FIGURE 22 Basic relationship of measured parameters with a simple sinusoidal vibration the sensor will most likely have problems in one to two years when mounted in field applications where vibration is high, especially vane passing frequencies.

The piezoelectric accelerometer is a very light and compact sensor that measures vibration using a mass mounted on a piezoelectric crystal. Its output is low and requires a charge amplifier in the lead even with very short leads. The accelerometer is small and can be mounted virtually anywhere; it has a 1 to 3% influence factor from transverse side forces. A good rule of thumb on the usable frequency range is one-fifth to one-third of the resonant frequency. The disadvantages are that the sensor is sensitive to mounting torque, although stud mount is the best method to mount accelerometers. A lot of data are produced, of which some may be the data from an excited accelerometer resonance or cable noise. An impedance matching device can be built into the accelerometer for use at temperatures below 250°F (120°C), and cable noise can be greatly reduced with the voltage and charge sensitivity greatly improved; for example, 100 mV/g and 50 to 100 pC/g (where g = number of accelerations of gravity). For higher temperatures, the accelerometer will need a separate charge amplifier and may need heat insulation, such as MICA wafers (refer to API 678, dated 1981).

Most so-called ultrasonic analyzers use the accelerometer as a structural microphone. Many have chosen a carrier frequency in the megahertz range to improve the signal-to-noise ratio and make characteristic high-frequency patterns. Most of these systems are still in the development stage.

Techniques for Taking Data The second most important part of a vibration analysis program is the type of data taken and the techniques used to take the data. The purpose of taking vibration data on a pump is to either perform an analysis because someone noticed a noise or increased vibration level, or as a part of a periodic preventative maintenance program. It has been proven from experience that the velocity measurement is the best method for determining acceptable levels of centrifugal pump vibration. This is not to say that displacement and acceleration are not measures of vibration severity; they are, but it is necessary to know the frequency of the vibration. Displacement is preferred by a few for frequencies less than 6000 cpm. Accelerometers matched with analyzers can be purchased with signal integration that will give reliable readings in velocity in the 3000- to 60,000-cpm range. For readings above 60,000 cpm, the sensor would generally be an accelerometer reading in gX peak or integrated to read velocity zero to peak or 0-P.

Vibration amplitude is an important parameter because it indicates the approximate severity of the dynamic stress levels in the pump. Experience has shown that the shaft bearings and seal will probably fail in a pump with a velocity reading of 0.5 in/s (13 mm/s) 0-P. Also, catastrophic failures will probably occur when a pump is at 1 in/s (2.5 mm/s) 0-P. Pumps with velocity readings of 0.05 to 0.15 in/s (1.3 to 3.8 mm/s) 0-P, will perform well mechanically. Vibration readings are taken in the horizontal and vertical planes on the bearing housing of horizontal-shaft pumps.

To have a worthwhile maintenance reliability program with pumps, vibration readings must be recorded regularly (that is, monthly). This can range from a trend plotting of unfil-tered vibration to a full vibration analysis using a real-time analyzer to generate the frequency spectrum. A standard method used by many companies consists of taping pump vibrations with a battery-powered cassette recorder using a velocity sensor. Readings can then be processed through a real-time analyzer and recorded on an XYY plotter. The best application of this method is during startup and repair evaluation.

As an alternate method, a spectrum analyzer/plotter that produces a spectrum on a 4 in X 6 in (l0 mm X 15 mm) card with a frequency plot versus amplitude can be used. This procedure has in some installations detected and corrected 95% of the mechanical problems before failure. Experience has shown that had unfiltered displacement readings been taken, only 60 to 70% of the mechanical problems would have been observed. During these recordings, emphasis should be placed on the change in vibration levels, which is a better indication of a mechanical problem than absolute vibration.

One of the best pieces of data available for the pump's equipment file is a vibration record taken during the manufacturer's test or during water batching or commissioning. It is advisable to request a witness performance test on key or critical pumps. The purpose of this test is to assure mechanical reliability along with performance.

The manufacturer should be asked about the availability and type of vibration analysis equipment and sensors. Regardless of the instruments used, the vibration data sheet for the tested pump should have a sketch of where all vibration points were taken. The manufacturer should also supply a complete mechanical description of the number of impeller vanes, number of casing volute cutwaters or diffuser vanes, type of coupling, length of coupling spacer, and so on.

There are several different methods for taking periodic vibration data on pumps:

1. Using a handheld battery-powered velocity probe/readout, a machinist or operator logs unfiltered readings taken at one or two points on the bearing housing. When the reading reaches 0.3 to 0.5 in/s (8 to 13 mm/s) 0-P, the pump is pulled for maintenance. Readings are usually taken every two weeks.

2. Vibration points in the vertical, horizontal, and axial directions are recorded on a tabulated chart in unfiltered and filtered velocity at the various peak amplitudes, using a battery-powered tunable analyzer with a velocity sensor.

3. Vibrations in the vertical, horizontal, and axial direction are taken at each bearing, using a velocity sensor. The signal is recorded on a tape recorder, preferably a battery-powered FM/AM cassette. These data are then processed through a real-time analyzer. A spectrum hard copy is made on an XYY plotter of velocity versus frequency.

4. Key vibration points are fed directly from a velocity sensor or an accelerometer/charge amplifier through a long extension cable to a safe area, where a real-time analyzer processes the signal into a velocity versus frequency spectrum or a g's (acceleration) versus frequency spectrum. Hard copies for records are made on a XYY' plotter. This method requires two technicians with radios.

