VIBRATION ANALYSIS

(PROFILES &
SPECTRUM)
LECTURE
03

Course Outline
LECTURE 01

Introduction to Plant Maintenance and its Regimes

LECTURE 02

Maintenance Planning, Failure Mechanisms, SHM

LECTURE 03

Vibration Analysis (Profile & Spectrum)

LECTURE 04

Corrective Techniques (Balancing, Align., Looseness)

LECTURE 05

Failure Mode Analysis

LECTURE 06

Visual Inspection (DPT & BT)

Week 01
Basic Maintenance & Damage Concepts

LECTURE 07
LECTURE 08

Vibration Analysis

Non-Destructive Testing
Radiographic Inspection (RT & NI)
Magnetic Electric Inspection (MPI)

Week 02

Week 03

Week 04

Lecture Outline
What is Vibration?
Why Monitor Machine Vibration?
How is Machine Vib. Described?

Vibration Spectrums
Energy in Spectrums
Lines of Revolution (LOR)
Analysis Techniques
Measurement Point Identification

Locating Turning Speed

Data Collection Techniques

Video-1

Most of us are familiar with vibration; a vibrating object moves

to and fro, back and forth. A vibrating object oscillates.

We often experience
examples of vibration in our
daily life

Rotating Machines Vibrate while Operating

Vibrating Belts

Vibrating Motors

Vibrating Fans

Vibrating pumps

Almost all machine vibration is due to one or more of

these causes:

Repeating Forces
Imbalanced

Misaligned

Worn machine

Looseness

Resonance

Improper
Driven

Imbalanced machine components contains heavy spot which

one rotated, exert a repeating force on the machine. Imbalance
is often caused by machine errors, non uniform material density ,
variation in bolt sizes, air cavities in cast parts, missing balance
weights, and broken deform, corroded, or dirty fan blades.

Uneven Rotor

Corroded tip

Bent Shaft
Dirt

Misaligned machine components create bending

moments which when rotated, exert a repeating
force on the machine. Misalignment is often
caused by inaccurate assembly, uneven floors,
thermal expansions, and distortions due to
fastening torque.

Parallel Misalignment

Angular Misalignment

Worn machine components exert a repeating force

on the machine because of the rubbing of uneven
worn surfaces. Wear in roller bearings, gears, and
belts is often due to improper mounting, poor
lubrication, manufacturing defects, and overloading

Worn Roller

Worn Teeth

Worn Belt

Improperly driven machine components exert a repeating

force on the machine because of intermittent power supply.
Examples include pumps receiving air in pulses, internal
combustion engines with misfiring cylinders, and
intermittent brush commutator contact in DC motors

Uneven
supply

Intermittent Brush
contact

Misfiring
Cylinder

Looseness can cause vibration in both rotating

and non-rotating machinery. Looseness is often
due to excessive bearing clearances, loose
mounting bolts, mismatches parts, corrosion,
and cracked structures

Excessive
Clearance

Loose
bolts

Resonance is an excitation of Natural Vibration.

Everything has a natural frequency fn, everything vibrates!
This fn does not cause any problems until another vibration source runs at a
same/similar frequency.

 A repeating force causing resonance may be small and may

originate from the motion of a good machine component.
Resonance, however, should always be avoided as it
causes rapid and severe damage.

. For example, whole bridges have collapsed due to their

natural oscillation rates being excited by the mere rhythm
of soldiers marching in unison across the bridges

The Tacoma Narrows Bridge in Washington opened to traffic

on 1st July 1940. Unlike most suspension bridges of its time
the Tacoma Narrows Bridge was unusually long and narrow.
The original design had called for trusses to stiffen the span
but funds were not available and inadequate stiffening was
used.
Wind blowing over and under the deck creates lift causing the
bridge to move up and down.
As the bridge was moving at its natural frequency the
resonance amplified the movement
Video- 2

 Monitoring the vibration characteristics of a

machine gives us an understanding of the 'health'
condition of the machine.
We can use this information to detect problems that
might be developing.

