Three- And Four-Channel Vibration Data Collection
EP Editorial Staff | July 1, 2009
With the right condition-monitoring program and the right tools, you can reduce the costs of your equipment maintenance and fully utilize your personnel.
A number of factors come into play any time a company selects diagnostic tools for a condition-monitoring program. These days, however, decisions that relate to capital investment require even more careful consideration—especially when those decisions impact the maintenance and operations of your plant.
When it comes to making the right choice, technical design, validity of data and compliance with international standards, as well as the related cost savings (or time savings) associated with a tool should be of the utmost importance. This article takes a practical look at each of these aspects. In particular, it compares the data-collection process using a three- or four-channel device with that using a traditional single- or dual-channel data collector.
Time is money
As far as predictive maintenance practices are concerned, “time is money” is not just a phrase—it’s a hard fact of life. Increasing the speed and efficiency of data collection and measurement can increase the productivity of valuable personnel. Plant engineers and maintenance personnel need a diagnostic tool that quickly and efficiently measures and stores readings to detect changes in a machine’s health and performance.
Working with a three/four channel data collector with a tri-axial accelerometer can greatly speed up the data-collection process, provide enhanced analytical data, and display more data to assist in making effective machine diagnostic decisions. The resulting benefits can include more efficient use of plant personnel, reduced maintenance costs, and increased plant uptime and overall profitability.
The data-collection process
Figures 1 and 2 illustrate best practice for data collection, in this case on a motor pump, when taking both single-channel and multi-channel measurements, and illustrate the differences and potential benefits you can expect to find.
Test A: Motor pump set: single-channel measurement practice…
To begin our comparison, we place the single vibration transducer—or accelerometer—on the non-drive end of the motor pump set in the horizontal plane. When the data collector provides its results, we compare them with a pre-defined alarm limit and store them in the data collector. In this particular case, a simple value of 1.4 mm/s rms is recorded.
After we yield results for the horizontal plane, we remove the accelerometer, reposition it to the vertical plane and repeat the process of measuring, comparing and storing. To complete the survey, we need to move and place the accelerometer a total of 10 times: once for each measurement point required along the machine train.
Also, to save time, some engineers collect readings along the horizontal plane only, which highlights a benefit of tri-axial measurement: Three planes can be recorded simultaneously in the same amount of time that it takes to record one reading with a single-channel data collector.
Test B: Motor pump set: tri-axial channel measurement practice…
In Test A, we completed the initial simple survey of taking single-channel measurements on just one machine. However, if we had a large plant with several machines to survey, the single-channel measurement practice would take exponentially more time to complete. Using a multi-channel data collector/analyzer with a tri-axial accelerometer can speed up the process significantly.
A tri-axial accelerometer consists of three independent accelerometers placed at 90 degrees to each other and wired out on a multi-pin connector. Using a tri-axial sensor eliminates the need to relocate the sensor in between the horizontal, vertical and axial locations, ultimately consolidating them into one location.
With the tri-axial accelerator placed on the non-drive end in the vertical position, we are able to measure both the values for the horizontal and the axial motion at this location, which saves a significant amount of time. The result recorded is 1.6 mm/s rms for the horizontal, 0.8 mm/s rms for the vertical, and 0.5 mm/s rms for the axial.
In comparison to the single-channel data collector, with which we had to move and place the accelerometer 10 times, the multi-channel data collector with the tri-axial accelerometer required us to move and place the sensor only four times, which is a 60 percent savings in personnel time and, ultimately, cost.
An important note to consider with the tri-axial measurement practice is that it is much easier to take incorrect readings if attention to detail drifts away from the data-collection process. To prevent this from happening, concentrate on the indicated orientation of the tri-axial sensor and keep it the same throughout the measurement sequence in order to achieve successful readings.
Operating deflection shape using four-channel instruments
Another benefit of using a multi-channel data collector is that it can help analysts view an operating deflection shape of a machine. An operating deflection shape is an animated picture that illustrates how a machine actually is moving and running.
Operating deflection shape is a very powerful analytical tool, and the traditional low-cost approach is to use a two-channel analyzer and measure the frequency response function (FRF) between two accelerometers as one of them is moved sequentially between multiple measurement positions around the machine. The FRF is measured and stored first with the sensor in the vertical plane, then in the horizontal plane and, lastly, in the axial plane.
If the FRF can be measured between the reference accelerometer and all three planes simultaneously, the time/costs for a vibration consultant to acquire the data can be reduced by approximately two-thirds.
Validity of the data and compliance with international standards
Extra channels of data do not necessarily equate to more information, especially if the recorded values are incorrect. Some engineers have justifiable concerns that tri-axial readings are not representative of the true vibration values.
When collecting machine data, one must clearly define the objectives of the data collection. Is it to confirm that a machine’s vibration levels are in compliance with an international standard and that the data is collected in compliance with the standard’s guidelines? Or is it to monitor a parameter that is representative of the health of a machine, so that one does not incur the unnecessary expense of unscheduled downtime?
Considering the first point, anybody who writes a standard, whether it be international or internal, needs to convey clearly and easily to the user of that standard where the vibration measurements are to be taken. The most logical place is on the geometric centerlines of the machine being tested.
The reason one is compelled to measure vibration on the centerline when comparing the values with a standard is that measuring the amplitude of the horizontal vibration at a position on top of the machine could yield different values. If we assume that most-but not all-machines are reasonably rigid during healthy operation, then we can illustrate their typical vibratory motion by comparing it with the motion of a metronome.
At position A, for example, we can see that the vibration amplitude at the foundation is close to zero, and that the measured value of the motion most likely is going to increase as we move higher from the foundation level. Then, when we measure at point B, there could be a difference in level when compared with the measurement at point C. All three of the positions, however, represent the health of the machine.
Remember, the primary objective of predictive maintenance is to detect changes in measured values that indicate the onset of deterioration in machinery health and performance. The consistent measurement of valid and representative data is important no matter where it comes from.
Another justifiable concern among engineers is whether the trend in data will be affected by changing to tri-axial readings. The answer is probably not if the tri-axial readings are commenced at the same time as a new trend-such as after an installation or after an overhaul.
If, however, conversion to tri-axial readings is done part way through an existing trend, expect to see a change. You would need to review and compensate by adjusting the alert and danger limits.
While the validity of data and compliance with international standards should be a justifiable consideration when switching to a multi-channel data collector, there are others to take into account.
With the measurement of more than one or two channels comes the inherent need to display more information. Three- and four-channel data collectors have larger display screens. Depending on the particular plant environment, one might prefer to use a smaller, lighter-weight device-which typically has a small display screen. A large screen, though, offers clearer data and simultaneous display measurement results for a more effective reading and machine analysis.
Changing to a three- or four-channel device requires very practical engineering considerations. As more data is collected and analyzed, more power is needed-something that might result in increased heat produced by the device. Since increased heat has the potential to cause a fire (or even an explosion), it is important to consider your operating environment.
In other words
There certainly are substantial benefits associated with using three- and four-channel data collectors. The ability to analyze and collect data from three points of interest most notably offers increased savings-including those associated with personnel time and maintenance costs.
What’s most important, though, is the health of your machinery. When your equipment systems are down, so is your productivity and overall profitability. Three- and four-channel data collectors increase the speed and efficiency of data collection and measurement to detect health and performance in a way that is easy to use, read and justify. MT
Robert Collyer is a principal engineer for SKF, working out of the SKF Condition Monitoring Centre in Livingston, Scotland.