In vibration analysis (with accelerometers) of rotating machines with rolling element

bearing, the process industries are interested to know the failure of the machine well

in advance to plan the spare inventory and maintenance. But in real world

most of the machines fails before the prediction of vibration analyst or Expert

analysis software. Presently the prediction of failure is based on ISO 10816 vibration

limits only. But this is not enough to monitor the failure of machines well in advance.

Because more than 40% of the machines will fail even the vibration readings are

within acceptable zone as per ISO 10816. Also we need to forced to run the machine

in order to avoid costly unplanned maintenance. How do we ensure that whether we

can take risk in running the machine above trip limit?

Hence it requires further detail analysis and different techniques to predict the failure

well in advance and some times allow the machine to run in critical situations.

1

The above chart is used in ISO -10816 method to determine the alarm and trip limit

for rotating machines equipped with rolling element bearings.

Till today, the world is using the above chart as guidelines to provide alarm and trip.

Also OEM recommends above chart as their threshold in their design documents.

The same thing is used in DCS and PLC logics to trip the machine whenever the

machine exceeds the threshold of trip level (Red) or action taken for maintenance.

2

The answers fro above questions start from slide 12

3

4

5

Speaker Notes:

This topic explains the basics of vibration measurement and the basics of how

that measurement is displayed and interpreted.

There are 3 types of vibration indicators.

1. Displacement – Oscillation of the object from front and back. Example

simple pendulum in Grandfather Wall Clock. This is measured in peak to

peak microns or mils. Basically this parameters are stress indicator.

Usually used in low speed machine and relatively massive weights.

2. Velocity – Rate of change of displacement. How fast the vibration is

occurring in a cycle. In Clock pendulum how fast the it is moving. The

speed is zero at extreme end and highest when it is in middle. Velocity is

measured in mm/sec or inc/sec peak or RMS. Basically this parameters are

fatigue indicator.

3. Acceleration – Rate of change of velocity. In pendulum clock example the

source of force is caused for the movement of pendulum. Usually the

pivot point is source of acceleration force. Acceleration is measured in

m/sec^2 or g peak or peak to peak. Basically this is force indicator.

4. Also any vibration parameters can be measured in peak to peak or peak

and RMS etc.

5. Also the vibration parameters are changeable mathematically from one

form to another form (Displacement-Velocity-Acceleration).

Note: I hope further explanation on this will increase the time and it is difficult to

understand without proper training in details. Please google for more details for the

appropriate topic and search.

5

The above chart shows relationship between Displacement, velocity and acceleration

measurement.

The above chart gives good amplitude capturing and relationship for analysis

between one type measurement to other’s in given range of frequency

measurement.

Displacement – Good response In low frequency ( 0 to 6000 cpm)

Velocity - good indicator in medium range of frequency ( 600 to 60000 cpm)

Acceleration – good indicator at high range frequency (>6000 cpm)

Basically velocity is good parameter when monitoring the machine problems like

unbalance, misalignment and looseness in the machine speed from 600 to 6000 rpm.

6

The above three spectrum shows relationship between the amplitude versus

frequencies impact at various types of measurements (displacement, velocity and

acceleration spectrum)

7

8

Bearing geometry overview to understand the components inside of rolling element

bearings.

9

The above slides shows how the envelope signals are processed from normal

acceleration waveforms.

Typically, the periodic impact pulses are of relatively small amplitude, and are

“buried” in the complex vibration waveform. It would be very unusual for them to be

visible in the raw vibration signal as clearly as they are shown in this drawing.

10

Speaker Notes:

This topic explains the basics of vibration measurement and the basics of how

that measurement is displayed and interpreted.

Statement of Relevance: Understanding how vibration signals are created is

essential to correctly reading dynamic data plots, which will be discussed in

the next few topics. This is a quick course in how Bently Nevada vibration

probes work.

Be sure to discuss the objectives stated below:

Manual Text:

Upon completion of this topic, the student will be able to perform the tasks in

the following objectives:

Recognize a vibration signal, and explain how it is generated by rotating

machinery.

Explain the definition and use of absolute phase, and find the absolute phase

from a timebase plot.

Explain the definition and use of relative phase, and find the relative phase

from a pair of timebase plots.

11

12

13

In the above charts, vibration readings collected from class 3 machine as per ISO

10816, The velocity trend readings are below 4.5 mm/sec throughout period from 12

April 2004 to 22 May 2005. Whereas the envelope readings are in increasing trend

over the period of 4 month before the bearing replacement made in November 2004.

The envelope trend has increased from 4g to 25g. The envelope readings are below

4g after bearing replacement.

