| |
DoctorKnow® Application Paper
| Title: | Field Balancing in the Real World, Part 1 |
| Source/Author: | Bob Bracher |
| Product: | Corrective |
| Technology: | Corrective |
| Classification: | |
Field Balancing in the Real World
Plant & Equipment Division, ORNL
Bob Bracher, Vibration and Training
Technical Training Department
Abstract: Field Balancing in the Real World
Field balancing can achieve significant results when other problems are present in the frequency spectrum and multiple vibrations are evident in the waveform.
Many references suggest eliminating other problems before attempting to balance. That's great—if you can do it. There are valid reasons for this approach, and it would be much easier to balance machinery when other problems have been corrected. It is the theoretical ideal in field balancing. However, in the real world of machinery maintained for years by reacting to immediate problems, the classic vibration signature for unbalance is rarely seen.
We make most of our decisions with limited information. The decision to balance or not to balance is usually made the same way. This paper will demonstrate significant results of field balancing in the presence of multiple problems. By examining the data available and analyzing the probabilities, a reasonable chance for success can be assured.
Bob Bracher, Vibration and Training Specialist
Oak Ridge National Laboratory
managed by
Lockheed Martin Energy Research Corporation
Post Office Box 2008
Oak Ridge, Tennessee 37831-6411
September 5, 1997
Notice: The submitted manuscript has been authored by a contractor of the U. S. Government under Contract Number DE-AC05-96OR22464. Accordingly, the U. S. Government retains a non-exclusive, royalty-free license to publish or reproduce the published form of this contribution, or allow others to do so, for U. S. Government purposes.
Oak Ridge National Laboratory is managed by Lockheed Martin Energy Research Corporation for the U. S. Department of Energy.
Acknowledgements
In a large, multifunctional company such as ORNL, it takes the efforts and contributions of several groups and individuals to bring about the change necessary to implement new techniques and ideas. Field balancing is not new. The idea of doing it on our own is new. Without the efforts of Eddie Baird, Wayne Castleberry, Doug Edwards, Bob Hileman, Blake Van Hoy, and J. J. Vivyan we would not have been able to perform the jobs mentioned in the case histories. Their efforts were made possible by the cooperation and support of managers, administrators, field maintenance and operations staff, and consultants of ORNL Plant & Equipment Division, ORNL Waste Management Remedial Action Division, and ORNL Environmental Restoration Division. Their continued efforts and support is appreciated.
CSI deserves some credit, too, for the many times they helped me resolve problems by telephone, in class, or in person. I particularly want to express my appreciation to the training, information desk, sales, software, hardware, and technical support groups, and specially to Darla Adcock.
Introduction
This paper will discuss case histories of field balancing jobs successfully accomplished at Oak Ridge National Laboratory (ORNL) in the presence of other vibration problems. A demonstration will explain by example the cases presented here. ORNL Journeyman Millwright Doug Edwards has agreed to assist me with this demonstration. I am grateful to him for his enthusiasm, skill, and dedication to predictive and proactive maintenance technologies. I enjoy working with him and I think you will too.
Overview
Ø List of CSI Equipment Used
Ø Field Balancing Basics: Decision to Field Balance
Ø Field Balancing Guidelines
Ø Case History One: The Underbalanced Overfire Blower
Ø Case History Two: The Confident Consultant
Ø Case History Three: The Doubtful Backup
Ø Demonstration of Multiple Problems
Ø Summary
Ø Recommendations
Ø References
Ø Appendices A & B
List of CSI Equipment Used
Ø CSI 2115 Machinery Analyzer (a single-channel analyzer)
Ø CSI 404 Infrared Phototach
Ø CSI 440 Strobe Package
Ø CSI FAST Bal II Downloadable Program
Ø CSI Cascade Downloadable Program
Field Balancing Basics: Decision to Field Balance
Appearance of Unbalance Signature
The 1x magnitude is always increased with an unbalanced condition in rotating machinery, and it is usually highest in the radial direction. A single peak of excessive vibration at 1x running speed (or 1 order) is the classic unbalance signature. An accompanying waveform with a nearly perfect sine wave all but confirms it, right? If all of these things are true, the decision to field balance ought to be fairly easy to make from one set of data.
