(CM) Techniques

Vibration Monitoring

Vibration monitoring is probably the most widely used predictive maintenance

technique and, with few exceptions, can be applied to a wide variety of rotating

equipment.

Machine vibration comes from many sources e.g. bearings, gears, unbalance etc. and even

small amplitudes can have a severe effect on the overall machine vibration. Each source of

vibration will have its own characteristic frequencies and can manifest itself as a discrete

frequency or as a sum difference frequency.

There are two ISO recommendations concerning machine condition monitoring by this
type of measurement, the much used ISO 2372 and the more recent ISO 10816, which is a
replacement of the older standard.

Features of ISO 10816

Measurements are made in three direction (horizontal, vertical, axial).

Machine condition is generally diagnosed on the basis of broad band vibration

measurements returning an RMS value. ISO 10816 keeps the lower frequency range
flexible between 2 and 10 Hz, depending on the machine type. The upper frequency is
1000 Hz.

ISO 10816 operates with the term vibration magnitude, which, depending on the machine
type, can be an RMS value of vibration velocity, acceleration or displacement. If two or
more of these parameters are measured, vibration severity is the one returning the relative
highest RMS value.

The standard consists of several parts, each treating a certain type of machines, with tables
of limit values differentiating between acceptable vibration (green range), unsatisfactory
vibration (yellow range), and vibration that will cause damage unless reduced (red range).
At low speeds, it is still possible to use

vibration but a greater degree of care

and experience is required and other

techniques such as measuring shaft

displacement or Acoustic Emission (AE)

may yield more meaningful results

although the former is not always easy

to apply. Furthermore, AE may detect a

change in condition but has limited

diagnostic capability.
Acoustic Emission (AE)

History, experience and familiarity

count for a lot where conditioning
monitoring is concerned, but that
doesn’t negate the need for change,
innovation and the advancement of
tried, tested and trusted techniques.
The late Steve Jobs commented:
“Innovation is the ability to see change
as an opportunity, not a threat.”

Steadily disrupting traditional vibration

techniques is acoustic emission (AE).
Acoustic Emission (AE)

AE technology spawned from the aviation industry where vibration analysis simply couldn’t
be easily applied. AE technique is based on frequencies much higher than are monitored in
the movement of vibration. These frequencies are the result of shock, friction and cracking,
for example. By this means, it is possible to detect impending failure before damage occurs,
as well as monitoring its progress thereafter.

For vibration analysis, the defect repetition frequencies are critically dependent upon the
machine component design and geometry, as well as the precise running speed. Vibration
can occur independently in the X, Y or Z axis, and so orientation of the sensor is as important
as location. For a detailed interpretation and to analyze the data before making a diagnosis,
it is also necessary to know internal machine geometries, shaft speeds, meshing frequencies,
etc. So in summary, vibration analysis is valuable but too often overly complicated.
Acoustic Emission (AE)

AE provides an earlier warning, detecting wear and small defects, whereas with vibration,
damage must have occurred to detect a signal. AE will pick up a lack of lubrication, friction
and cracking, which vibration will not, although it must be acknowledged that the totality of
information obtained from AE will be more limited than that derived from vibration.
AE Inspection Usage

Welding Cracks Bridges Gas tubes

Aerospace structures
Pipeline Monitoring
Vibration Monitoring for wind turbines

Vibration is used successfully on wind turbines where the main rotor speed is typically

between 5 and 30 rpm. In a wind turbine, there are two main groups of vibration

frequencies generated - gear and bearing defect frequencies. This can result in

complex vibration signals, which can make frequency analysis a formidable task.

Bearing defects can excite higher frequencies, which can be used as a basis for

detecting incipient damage.

Vibration measurement can generally be characterized as falling into three categories –

detection, diagnosis and prognosis.

Detection generally uses the most basic form of vibration measurement, for example,

of 10-1000Hz or10-10000Hz.

In machines, where there is little vibration other than from the bearings, the spikiness

of the vibration signal, may imply incipient defects, whereas the high energy level may

indicate severe defects. This type of measurement generally gives limited information

(other than to an experienced operator) but can be useful for trending, where an

increasing vibration level is an indicator of deteriorating machine condition.

Trend analysis involves plotting the vibration level as a function of time and using this to

predict when the machine must be taken out of service for repair or at least a more in

depth survey must be performed.

Another way of using the measurement is to compare the levels with published vibration

criteria for different types of equipment. Although, broadband vibration measurement

may provide a good starting point for fault detection, it has limited diagnostic capability

and, while a fault may be identified, it may not give a reliable indication of where the

fault lies, for example in bearing damage, unbalance, misalignment etc.

Generally, rolling bearings produce very little vibration when they are free of faults

and have distinctive characteristic frequencies when faults develop. Where machine

speeds are very low, the bearings generate low energy signals, which may also be

difficult to detect. Furthermore, bearings located within a gearbox can be difficult to

monitor because of the high energy at the gear meshing frequencies, which can mask

the bearing defect frequencies.

Where an improved diagnostic capability is required, frequency analysis is normally

employed which usually gives a much earlier indication of the development of a fault and

also its source. Having detected and diagnosed a fault, it is much more difficult to give a

prognosis on the remaining useful life and possible failure mode of the machine or

equipment. This often relies on continued monitoring of the fault, to determine a

suitable time when the equipment can be taken out of service, and/or on experience

with similar problems.

Overall Vibration Level

This is the simplest way of measuring vibration and usually involves measuring the RMS
(Root Mean Square) vibration of the bearing housing or some other point on the machine
with the transducer located as close to the bearing as possible. The vibration is measured
over a wide frequency range, such as 10-1000Hz or 10-10000Hz.

