Filling The Gap

Filling in the Gap
An easy-to-use vibration screening tool for identifying basic machine faults fills the gap between inexpensive technologies with limited capabilities and expensive ones with advanced capabilities. By Alan Friedman, Predict/DLI Many plants and facilities around the world realize the benefits of predictive maintenance and use some sort of vibration analysis equipment to monitor the condition of their machines. In a typical plant, the vibration analysis program focuses on machines that are either critical to operations or costly to repair machines for which the costs of maintaining a database of trend information or using the valuable time of a vibration analyst can be justified. Even in the most aggressive vibration programs, a large percentage of low-cost machines, such as vent fans and pumps, go unmonitored. The device shown and described here provides a cost-effective method for monitoring these low-cost machines by using lower-cost diagnostic technology and personnel with fewer skills, allowing the benefits of periodic machinery monitoring to be extended to cover all of the assets in a plant. Monitoring all equipment reduces down-time and maintenance costs and aids in asset management and repair planning. This technology is also useful to those who do not have the resources to purchase and use an advanced vibration monitoring system, which typically costs $20K or more to purchase with an additional $20K going for training and labor. If a facility is outsourcing monitoring functions or periodically relying on an outside consultant to troubleshoot machinery, this less expensive technology can aid in determining which machines are in most need of attention, thereby lowering the number of machines a consultant must monitor.
Figure 1a
Limitations of RMS broad-band vibration meters

Currently, many facilities use broad-band RMS meters to fill the gaps described above. An RMS meter returns a single number that represents the total amount of vibration being emitted from a machine in a certain frequency range. The idea is that the more a machine vibrates, the worse its condition is. There are a number of problems with this concept. First, a single number gives no indication of what the problem may be and, therefore, offers no clue about what action to take. Imbalance, misalignment, and bearing wear account for 80% of machine faults, and each requires a different repair action. One does not have to change the bearings in a motor or replace the entire unit to deal with balance or alignment problems. The inability to differentiate between these three common faults (or any other faults for that matter) using an RMS meter is a significant limitation of this type of device. Secondly, there are a number of cases in which increasing vibration levels do not indicate mechanical faults. Also, there are a number of faults that can be present in a machine that do not add significantly to the overall RMS vibration level.
Limitations of shock-pulse technology

Another device that often complements a vibration analysis program uses shock-pulse technology. This technology is fairly adequate in accomplishing the specific task for which it was designed identifying bearing wear. However, its ability to identify only bearing wear is a serious limitation. Remember, imbalance and misalignment are also common machine faults, and they accelerate bearing wear. Therefore, if one can identify these additional faults and correct them, the bearings will last much longer. Thus, it is more cost effective to know that a fan is out of balance and clean the blades than to wait until the bearings begin to fail and then replace them or the entire unit. Even if one uses a shock-pulse device to avoid a catastrophic failure, the cause of the failure has not been addressed or avoided. A proactive approach to maintaining ones machines can be accomplished just as easily and inexpensively using the new diagnostic screening device.
Using the diagnostic screening device

The device (Figure 1a) is easy to use. One simply selects the appropriate machine running speed from the keypad, places the magnetically mounted accelerometer on the bearing housing of the machine, and compares the levels of the three bars on the display. If all of the bars are green, the machine is in good health. If one or more of the bars is red, one consults a cheat sheet to determine what the problem is. The relative heights of the three bars give an indication of what frequencies of vibration are present or exceed alarm levels. The cheat sheet relates these patterns to specific machine faults depending upon the type of machine and the axis of the measurement. The sheet includes pictures (e.g., Figures 2b, 3b, 4, and 5 in the box on the next page) along with fault descriptions by machine type. Each reading usually takes less than a minute to complete. The numerical display can be configured in Quick Trend mode to show the level of each of the three bars in terms of how much they exceed the alarm level. Thus a reading of 770 means that the first two alarm bars exceed the threshold by 7 and the third bar is exactly at the alarm level. This three-digit number may be trended over time to monitor changes in the machines status. Each LED indicator on the bars denotes a change in vibration level of 3 dB (Note: an increase of 6 dB is a doubling of amplitude). A demonstration program of the screening device is available on the companys Web site, www.PredictDLI.com. Click on Products then Vibration Data Collection & Analysis Products then DLI Watchman ST-101 - Vibrations Screening Tool.
How the device works