Other Factors
The problem arises when the forcing functions causing excessive 1x peaks are examined. Some, such as (1) bent shaft, (2) eccentricity, and (3) unbalance (center of gravity and center of rotation do not coincide), will respond to field balancing. Others, such as
(4) severe looseness, (5) resonance near running speed, and (6) cracked shaft, will not respond successfully to field balancing. To complicate matters, misalignment can be mistaken for unbalance, and unbalance can be mistaken for misalignment. Misalignment can exhibit a 1x peak only. An overhung rotor can exhibit a high 1x in the axial direction or the radial direction, as well as a high 2x running speed peak.
Additional Tests to Confirm Unbalance
There are several ways to confirm that a field balance correction would decrease the vibration in rotating machinery: (1) On a single-plane rotor, phase data between the two bearings can help to confirm an unbalance condition. If the vibration of the two bearings is in phase, it is likely unbalance rather than misalignment. If the vibration is 180 degrees out of phase, the problem is likely misalignment. (2) Bump tests can rule out resonance within 10-20% of running speed (difficult or impossible to balance successfully).
(3) Waterfall or cascade plots of coastdown and runup data can also help. All of the tests mentioned can be performed with a single channel analyzer, such as CSI's 2115, and a tach trigger.
A Note on the CSI Cascade Downloadable Program
When properly set up and operated, this program and the cascade plots produced can provide a wealth of reliable information quickly. CSI's Cascade Downloadable Program can reveal a resonant frequency too close to running speed, some transient events, and magnitude in relation to varying running speeds.
Balancing Standards
The decision to balance must take into consideration some standard that can be met. Currently, the primary standard used on commercial fans and pumps at ORNL is customer satisfaction. Objective standards for assessing field balancing results can be stated in maximum ounce or pound inches of residual unbalance remaining, or they can be stated in terms of a vibration magnitude limit at a certain location at 1x turning speed. Balancing standards will vary greatly with application, consequences of a machine failure, and customer requirements. The balancing of rigid rotors is covered in ANSI Standard S2.19-1989.
Field Balancing Guidelines
I. Preliminary Checks:1. Complete Safety Work Permit according to YOUR COMPANY'S procedure.
2. Lockout/tagout equipment according to YOUR COMPANY'S procedure.
3. Inspect machine to be balanced even if it has already been inspected for any of the following problems that could preclude a satisfactory balance of the machine:
A. Cracks in foundation, grouting, welds, mounting hardware, rotor, etc.
B. Loose mounting bolts, debris, hardware, bearings, or couplings.
C. Missing parts: keys, set screws, balance weights, etc.
D. Damaged or broken parts: coupling, belts, bearings, etc.
E. Deformed parts: bent blade, misshapen belts, etc
F. Discoloration of moving parts
G. Excessive wear of visible moving parts.
H. Too much, too little, or burnt lubrication.
I. Dirt or debris on rotor
4. If any of the above items were checked and not corrected, report findings.
5. Clean the rotor of the machine to be balanced if it is dirty.
II. Typical Machinery Vibration Analyzer Equipment Setup:
1. Mount two vibration transducers radially at each bearing.
2. Install reflector tape for PhotoTach reference on shaft connected to rotor.
3. Mount PhotoTach to read reflector tape on center and perpendicular to rotor shaft axis.
4. Connect Phototach, multiplexer, and cabling to machinery analyzer and accelerometers.
5. Set up the job in the machinery analyzer.
6. Ensure that cords, accelerometers, and equipment will not interfere with shaft rotation.
CAUTION: From this point on, permits and lockout/tagouts may need to be suspended temporarily and then reinstated until the field balancing job is complete. Follow YOUR COMPANY'S procedures applicable to this job. Actual step-by-step of balance job will vary.
III. Typical Steps for Single-Plane Balancing
1. Acquire reference data.
2. Add one or more trial weights.
3. Acquire trial run data with trial weights in place.
4. Remove trial weights.
5. Determine whether to add or remove weight to balance.
A. Calculate correction weight and location required to balance machine (add or remove).
B. If adding weight, permanently affix correction weight; if removing, calculate area needed for removal and remove weight.
