Alignment Basics, Part Two Ask
2
Ask the experts Q: What causes misalignment? Alignment Basics, Part Two Vol. 9, July 2003 A: Have a question for the experts? Contact us at emersonmotorhelp @usmotors.com continued on back > In theory, machine alignment is a very straightforward process. With some type of measuring device extended across the coupling, the shafts are rotated to several posi- tions (at least three) to determine the relative position between them. Since alignment is an iterative process (meaning that the mis- alignment should continuously decrease with each machine move), it is theoretically only a matter of sufficiently repeating alignment corrections until an acceptable solution is achieved. In fact, quality alignment is not dependent on the type of meas- urement system used. Any good dial indicator set or laser system should be sufficient to perform quality alignments. Therefore, in heavy industrial applications, where the cost of downtown can be in excess of $10,000 per hour, the fundamental question for an alignment program is not simply “Can I successfully align the machine?” but rather “Which method will provide the fastest alignment solution so that I can start production again?” Furthermore, since misalignment is often compounded by structural faults such as ‘soft foot’, piping strain, induced frame distortion, excessive bearing clearance, shaft rub, etc., it may not be possible to align the machine without first addressing these additional prob- lems. These pitfalls can turn an otherwise simple alignment job into an all day affair - frequently with unsatisfactory results despite conscientious effort and a consid- erable investment in manpower and downtime. For this reason, it is crucial for the personnel performing alignments to be aware of the kinds of structural faults that can complicate the align- ment process and that they learn to recognize the tell-tale signs of bad measurements before they invest valuable downtime in an unproduc- tive exercise. COLLECTING VALID DATA Some fairly simple yet powerful techniques can be applied to deter- mine the validity of alignment read- ings before investing time execut- ing a machine move that may be wrong. If using a dial indicator set, it is useful to apply the data validi- ty rule to each set of readings. The data validity rule compares the Installation (human) errors, thermal growth, dynamic twist of bases under load, settling bases and/or foundations. Machinery alignment is a process. The process begins with Pre-Shutdown and Pre- Measurement preparation. The pre-shutdown preparation should include an inspection for machinery problems. Any problem that can affect the quality of the alignment should be corrected before the align- ment is attempted. This inspection should include: Complete visual inspection of the machine including the foundation, baseplate, bolts, welds, etc. Vibration and phase analysis of the machine. Motor electrical data for later efficiency calculations. Thermal temperature readings or infrared thermography on the machine.
Transcript of Alignment Basics, Part Two Ask
Ask theexperts
Q: What causesmisalignment?
Alignment Basics, Part Two
Vol. 9, July 2003
A:
Have a question for the
experts? Contact us
at emersonmotorhelp
@usmotors.com
continued on back >
In theory, machine alignment is avery straightforward process. Withsome type of measuring deviceextended across the coupling, theshafts are rotated to several posi-tions (at least three) to determinethe relative position between them.
Since alignment is an iterativeprocess (meaning that the mis-alignment should continuouslydecrease with each machinemove), it is theoretically only amatter of sufficiently repeatingalignment corrections until anacceptable solution is achieved. Infact, quality alignment is notdependent on the type of meas-urement system used. Any gooddial indicator set or laser systemshould be sufficient to performquality alignments.
Therefore, in heavy industrialapplications, where the cost ofdowntown can be in excess of$10,000 per hour, the fundamentalquestion for an alignment programis not simply “Can I successfullyalign the machine?” but rather“Which method will provide thefastest alignment solution so that Ican start production again?”
Furthermore, since misalignmentis often compounded by structuralfaults such as ‘soft foot’, pipingstrain, induced frame distortion,excessive bearing clearance, shaftrub, etc., it may not be possible to
align the machine without firstaddressing these additional prob-lems. These pitfalls can turn anotherwise simple alignment jobinto an all day affair - frequentlywith unsatisfactory results despiteconscientious effort and a consid-erable investment in manpowerand downtime.
For this reason, it is crucial for thepersonnel performing alignments tobe aware of the kinds of structuralfaults that can complicate the align-ment process and that they learn torecognize the tell-tale signs of badmeasurements before they investvaluable downtime in an unproduc-tive exercise.