The measurements can be trended over time and compared with known levels of
vibration, or alarm levels can be set to indicate a change in the machine condition.
Alternatively, measurements can be compared with general standards or charts.

Although this method represents a quick and low cost method of vibration monitoring, it
is less sensitive to incipient defects i.e. it is only really suitable for detecting defects in the
advanced condition and has limited diagnostic capability.
In the early stages of a bearing defect, the vibration is normally low compared with

other sources of vibration present and is therefore easily influenced, so any changes in

bearing condition are difficult to detect.

Therefore, in some situations, the Crest Factor (Peak-to-RMS ratio) of the vibration is

capable of giving an earlier warning of bearing defects.

As a local fault develops, this produces short bursts of high energy, which increase the

peak level of the vibration signal but have little influence on the overall RMS level. As

the fault progresses, more peaks will be generated until finally the RMS vibration

increases but the Crest Factor decreases.

Vibration Measurement

The first task for the vibration analyst is

to obtain frequency selective

amplitude data to identify the source

of vibration. The frequency is the key

information that establishes the

possible causes and then the

amplitude is used to judge the severity.

This means having a vibration analyser

at one’s disposal. This takes us to the

first fork in the analysis flowchart.

The appropriate question to ask when armed with the frequency and amplitude data is

whether the vibration is benign or serious. If it is benign, then we can safely ignore it.

The reason for determining early whether the vibration is forced or natural is because
the fixes are completely different. Forced vibrations on machines are corrected by mass
balancing, aligning, or changing the bad parts. Natural vibrations are a structural effect,
where some structure behaves like a mechanical amplifier that is frequency sensitive.
The symptoms of natural vibrations, or resonance:

1. The vibration is very bad, in other words, abnormally high amplitude.

2. The vibration is strongly directional.
3. The amplitude is not steady, but varies up and down.
4. Rumbles during run-up or coast-down.
Once a resonance is suspected from the symptoms, then it needs to be verified with
some additional tests. These resonance tests are:
1. Impact testing of the major components to find their natural frequencies.
2. Run up or coast down.
3. Operating deflection shapes.
4. Variable speed shaker.
The four known fixes for resonance are:
1.Change speed.
2.Change the natural frequency of the responding part with added mass or stiffness.
3. Add damping or dynamic absorber.
4. Reduce the source vibration input.

These fixes are very different from the balancing, alignment, or changing parts for
forced vibrations, so it is important to identify this fork in the road before proceeding
down the wrong path.
Root Cause Analysis
Machine vibration has several categories of causes that are discovered sometimes

after repair. The major categories are,

• Design defects.

• Manufacturing defects.

• Operational stresses.

• Maintenance actions.

• Aging.
Design defects are mostly structural related with active resonances built-in because

of improper sizing and proportioning of the parts. Statically, the structure is O.K.,

but is dynamically weak. This is not discovered until the machine is energized and

brought up to speed. This is more common than it should be, but designers are not

well equipped to predict or test for natural frequencies. In addition, the owners’

foundation or base has a significant effect on natural frequencies, which the

designer has little control over. Hence, resonances are best detected during start up

testing and corrected on-site with strategic stiffeners added.

Manufacturing defects are built-in during the casting, machining, heat-treating, and

assembly processes. They are latent defects that may show up in the first 24-hours of

running, or they may not be obvious during the run-in period, rather appearing years

later. The machine does not survive to a normal life expectancy. Vibration may or may

not be present.

An example is residual stresses in a shaft that gradually distorts the shaft over a period

of years.

Manufacturing defects are difficult to control, impossible to predict, and elusive to fix.
Excessive operational stresses can develop due to material build up or

erosion, that changes the balance condition, or thermal expansion that

changes component alignment. Both of these cause high dynamic loads at

the bearings which lead to accelerated wear out. These defects are easily

detected with periodic vibration measurements and there are well

established methods to correct them on site.

Maintenance actions, or inactions, are the most common cause of machine failure. It is
well known in the repair business that a machine never goes back together the same way.
Some of this is due to rough handling, but some is simply the fact that field repair is less
controlled than the original factory build.

Other maintenance activities that affect vibration are –

• Excessive localized heating, like welding on a shaft.

• Too high belt tension.
• Shaft, or bearing, misalignment.
• Substandard replacement parts.
• Lack of lubrication.
• Loose hardware.
• Replacing hardware with different weights that affect balance.
• Hammering on a bearing.
• Unclean, or burred, precision machine surfaces.
Aging effects can only be detected with long term vibration monitoring. The two

dominant aging effects are residual stress relaxation and softening of structural

joints. The residual stresses left behind in machine components will always relieve

themselves over time. This process is accelerated at higher temperature. Shafts,

being long and slender components, are particularly vulnerable to bowing. The

symptoms are an increase in 1xRPM balance condition and beating up of the

bearings. Bearing replacements do not restore the original smooth running condition,

and mass balancing is unsuccessful, until the shaft is replaced.

All joints soften over time, and joints are the weak links in any structure. The subtle

symptom of this is lowering of the natural frequencies. This is usually first detected

with high vibration when the lowest natural frequency drops down into the operating

speed range of the machine.

Corrective Methods
The good news is that almost all machine vibration problems can be corrected in
place.

• Disassembly, visual inspection, cleaning, and re-assembly can fix some elusive
problems.
• Bearing replacement.
• Identifying other bad parts and replacing them.
• Mass balancing.
• Alignment.
• Lubrication. Just greasing noisy bearings can quiet them, but changing the
lubrication schedule can extend their lives.
• Structural stiffening to raise natural frequencies (Mass loading and stiffening can
reduce any measured vibration).