Below we discuss some common problems that can

be avoided by monitoring machine vibration.
Severe Machine damage

High power consumption

Machine unavailability
Delay shipments

Accumulation of unfinished goods

Unnecessary maintenance
Quality Problems

Bad Company image

Occupation hazards

Machine vibration that is not detected early enough will

often lead to severe machine damage requiring costly

repairs or even total machine replacement.

We cover all services and you

have a choice; a $200.000
operational or a $5000 funeral.

Just as it is costly to
detect medical problems
at a late stage, it is also
the same for machines.

It might be cheaper
to replace it!

A machine that is vibrating consumes more power.

We can minimize this problem by regularly
monitoring and maintaining the machine.

Because an unmonitored machine is more likely to break

down, it is more often out of action.

A machine should be consistently available to generate

the money to justify its investment.
This is so often
unavailable to make
money it is actually
losing money.

How?

Its unavailability does not

justify its initial cost, the
floor space it occupies, the
business opportunities we
miss out on!

Because an unmonitored machine is more likely to break

down, it is also more likely to cause delays in the shipping
of goods.
Ive got a dinghy!
Maybe we can
catch it.

Because an unmonitored machine is prone to breaking

down it is often unavailable.
Goods still in the making tend to get stuck at the input point of
the machine. This leads to unnecessary wastage.
Is there a
market for
sour milk?

This needs
processing now!
By tomorrow itll
be no good

Hmm!

To constantly ensure proper machine condition, some

companies stop running machines according to
predetermined schedules to adjust and replace parts
regardless of whether or not the machines are malfunctioning.
Its good that youre
getting full checkups
every week.

Its even better that

youre making me rich
unnecessarily!

Sometimes a machine can be running into trouble even

though it appears to be functioning normally.
If not caught early, the problem could lead to poor
quality products being made.

These are all wrong

because the machine
bearing is worn

And you didnt

know?
Have you any idea
what this will cost
the company?!

More than a vb
instrument?

Machines that are not regularly monitored can lead to shipment delays
and produce goods of poor quality.
Just one incident of shipment delinquency or product defect is often
enough to seriously strain or even end relationships with customers.

Im out of here.
Delayed
Quality
System
LTD.

A bad company image associated with shipment delays and poor

quality is something to be avoided.

Due to the noise and shaking they create, vibrating

machines can cause occupational hazards and human
discomfort.

Now What?

Vibration transmits as a sine wave.

We can measure this sine wave in an

Amplitude versus Time domain
This is called a Time Waveform

Video-3

Amplitude

How Much

Frequency

How Often

Phase.

When

Unfortunately there can be

multiple sine waves emitting
from a machine, So our
wave form becomes
complex

We call this a complex sine

waveform
Analysis is very difficult on this
type of waveform.
Fortunately we have whats
called an FFT which converts
the complex waveform back into
several simple waveforms

Video-4

Transducer
Amplitude

Overall
Energy

Waveform

Time

FFT
Amplitude

Spectrum

Frequency

The waveform is
converted to an
amplitude/frequency
domain

Tim
e

Amplitude

Amplitude

Amplitude

Complex waveform
changes to a
simple waveform

Before we learn how to diagnose

potential faults within a spectrum,
we need to understand the units of
measurement.
The vibration data that is
converted from the waveform by
the FFT process can be seen very
clearly
However there are a few
considerations we need to take into
account first.
The amplitude scale and the
amplitude units are important

As well as the frequency scale

and units

There are three different

amplitude measurements we use,
these are:
Acceleration

Velocity
Displacement

Acceleration
Measures the change in velocity over a period of time

Velocity
Rate of movement

Displacement
Measures total movement in relation to a reference point

A good way of remembering this is to think of a car:

From rest to 60mph is your acceleration (change in velocity over time)
Drive at 60mph for x-time (this is your velocity)
From start to finish is the total distance traveled (Displacement)