COLLECTING VALID DATA Some fairly simple yet powerful
techniques can be applied to deter-mine the validity of alignment read-ings before investing time execut-ing a machine move that may bewrong. If using a dial indicator set,it is useful to apply the data validi-ty rule to each set of readings. Thedata validity rule compares the
Installation (human) errors, thermal
growth, dynamic twist of basesunder load, settling basesand/or foundations.
Machinery alignment is aprocess. The process beginswith Pre-Shutdown and Pre-Measurement preparation. Thepre-shutdown preparationshould include an inspectionfor machinery problems. Anyproblem that can affect thequality of the alignment shouldbe corrected before the align-ment is attempted. Thisinspection should include:
Complete visual inspection ofthe machine including thefoundation, baseplate, bolts,welds, etc.
Vibration and phase analysis ofthe machine.
Motor electrical data for laterefficiency calculations.
Thermal temperature readingsor infrared thermography onthe machine.
product service BULLETIN
readings taken at the four cardinalpositions: Top + Bottom = Left +Right. It provides a quick way todetermine the validity of an align-ment solution before moving themachine. This simple check is ableto catch many set-up errors andmechanical faults such as:
• loose brackets • sticking indicators • indicators set too high or too
low • improperly recorded data values
and/or signs • sleeve bearing float • surface irregularities or
eccentricities • excessive bearing clearance
Small deviations from the validityrule are to be expected. If the dif-ference is more than 10%, it is pos-sible that the coupling may be looseenough to provide excess torsionalplay (“backlash”). To reduce theeffects of torsional play keep thecoupling engaged while rotating theshafts from the driven machine inthe normal direction of rotation.
If the error is greater that 20%,the cause should be determined.This could be a problem with thealignment fixture(s) or a concernwith the machine being aligned.Alignment problems occur fromloose fixtures or improper use offixtures. Possible machine concernsinclude locked couplings, spalledbearings, machine binds, etc. If thedata validity rule is not checkedwhen such a problem exists, thesepotential machine faults will remainundetected and substantially com-plicate the alignment process. Evenworse, the objective of increasingmachine reliability through qualityalignment will not be accomplished.
< continued from page one
When using a laser alignmentsystem, the potential for user erroris greatly reduced due to the auto-matic measurement and recordingof readings. However, the datavalidity rule can still be very usefulto identify structural faults such asexcessive bearing clearance andother forms of structural looseness.
To apply the validity rule with alaser system, it is necessary torecord all four cardinal readings(top, bottom, left, right) and plugthem into the formula. If, however,the alignment solution is based ononly three of the four cardinalreadings, the user will not have theability to check the validity of thesolution.
In one such example involving afeed water pump in a power plant,an alignment was attempted usingonly three of the four cardinalmeasurement (top, left and right -the bottom reading was omitted).The machine was moved as indi-cated by the laser system but noimprovement in the alignment con-dition was achieved. Numerousreadings and machine moves wereimplemented but failed to result inany improvement in the alignmentcondition.
When the reading for the fourthposition (on the bottom) was man-ually collected and the values wereplugged into the equation, it wasclear that the validity rule wasbeing violated. Visual inspection of
the machine train indicated thatone of the feet on the gearbox hadbeen bolted down with the wrongsize bolt head - thereby substan-tially reducing the hold-down forceat this foot. This allowed the foot tolift slightly during shaft rotationcreating substantial error in thereadings. After replacing it with theproper size bolt, the operator wasable to align the machine in just afew moves. (Note: more advancedsystems are currently availablethat will automatically apply thevalidity rule to the obtained read-ings and indicate whether accept-able levels for deviation have beenexceeded.)
It is important to realize that oth-erwise straight-forward alignmentjobs can become highly complexand yield unacceptable results ifthe technician does not addressthe quality of the alignment meas-urement and potential frame stressconditions (frame distortion, softfoot, and piping strain) during thepre-alignment check. These stepsshould all be conducted before thetechnician ever begins to move themachine.
Asset Optimization is possiblewith a balance of Technology,Expertise, and Work Processeswhich only Emerson Process Man-agement offers for each categoryof plant assets: mechanical equip-ment, electrical systems, processequipment, and instruments andvalves. Emerson’s MachineryHealth Management disciplineplays an important role in AssetOptimization by increasing the per-formance and availability ofmechanical equipment.
For more information on Emer-son’s Machinery Health Manage-ment, visit compsys.com.
Figure 1: Examples of good and badalignment data