The most accurate are methods 3 and 4. The most costly to run in workerhours per point is method 2. The least accurate is, of course, method 1, but it is a popular screening technique.

Use of Vibration Sensors The use of a handheld velocity sensor with an aluminum extension rod or a light-duty vise grip with the probe mounted on the top of the grip has produced some high and misleading vibration readings because of extension resonances. For instance, the vise grip should not be used because of a 5000-cpm resonance. A 9-in (23 cm) long by §-in (0.95-cm) diameter extension to the velocity pickup should not be used above 16,000 cpm. The approximate axial natural frequency in cycles per minute for a rod extension from the probe, in tension and compression, can be expressed as in SI units in USCS units fn = 946

where W = pickup weight (force), lb (N) L = length of rod, in (m) A = cross-sectional area of rod, in2 (m2) E = modulus of elasticity of rod,° lb/in2 (kPa)

Use of attachments above these listed frequencies will produce a higher amplitude.

The best and simplest method of holding a velocity probe to a pump is a two-bar magnetic holder on the end of a velocity probe. Proper cleaning and some paint removal are generally necessary for good attachment. Periodic wiping of the magnetic bar to remove iron filings is also necessary. After mounting the probe, give it a light twist and a rocking motion; if it twists easily or rocks, change locations or reclean the surface. This location should be marked and future readings taken on the same spot; otherwise, the trend plots will vary.

If there is a concern about an extension or magnetic holder resonance, test this by holding the sensor, without the extension or magnetic holder, directly to a reasonably flat spot and noting any differences. Holding the sensor on a flat spot is generally safe up to 60,000 cpm.

When measuring vibration on an electric motor, there is always the possibility of false readings at 60 and/or 120 Hz due to electrical induction by the motor. This can be checked by two methods:

1. Hold the sensor by its cord and move it toward the motor, noting any increase in amplitude.

2. Using a two-channel oscilloscope, trigger the filtered signal against line voltage. Inphase signals mean the vibration is electrically induced.

For field use, one usually does not have to contend with temperatures above 250°F (120°C) direct to the sensor; thus, accelerometers with built-in impedance matching devices can be used and 100 mV/g voltage sensitivities can be obtained and transmitted 300 ft (90 m) if necessary. If the frequency range is low, and it is for pumps, a charge sensitivity in the order of 100 pC/g can also be obtained.

Techniques for Taking Preliminary Vibration Readings Some key points to remember before you start your analysis of the vibration problem:

1. Do not reach a decision on what the problem is before you record and analyze the data. By deciding too quickly what the problem is, you will most likely neglect other important factors.

2. Before you take the data, take time to review maintenance logs, talk with the area mechanic and operator, and make notes on the following:

a. Are there any unusual sounds (cavitation, bearings, and so on)?

b. Is there any movement in the discharge pressure gage?

c. What is the direction of rotation?

d. Are the flush and cooling lines lined up properly?

e. Is there any movement in the coupling shim pack?

f. Are there any foundation cracks?

g. Are pipe supports functioning properly?

h. Has a suction screen been installed?

°For aluminumm, E = 10.3 X 106 lb/in2 (71 X 106 kPa). For steel, E = 30 X 106 lb/in2 (2 07 X 106 kPa)

i. What is the magnitude of the liquid velocity in the suction line? j. Is the automatic oiler level adjusted correctly? k. Where is the pump being operated? l. What are the flow, suction, and discharge pressure? m. Is the pump's minimum flow bypass system in service? n. Has the process changed? o. What is the suction valve stem orientation? p. Have there been any color changes in the paint? q. Are there any loose parts, including the coupling guard? r. Determine the color and feel of the oil (if possible). s. What is the bearing housing temperature?

t. How is the coupling guard attached (attachment to the bearing housing is poor practice)?

You will be surprised how much this information aids in an analysis. Example: you record a high 1X in the radial plane, low values of 2X and 5X, and several high-frequency components at about 0.15 in/s (4 mm/s) 0-P. The 1X could be a bent shaft, loose coupling, plugged impeller, bad coupling unbalance, upper and lower case halves misaligned, and a whole list of running frequencies symptoms. During your review of maintenance logs, you noted that the impeller had been replaced because there had been distillation column tray part damage. The next questions you should ask are, "Did maintenance reinstall the suction screen?" (if not, another tray part may have lodged in the impeller) and, "Was the impeller rebalanced after it was trimmed from maximum diameter?"

3. Do not try to interpret partial vibration readings for someone looking over your shoulder before you have even taken all the readings. Sit down in a quiet place with your notes on installation and maintenance, a symptoms list, and a severity chart and then make the analysis. Analysis is not a simple task, but with some experience you will build confidence and it will become second nature.

Renewable Energy 101

Renewable Energy 101

Renewable energy is energy that is generated from sunlight, rain, tides, geothermal heat and wind. These sources are naturally and constantly replenished, which is why they are deemed as renewable. The usage of renewable energy sources is very important when considering the sustainability of the existing energy usage of the world. While there is currently an abundance of non-renewable energy sources, such as nuclear fuels, these energy sources are depleting. In addition to being a non-renewable supply, the non-renewable energy sources release emissions into the air, which has an adverse effect on the environment.

Get My Free Ebook

Post a comment