Conclusion: There is no indication of bearing failure from velocity reading. If some

one follow the velocity reading is as a guidelines to stop the machine for

maintenance. The machine will lead to catastrophic failure or costly maintenance as

the impeller or casing might damage during the breakdown of the machine. Hence

envelope trend helped here to replace the bearing and saved other components

failures.

14

15

16

The above tables are readings collected from a pump with velocity and

envelope readings. The pump is fall under class 2 as per ISO10816. The

velocity reading in April 2008 is 2.7 mm/sec and envelope reading is 16.95g.

The velocity reading is in green zone as per ISO 10816. Whereas the envelope

readings are in extreme level as per experience knowledge.

Hence recommended to replace the bearings.

After replacement the velocity readings reduced from 2.7 to 1.26 mm/sec

and envelope reading has reduced from 16.95 to 2.8g. After bearing

replacement both readings (velocity and envelope) are in normal condition.

17

The above plots are recorded before bearing replacement. The velocity

spectrum shows domination of 1x vibration amplitude and found no other

abnormal peaks. Hence it could be considered that 0.8 mm/sec of 1x vibration

is due to rotational force and considered to be normal for long-term

unrestricted operation. Also noticed cluster of peaks in mid range and higher

range of frequency in the spectrum.

But the envelope spectrum shows highest dominant amplitude 10.9 g @

1xrpm. Basically the 1x vibration in envelope readings are not interpreted as

like velocity spectrum as rotational force. This should be interpreted as high

impact or severe damage in one of the location in the raceways. Hence the

bearing need to be replaced at the earliest.

18

The plots shown in the slide is after replacement of bearings in the pump.

The velocity spectrum shows almost same 1x vibration amplitude before and

after bearing replacement. The overall vibration reduced from 2.7 to 1.6

mm/sec. The impact of overall vibration amplitude can be seen as no cluster

of peaks in mid range and higher range frequency in the spectrum. Hence the

cluster of peaks seen ( before bearing replacement) in the mid and higher

range frequency are belongs to impact of bearing failure.

The envelope spectrum after replacement also shows 1x rpm amplitude. But

the amplitude has reduced from 10.9g to 1.1g. Now this plot should be

interpreted like that the high impact in one location of the bearing due to

rotational force and load zone of the bearing.

19

This machine was running always >25 mm/sec over the period of 12 months. As per

ISO 10816 class 3, this machine should be stopped for maintenance when the

velocity readings are >11.2 mm/sec. but we ran this machine with confidence more

than a year based on envelope readings. As the envelope overall readings are <5 g.

Conclusion: This type of scenario is expected in any plant. When the machine need

for production it would hit trip limit as per ISO 10816. As per standard we need to

stop the machine for maintenance. But in this scenario, the source of high velocity

reading is due to casing distortion in the fan bearing NDE side. Actually this machine

was mounted on fabricated metal structure and the base plate of the mono block

mounting was bend and found early stage of crack. Hence we need to fabricate and

need long duration maintenance to re-install the structure of the fan. But this fan is

required for production and stoppage of this fan leads to stopping the whole

unit(plant). Hence we desired to run the machine till the bearing condition become

worse or started deteriorating.

Hence the combine analysis helped us to run the fan without any fear for more than

12 motnhs.

20

21

The above plots shows relationship between unbalance and envelope readings. When

unbalance increase the load on the bearing used to increase slightly. Hence the

envelop reading was increasing from 2 g to 5g.

In the above case the velocity readings are increasing depends upon the dust

accumulation in the fan. Hence the fan cleaning always planned based on the velocity

reading. When the envelope readings are increased planed for bearing replacement.

Hence the combine analysis helps to plan cost effective and correct type of

maintenance whenever opportunity comes for maintenance.

Based on this approach we reduced breakdown hours on account of rotating

machines from 300 hours (only velocity measurement) to 12 hours (velocity and

envelope) in a year after establishing the above concept.

22

Based on the above 4 case studies we can conclude that we can move forward in

determining the life of equipment or duration to plan the maintenance.

Hence we provide 6 month notice for spare procurement and 3 month notice for plan

the maintenance and one week notice for immediate action to operation and

maintenance team.

After applying the above concept we could able to reduce breakdown on account of

rotating equipment from 300 hours to 12 hours in the first year and 2nd year we

brought down further to zero.

Hence the combine analysis helped us to predict the machine problem well in

advance and we able to determine the duration for maintenance. Hence we could

able to plan the spares and manpower to carry out the job within available window

for routine maintenance. Hence we made zero surprise breakdown on account of

rotating machines.

23

Slide is self explanatory

24

25

Slide is self explanatory

26

27

28