You can measure amplitude from a time waveform as shown:

0 to Peak
RMS

Amplitude

Average

Time t

Peak to Peak

The period t is the time required for one revolution of the shaft in this
illustration, which equals one cycle of the waveform
During this period, the amplitude of the waveform reaches a positive (+) peak,
returns to rest, and reaches a negative (-) peak before returning to rest

RMS

Peak (Pk) Amplitude measured from the at rest position (0) to the
highest value (0 to Peak)
Peak to Peak (Pk-Pk) Amplitude measured from the peak positive (+)
value to the peak negative (-) value
RMS (Root Mean Square) obtained by averaging the square of the signal
level over a period of time and then taking the square root result
Average (Avg) Amplitude value that averages the peak values of the
waveform
0 to Peak

RMS = 0.707 times the peak value

Avg = 0.637 times the peak value
Pk-Pk = 2 times the peak value

Average

Amplitude

You can calculate the different amplitudes

when one of the values are known:

Time
t

Peak to
Peak

Displacement - microns
Total movement, value is from
Peak to Peak
Ignores all high frequencies and
looks at the low frequency

Acceleration - G-s
Value from the base line to the peak
amplitude
Looks a force generated in our machine
(High frequency domain)

Velocity RMS - MM/Sec

RMS - root mean square,
appears at 0.707 the value of
the amplitude
Gives a good overall picture,
of the vibration in our machine

100
0.8
0.24

E6
E6
E6

19T - Example 6
19T - Example 6
-G3H19T
GBox
Shaft 26Inboard Horz
- Example
-G3H GBox Shaft 2 Inboard Horz
-G3H GBox Shaft 2 Inboard Horz

Route Spectrum
Route Spectrum
31-Jan-01
10:57:43
Route
Spectrum
31-Jan-01 10:57:43
31-Jan-01 10:57:43
OVRALL= 1.32 V-DG
OVRALL= 1.32 V-DG
P-P = 58.12
OVRALL=
1.32 V-DG
RMS = 1.31
LOAD
100.0
PK = =.6325
LOAD = 100.0
RPM
LOAD== 1052.
100.0
RPM = 1052.
RPS
RPM == 17.54
1052.
RPS = 17.54
RPS = 17.54

G-s
Acceleration
PK
Microns
in
mm/Sec
Displacement
S Velocity inin
P-P
RM

0.7
0.21
80

0.6
0.18

0.5
60
0.15

0.4
0.12
40

0.3
0.09

Velocity RMS
Good overall value
Acceleration G-s
Accentuates the high
frequencies
Good for anti-friction
bearings & Gears)

0.2
0.06
20
0.1
0.03
0
0
0
0
20000
0
20000
0
20000
Label: Chipped Gear Tooth
Label: Chipped Gear Tooth
Label: Chipped Gear Tooth

40000
40000
40000Frequency in
Frequency in
Frequency in

60000
60000
60000
CPM
CPM
CPM

80000
80000
80000

100000
100000
100000

Displacement microns
Accentuates the low
frequencies

Vib

Acceleration
accentuates HIGH frequencies,
and attenuates LOW frequencies.

Acceleration

Velocity
flat treats all frequencies equally.

Velocity
Displacement
accentuates LOW frequencies,
and attenuates HIGH frequencies.

Displacement
Freq
Video-5

Frequency refers to how often something occurs:

How often a shaft rotates?

How often a rolling element hits a defected race?

There are three ways to express frequency:

1.

CPM Cycles Per Minute

2.

Hz Cycles Per Second

3.

CPM / 60

Orders Multiples of Turning Speed

1CPM = 1RPM

Frequency/Turning Speed

Consider a motor has a rotational speed of 1485RPM, in terms of frequency this

equates to:
1485 CPM
24.75 Hz
1 Orders

(1rpm = 1cpm)
(1485/60) (minutes to seconds)
(1 x revolution of the shaft)

Synchronous energy - related to

turning speed.