6. Check results and trim balance as necessary.
7. Repeat trim runs as necessary until within tolerance or acceptable to customer.
8. If trim balance calls for too much weight (>50% of correction), re-evaluate job.
9. Return machine to customer according to YOUR COMPANY'S procedure.
Case History One: The Underbalanced Overfire Blower
Nature of Involvement
No vibration readings had been taken on this fan before September 27, 1996. The resident millwright—Eddie Baird—had known of our program and some of our equipment, and he recommended our services to the relief supervisor.
Known Facts
New bearings had been installed in the motor, and the fan rotor was cleaned. The fan shook the entire second floor of the steam plant when the operators tried to run it. This fan had motor bearings only.
A set of vibration data was taken to verify that the fan was out of balance. The vibration was directional to some extent, with the horizontal reading >4x the vertical reading.
Relevant Spectrum
Spectrum After Balance
Fig. 1. Initial spectrum.
Fig. 2. Final spectrum.Decision to Balance
The decision to balance was made on the basis of work already performed, major 1x vibration, and the comparatively low magnitude of the 2x-4x harmonics. During the balance job, the software help message (CSI FAST Bal IIÔ) advised that the vibration was directional at one measurement location out of four. We continued to balance, and the spectrum above shows the pleasing result. Incidentally, the weight had to be placed at the same phase location as the original weight, hence the name for this case history.
Other Fan Problems
After field balancing, the 2x-4x fan harmonics are now ~20% of overall vibration. Additional data reveal that the vibration magnitudes are no longer directional. This suggests excitation vibration due to original high magnitude. Tracking and trending is the order of the day for this machine.
Case History Two: The Confident Consultant
Nature of Involvement
An outside consultant was hired by the ORNL Environmental Restoration Division to determine the "upgradeability" of two exhaust fans for one of our old reactor facilities, the Molten Salt Reactor Experiment (MSRE), nicknamed "ole salty". Our resident millwright—Bob Hileman—for that area advised his supervisor and the consultant of our services. We had already recommended a data collection route be started there.
Known Facts
Previous readings had been taken on this belt-driven, centrifugal fan, and the sheaves were realigned with a matched set of belts installed. New bearings were installed previously. The vibration readings were still too high. The data did not show a clearly unbalanced condition. However, in the face of increased speed, we recommended a coastdown test to determine the need to balance in order to reduce the vibration.
Relevant Spectrum
Fig. 3. Initial spectrum.
Fig. 4. Selected cascade spectra.Decision to Balance
The highest magnitude of our data set was the 1x in Fig. 3. at .318 ips peak velocity. This represented about 80% of the total, with some subsynchronous energy (<1x frequencies) and the remainder mostly synchronous energy (whole number multiples of 1x running speed frequencies). The cascade spectra from the coastdown (Fig. 4.) provided further assurance we could significantly decrease the vibration by balancing. The only possible resonance was at .543x running speed, or slightly greater than 675 rpm, so it would have no adverse effects on our balancing. Also, although difficult to see, the 1x peak continues to decrease in magnitude as the fan slows down. This is a sure sign that field balancing can reduce vibration magnitude.
Sounds good, huh? Well, the problem was that this fan did NOT have an access to the fan rotor (or fan wheel, if you prefer). The consultant was confident that our diagnosis was correct and that we could reduce the vibration. A cover was made, a hole was cut in the fan housing, and field balancing was the next step. We had help from our Central Engineer—Blake Van Hoy—by telephone, but I was sweating this one out!
Spectrum After Balance
Fig. 5. Second spectrum.
Fig. 6. Final spectrum.
The vibration for this fan was significantly reduced, but to my way of thinking, it was still too high (see Fig. 5.). It was running at only 1250 rpm, and it was headed for a speedup. We were fortunate to have two millwrights observing during this balance job. Good thing, too, because one of them—J. J. Vivyan—discovered the source of the remaining 1x vibration. During our trial weight run she noticed that a base bolt was loose.
After all of the base bolts were tightened, the vibration was reduced even further, and this fan continues to operate smoothly. The happy customer, ORNL Environmental Restoration Division, placed this fan and its mate on periodic vibration data collection.
