few mils off-center, it creates static unbalance. If the bore angle is not square with the coupling faces within a few mils, it causes couple unbalance.

The resulting static unbalance may not seem to be large enough for concern. The static unbalance centrifugal force may be in the plane of the coupling's center of gravity, causing no couple reaction within the coupling itself. However, couplings are most often mounted overhung compared to the primary rotors. The total assembly's center of gravity is usually an appreciable distance away from the coupling's static unbalance centrifugal force. This results in a couple unbalance, which is determined by the unbalance oz•in or g•mm times the distance to the plane of the total rotor assembly's center of gravity. As a coupling's static unbalance is at a relatively large distance from the armature's CG, it usually results in vibration from a couple unbalance that is considerably larger than that caused by the original static unbalance that created it.

The information above refers to couplings manufactured in or influenced by North American thinking. For other parts of the world, such as Europe and those influenced by European thinking, there was a different story (notice the past tense). Balancing the main rotors, such as armatures, pumps, impellers, etc., by European manufacturers required that they be balanced with keys of full length and full cross section. North American main rotors were balanced with full length keys, but with only half its cross sectional height (half keys). Half the key was assumed to be part of the main rotor's balance and the other half key was assumed to be part of the minor rotor's balance. Therefore, the coupling was assumed to be balanced when the key was in place. (The assembly precautions relative to keys are covered in the section on "Key Length Considerations" and will not be repeated here.) However, the European method is to balance with all of the key on the main rotor and no key on the fitment, such as the coupling. As a result, a coupling's empty keyway required that each and every coupling be dynamically balanced on a balancing machine. Balanced European main rotors assembled to dynamically balanced fitments resulted in a total balanced combination. Many North American couplings probably were within the balance tolerances of ISO-G 6.3 and G 2.5, even though they were not balanced in a dynamic balancing machine. A reasonable percentage did not pass the tolerance requirements, but the machinery users were accustomed to the typical resulting vibration levels. However, almost all machinery users are now under pressure to obtain the longest production running times between bearing or seal changes. Precision balancing of couplings is an absolute requirement if precision balancing is required of the main rotors.

Although most participants in Update's vibration seminars are convinced of the need for precision balanced couplings, others at their plants may have difficulty understanding the need when it appears to them that most of the plant's machinery seems to be running okay. The following demonstration is therefore suggested:

1. Precision balance a rotor such as a motor armature to ISO G 1.0 or the API Standard (using the key length procedures as described in the section "Key Length Considerations").

2. After the motor has been fully assembled, place on a rubber mat. Complete the electrical connections. Add half key that is safely taped in place.

3. Run the motor and record the vibrations at the frequency of 1 x rpm. This is to avoid measuring amplitudes at frequencies not related to unbalance, such as electrical hum ( 2 x lines frequency).

4. If the 1 x rpm is well within precision tolerance, such as is called for by the API, continue on to next step. If not, perform a trim balance in-situ (in-place).

5. Using proper key length, attach a coupling half to the armature.

6. Run the motor and record the new amplitude.

Most often the coupling will increase the total assembly's unbalance and, therefore, the amplitude at 1 x rpm. However, this may not always be so. In some situations, the addition of the coupling may actually improve the amplitude at 1 x rpm. It must be remembered that the residual unbalance in the armature is represented by a vector. The unbalance in the coupling half or any other fitment, such as a pulley, is also represented as a vector. Vectorial addition is different from that of regular arithmetic. The two vectors are "added vectorially," which means that the resultant due to both could cause an increase, decrease, or even remain the same (as if the coupling had no unbalance, when actually it does). As a result, it would be best to perform this test with several couplings, rather than just one.

Another test can be conducted on a balancing machine. Using a balancing machine, simply check the dynamic balance of at least ten new couplings before they are assembled to their main rotors. Determine what percentage is already balanced to precision tolerances. The writer has enquired about this point with balancing machine operators in several parts of the world, who are actually precision balancing each and every coupling. Out of at least 50 operators, only one has ever indicated that a coupling did not require at least a trim balance. And for that one operator, it was for only one coupling! Many plants are already precision balancing coupling halves (such as requiring that each half coupling be mounted on its accompanying main rotor while it is being dynamically balanced). Yet, they are not balancing the coupling's loose piece, such as what is often called the "spool piece." They erroneously assume that, for example, spoolpieces will center themselves when running via the proper mating of gear teeth. However a rotor that is running "free-in-space" does not necessarily rotate about its centerline, but instead rotates about its "axis of rotation." The axis of rotation will be located at the spool piece's "center of mass" (see "Vibration Phase Relative To Resonance.") If there is any unbalance in the spoolpiece, the center of mass, and therefore its axis of rotation, will not coincide with the spoolpiece centerline. Therefore the spoolpiece will not center itself based on the gear teeth. The result will not only be increased vibration that is transmitted to both main rotors, but also an increased wear of the coupling itself. Loose pieces have to be balanced to the same tolerances as coupling halves.

It is not the purpose of this section to show how to mount loose pieces on balancing arbors and fixtures except to indicate that there are good balancing shops that are already using such tooling. In some situations, the coupling manufacturer may assist (although not believing in its necessity).

For simple couplings such as "rubber tire" type or those with a flexible portion of sheet plastic between coupling halves, a simple straight balancing arbor is thrust through both halves while the "rubber tire" or plastic remains mounted in place. Weight corrections are made on each coupling half as if they formed one rotor.

For maximum machine life, precision balancing and precision alignment are all-important. Many vibration troubles originate at the coupling. Couplings that are precision balanced, precision aligned and properly assembled with all the necessary precautions will probably contribute to increased bearing, seal and coupling life as much as or more than just about anything else you can do.


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.

Home PageEmail Update