Pumps & Systems, July 2007
When measuring vibration, one of the most important factors in obtaining accurate information involves selecting the proper vibration transducer. Selecting the wrong vibration transducer can yield results that are misleading or, worse, mask a problem.
First, we will look at some of the basic considerations involved in selecting the proper vibration transducer:
- Type of machinery to measure
- Frequency range to be measured
- Environmental considerations
- Permanent or portable data collection
If the machine of interest is a typical pump driven by an 1800-rpm or 3600-rpm motor and the bearings are rolling element type, one would normally select an accelerometer for vibration measurement (see Figure 1).
Figure 1
Today's accelerometers are vastly improved over the accelerometers of 25 years ago. The accelerometers of today are mass produced, low cost, have excellent frequency ranges, and are very durable. In almost all applications they are superior to the old school velocity transducers. With modern day instrumentation one can easily view the acceleration data in velocity or even displacement units. By attaching the accelerometer to the bearing housing, the vibration from the machine's moving parts is transferred through the bearings to the accelerometer.
The next thing one must consider when selecting an accelerometer is its frequency range. One must anticipate the frequencies that would be emitted by machinery problems for the specific application. Then one must match the accelerometer so that it will have the frequency range to measure the anticipated frequencies. Figure 2 can be used as a general guide for the frequency ranges to expect and the best units of measure.
Figure 2
General machinery fault frequencies such as out-of-balance, misalignment, rolling element bearing, and blade or vane pass will typically be between 120 and 60,000 cycles per minute. Gear Mesh and three times Gear Mesh can easily exceed 1 million cycles per minute. Bearing Lubrication condition is often measured in Shock Pulse, a unit which requires the accelerometer to resonate at 2.16 million cycles per minute. Figure 3 displays two accelerometer response curves.
Figures 3 and 4
The first thing one may notice is how similar they look. What we are interested in is the flat or linear portion of the curve. Most of the time we want our fault frequencies to lie within this frequency range. Upon closer inspection, we can see that the top accelerometer has a higher linear frequency range (shifted to the right) than the bottom accelerometer. If, for example, we have a frequency of interest at 900,000-cpm or 15-kHz, the VIB 6.146 would not have a linear response in this range. Thus it would be a better fit to select the VIB 6.140 accelerometer. This is easier to see in Figure 4 which is from the specification sheets of the accelerometers.
So why would anyone choose the VIB 6.146? Notice from the specification sheet that the output is approximately 5 times that of the VIB 5.140. In areas with high amounts of electrical noise the VIB 6.146 is better, provided we do not need high frequency information above 10 kHz. The outputs are not voltage, but current. Again, this makes for a more noise resistant vibration signal. In addition, the accelerometer resonant frequency of the VIB 6.140 is ideal for measuring Shock Pulse for bearing lubrication.
Another important consideration is the environment that the accelerometer will be in. If the application is portable, meaning the accelerometer is placed on the machine with a magnet or VIBCODE cam-lock mounting pad, the environment is less critical. If the accelerometer is to be permanently mounted due to limited access or connection to an automated data acquisition system, then the environment is more critical. Some of the key factors are:
- Heat tolerance
- Moisture
- Chemical exposure
- Electrical interference
- Intrinsically safe requirements
- Shock limit
Again, a quick look at the accelerometer specification sheet in Figure 5 can reveal many of the environmental capabilities of the accelerometer.
Figure 5
In this case, we can see that the standard accelerometer is capable of sustained temperatures of 212-deg F while the EX version is limited to 176-deg F. The case material is stainless steel and the accelerometer is rated as IP65. We can even see that the connector is a TNC type which is quite durable in high moisture environments.
As can be seen, there are a variety of factors that influence the selection of the proper vibration transducer. The first critical consideration is to make sure the accelerometer you select has a frequency range that includes the potential machine fault frequencies. The second key factor is to make sure the accelerometer will perform in the environment you intend to use it in.
We have seen accelerometers last only three weeks in a semi-moist environment because the connectors were not moisture-proof. We have also seen accelerometers last over 10 years (and still going) in a 100 percent humidity environment on a paper machine because care was taken to ensure the accelerometer and connector were rated for the environment.