While a user needs no background in vibration analysis to successfully use the diagnostic screening device, knowledge of vibration analysis is necessary to understand how it works. When the user enters the machine speed, three separate alarm bars are configured (Figure 1b see box on next page) such that the crossover point of the Low and Mid bars corresponds to the running speed of the machine. The crossover of the Mid and High bars corresponds to 10x the running speed of the machine, and the crossover of the Low and High bars corresponds to approximately 3x the running speed. These alarm bars directly relate to the three bars on the device display (Figure 1a - previous page). If one overlays a spectrum on this alarm plot, it becomes obvious how the relative levels of the alarm bars correspond to high vibration levels at discrete frequencies. As an example, imbalance produces a high 1x peak in the vertical or horizontal direction. If a spectrum with a high 1x peak is overlaid on our alarm graph, the peak will exceed the thresholds of the Low and Mid bars by exactly the same amount. This is because the crossover point of these two alarms was set to the 1x speed when the machine speed was entered on the device. Therefore if the Low and Mid bars on the device are the same level, it can be said that the machine is unbalanced. If the same measurement were taken in the axial direction and these bars were the same height, the result would correspond to angular misalignment, which shows up at that same frequency in the axial direction.
Figure 1b
Figure 2a Imbalance Spectrum
Figure 2b Device Display Corresponding to 2a
Figure 3a Parallel Misalignment Spectrum
Figure 3b Device Display Corresponding to 3a
Figure 4 Rolling Element Bearing Wear
How The Device Works (cont.)

Parallel misalignment typically corresponds to a high 2x peak in the vertical or horizontal direction. If a high 2x peak is drawn on the alarm graph (Figure 3a), the peak may still cross over the Low and Mid thresholds. However, this time, the peak will exceed the Mid bar alarm by more than it exceeds the Low bar alarm. Thus, on the device, the Low and Mid may be in the red zone, but the Mid bar will be higher than the Low bar if we have this kind of fault in the machine. The High bar will be low. Roller bearing wear typically produces high frequency vibration.
Figure 5 Sleeve Bearing Wear or Belt Wear
The presence of peaks above 10x will exceed the High bar alarm more than the Low and Mid alarms. Therefore, a large High and small Mid and Low bars on the device would correspond to bearing wear (Figure 4 - see box above). Problems with sleeve bearings and belts or chains appear in a vibration spectrum at frequencies below the run speed of the machine. These faults produce a pattern on the device with a high Low bar and low Mid and High bars (Figure 5 - see box above). Finally, in rotating machinery, faults on components such as fan blades, pump impellers, and gears typically appear at a frequency equal to the number of elements multiplied by the shaft rate. For example, if one has 6 blades on a fan, fan blade problems will appear at 6x the shaft rate. One can add peaks to the alarm graph that correspond to these components to see how they would affect the relative heights of the alarm bars. About the author: In eight years at Predict/DLI, Alan Friedman has worked in software development, expert system development, data analysis, training, and installation of predictive maintenance programs. He is a graduate of Tufts University with a B.S. in mechanical engineering.

Filling The Gap

Uploaded by

Copyright:

Available Formats

Filling The Gap

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Filling The Gap

Uploaded by

Copyright:

Available Formats

Filling in the Gap

Limitations of RMS broad-band vibration meters

Limitations of shock-pulse technology

Using the diagnostic screening device

How the device works

Figure 2a Imbalance Spectrum

Figure 2b Device Display Corresponding to 2a

Figure 3a Parallel Misalignment Spectrum

Figure 3b Device Display Corresponding to 3a

Figure 4 Rolling Element Bearing Wear

How The Device Works (cont.)

Figure 5 Sleeve Bearing Wear or Belt Wear

You might also like