Vibration Analysis Reporting
Bearing Failure Stages & Responses
This technical note has been written to act as a guide to vibration analysts and maintenance personnel. It outlines the expectations that maintenance personnel should have w.r.t. reporting of bearing defects identified in vibration spectrums.
The Stages of Bearing Failure and Levels of Response below are to be used in conjunction with the Criticality Definitions as a guide for vibration analysts.
The criticality of a machine will determine the level of maintenance response to each stage of a suspected bearing failure. The different levels of response are required to accommodate operational factors such as:
i. When can it be shut down and what else will need to be shut down with it?
ii. How long will it be down and at what cost?
iii. What are the lead times for parts and labour?
Example 1. A furnace combustion fan, which supplies billets to a rolling mill, is highly critical and a Stage 2 defect will require that planning begin to replace the bearing. The next opportunity for replacement could be in 6 to 12 months, by which time the fault may have progressed to a higher stage.
Example 2. A gearbox driving a conveyor is a critical machine and a Stage 2 defect could be monitored until it reached Stage 3. It would then depend on the location of the bearing, Input Shaft, Intermediate or Output Shaft, the rate of deterioration and the nature of the fault; Raceway, Rolling Element or Cage; as to how urgently the bearing needed to be replaced and how often the machine needed to be monitored.
Modern vibration analysis provides a means of identifying bearing faults, at various stages in the progression to failure, by breaking up the vibrations into their component frequencies. Generally, the appearance of fault signatures progresses from very high to lower frequencies as deterioration progresses and the patterns produced in the spectra are recognisable to properly trained vibration analysts and their development can be trended.
The stages defined below are based on accepted industry practice and it is by clearly defining which stage of failure a bearing is at, that an analyst can prompt maintenance staff to take appropriate and timely action.
Four Stages in Bearing Failure and Levels of Response.
- The first stage (normal operation) appears at ultrasonic frequencies from about 1,200K to 3,600K CPM (20,000 â€“ 60,000Hz). At this stage, Spike Energy and Shock pulse instruments can evaluate the energy.
Indicates metal-to-metal contact. May indicate a lack of lubrication.
- At the second stage of bearing failure, minor defects excite a mounted resonance response (of the magnet / accelerometer assembly), which is picked up with a spectrum analyzer in the middle of the spectrum, 12OK-480K CPM (2000 â€“ 8000 Hz).
At the end of stage 2, bearing defect frequencies appear, sideband frequencies may also be present above and below the defect frequencies.
Highly critical machines should be replaced/repaired at the next available opportunity, this may be months ahead. A reduced monitoring interval should be established with the CM provider.
Critical machines should have additional VA taken for assessment.
20x Magnification showing pitting and micro pitting.
- In the third stage of failure, bearing defect frequency levels increase and their harmonics appear on the spectrum. As wear progresses, sidebanding increases around the defect frequencies and can be seen more clearly as raised levels and harmonics in the mounted resonance area. At this stage, if you remove the bearing, you can clearly see the defects in the raceways and / or rolling elements.
Bearings in critical and non-critical applications should be replaced at this stage. Additional VA should to be taken by the CM provider to establish a deterioration rate if maintenance action is to be delayed.
Spalling in raceway
- Stage four appears toward the end of bearing life. It shows up as random high frequency vibration, which lifts the noise floor in the higher frequencies. Discrete bearing defect and mounted resonance peaks begin to disappear and are replaced by a random, broadband noise floor in the lower frequencies.
Bearings requiring routine vibration monitoring should not be allowed to reach this stage.
Machine Criticality Definitions
A. Highly Critical â€“ Machines whoâ€™s operation cannot be interrupted and upon whose operation production systems depends.
B. Critical - Machines which are key to meeting production quotas and whose operation cannot be interrupted during a production schedule. Machines may be classified as critical if:
i. Repair down time is considered unacceptable by production.
ii. Long lead times exist for spares.
C. Non-Critical â€“ Machines which may be shut down during production for repair. Machines may be classified as non critical if:
i. They have backup systems that allow change over during production.
ii. The repair or replacement time would not affect production and spare parts and labour would be available if a repair or replacement was required.
Where possible, indications of a bearing fault should be identified by:
i. Stage of the fault as defined above.
ii. Nature of the fault. Raceway, Rolling Element or Cage.
iii. Location of fault. E.g. Input shaft NDE.
E.g. Stage 3, outer race defect on output shaft, load-side bearing.
The use of the same bearing at two different locations on the same shaft may prevent the exact bearing being identified, particularly if vibration data is collected from only one side, but it is usually sufficient to know that at lease one bearing is failing rather than which one.
Recommendations for corrective measures should be appropriate to the level of fault, the mechanics and criticality of each machine. Where the analyst does not know the criticality of a machine, a guess based on reasonable experience is acceptable but it is the responsibility of both the analyst and plant engineer to ensure that as much detail as possible is collected prior to analysis.
Vibration analysis requires special skills and training as well as experience. Good analysts with machine and plant experience regularly move on to fill other technical and leadership rolls, often in other companies. An analystâ€™s recommendation to either act or not act on the vibration data can have significant impacts on a plants operations; it is therefore reasonable for a plant engineer to ask the service provider for a brief summary of the following:
Analystsâ€™ formal training and level of experience on similar plants and machines.
Supervisorsâ€™ background in vibration analysis. This is particularly important if the analyst is new to the field and may need support.
The documented procedures used by a service provider to establish a vibration program and any guidelines, standards or reference material available to the analysts. It should be noted that this information will almost certainly be commercial and in confidence but service providers who have established these procedures and reference libraries should be more than happy for you to sight them and ask questions. Be objective in your assessment of them, they are critical.
Effective analysis requires objective interpretation of information, the better the information available, the more likely the analyst is to correctly assess the situation and provide meaningful recommendations on follow-up action.
It is important that the following information be available:
A sectional assembly drawing showing bearing numbers and locations
The kW rating of the machine (to indicate size)
The input shaft speed and any subsequent reduction ratios.
Component details such as; gear teeth numbers, pump and an blades numbers, etc.