Other Fan Problems
There is some nonsynchronous energy (mixed-number multiples of running speed) present in the last spectrum, Fig. 6. Tracking and trending with realistic baselines, fault frequencies of bearings, belts, blade pass, and parameter banding will help us correct the problem at the customer's convenience before a breakdown.
Fig. 7. Fan discussed in Case History Two.
Case History Three: The Doubtful Backup
Nature of Involvement
This particular fan was placed on routine periodic vibration monitoring along with several others, thanks in large part to our resident millwright in the area—Doug Edwards—who will assist me with the demonstration today. The first reading taken to establish baseline data showed serious problems.
Known Facts
The turbine-driven, direct-drive fan had recently been serviced. The turbine was rebuilt and it was aligned low to allow for thermal growth. No history was provided except the memory of the craft people working on this and similar jobs. This overhung fan served as a backup exhaust fan to an electric motor-driven primary, and it was operated once a week for about 30 minutes to verify that it was usable.
Relevant Spectrum
Fig. 8. Initial baselining the hard way.
Spectrum After Balance
Fig. 9. Final spectrum.Decision to Balance
The fan was exhibiting a 1x peak with a magnitude >85% of the overall vibration, and it was very high. Misalignment was not immediately suspected because the turbine was just rebuilt. Fan operation was improved but excessive vibration was noticeable.
Other Problems
This fan was difficult to balance because of the underlying misalignment, and we wound up with three weights in each plane. Using the "Add Weights" feature in the FAST Bal II software, we calculated a single correction weight for each of two weight planes, and the weights were welded in place. After balancing, the misalignment was evident (Fig. 9.). However, the customer was happy, the operators on the job said this fan had never run this smoothly, and the vibration was reduced by >5 times, a significant improvement. The machine awaits alignment when time permits. If pressed into service now, however, it would run much longer than before it was balanced.
Demonstration of Multiple Problems
The machine used for the demonstration is the Baseline Trainer available from CD International. For more information on this machine, contact Mr. Zane Dreja at CD International, Lewiston, New York (1-800-388-7972).
Summary
As demonstrated, field balancing can achieve significant results when other problems are present in the frequency spectrum and multiple vibrations are evident in the waveform (see Appendix A for waveform data.)
CAUTION: Always correct other known problems if able to correct them before field balancing. Never attempt to field balance a machine if you are not reasonably certain it will significantly reduce the vibration.
Recommendations
Ø Adopt a troubleshooting strategy such as the one mentioned earlier ("…Decision to Field Balance"), or develop your own.
Ø Follow a checklist or guideline plan to avoid the frustration that comes from futile field balancing effort.
Ø Maintain records of balance jobs. If you have no record of where you have been no one will really know how far you have come except you.
Ø Use the reference materials listed to improve knowledge and skills.
References
The following publications are highly recommended for increasing your knowledge and skills in field balancing. Most also cover other vibration topics.
The Simplified Handbook of Vibration Analysis, Vols. 1 & 2, Crawford, CSI , 1992
Machinery Vibration Balancing, Wowk, McGraw-Hill, 1995
Shock and Vibration Handbook, Fourth Edition, Harris, Industrial Press, Inc.,1996
Reliability Magazine, Industrial Communications
P/PM Magazine, Second Childhood, Inc.
Vibrations, VI Press, Inc. (A Vibration Institute publication)
Appendix A
Spectra and Waveforms for All Case Histories; Bump Test Example
Fig. 10. Case History One.
Fig. 11. Case History One final.
Fig. 12. Case History Two.
Fig. 13. Case History Two second.
Fig. 14. Case History Two final.
Fig. 15. Bump test resonance at 1.1x.
Fig. 16. Case History Three.
Fig. 17. Case History Three final.
Appendix B
On the Job Assistance
The remaining pages contain copies of job aid cards for performing a single plane field balance using CSI FAST Bal II. They were developed and field tested by the ORNL Plant & Equipment Division Technical Training Department. ORNL is managed by Lockheed Martin Energy Research Corporation for the United States Department of Energy. The CSI FAST Bal II User's Manual and our CSI 2115 Machinery Analyzer were used to develop these. Accuracy of information contained cannot be guaranteed. CSI does not necessarily endorse the information, and CSI is not responsible for content.
All contents copyright © 1998 - 2006, Computational Systems, Inc.
All Rights Reserved. |