We can see from the

spectrum that the first peak
is at 1 Orders (which means
it is 1 x turning speed)

All the other peaks are

harmonics off, which means
they are related to the first
peak

Examples of synchronous energy:

1) Imbalance 2) Misalignment

3)Looseness

4) Gears

Non-synchronous energy not related to turning speed

We can see from the

spectrum that the first
peak is at 10.24 Orders.
This is not related to
turning speed.

Examples of non-synchronous energy:

Bearings
Multiples of belt frequency

Other Machine Speeds

Sub-synchronous energy Less than turning speed

The spectrum shows the first

impacting peak below 1 Order.
This is sub-synchronous
energy
Examples of sub-synchronous
energy are:
Belt Frequencies
Other Machine Speeds
Cage Frequencies

Lines of Resolution (LOR) determine how clear the peaks(data) are

defined within our spectrum.
The more lines we have over the same F-max (Maximum
frequency scale). The more accurate our data will be
Example.
The diagram below shows data that has been collected using 400 LOR.
Notice how the top of the peaks are capped. When the LOR are
increased the data becomes more accurate.

TA16

0.20
0.5

L2 - TA 16
-M1H Motor Outboard Horizontal
Analyze Spectrum
13-Mar-01 09:14:16
09:13:53
PK = .3852
.7078
LOAD = 100.0
RPM = 1497.
1496.
RPS = 24.95
24.94

G-s
in G-s
Acceleration in
PK Acceleration
PK

0.16
0.4

The spectrum shown

displays data at 800 L.O.R
with an Fmax of 1600 Hz

0.12
0.3

0.08
0.2

0.04
0.1

0
0

400

800
Frequency in Hz

1200

1600

The second spectrum

displays the same data but
with 3200 L.O.R over the
same Fmax

There are 8 LOR settings we can choose from on the analyzer.

These start at 100 Lines and go up to 12800 Lines.

The average number of LOR is around 800 Lines for a typical

motor/pump set up

To change the LOR settings we need to alter our parameter set.

This is done in the Database Setup program

Remember. If you double your lines of resolution you double your

data collection time.

Measurement Point Identification

Locating Turning Speed

What is a Measurement Point?

A reference to a particular place/plane on
the machine to be monitored
What do we need them?
To know which machine is inspected?
Where the data was collected?
To ensure same data collection points
To ensure measurement repeatability

A measurement point is determined by three characters

M1H
side of the machine

Type of
Machine

Measurement
Plane

MOTOR INBOARD HORIZONTAL

H=
Horizontal

M = Motor

P = Pump

G=
Gearbox

V = Vertical

A = Axial

1 = Inboard(Drive End)
2 = Outboard (Non-Drive End)

Point identification of Pump

MOTOR (M)

PUMP (P)
Coupling
Bearing
(CB)

Point Identification Horizontal Mount Fan

Fan Bearing 1 (FB1)
MOTOR (M)

Coupling
Bearing
(CB)

Fan Bearing 2 (FB2)

When performing analysis on spectrums and waveforms, it is of utmost

importance to set the turning speed (running speed) correctly

When the turning speed has been located, the software will recalculate all the frequencies to this exact speed.

Once the turning speed has been set, it is now possible to

determine what is Synchronous/Non-synchronous and Subsynchronous energy.

The spectrum is showing

numerous impacts appearing at
different frequencies.

By locating the turning speed, it

is very clear that the impacts are
Non-synchronous

BFI - Example 6
-P2V PUMP OUTBOARD VERTICAL

Ex6

12

Route Spectrum
30-Jan-96 15:14:51
OVRALL= 13.52 V-DG
RMS = 13.46
LOAD = 100.0
RPM = 2987.
RPS = 49.78

10

RMS Velocity in mm/Sec

0
0

20

40

60

80

100

Frequency in kCPM

120

140

160

Have a look at the spectrum

below.

Where was the data taken?

EX3

1.0

When the turning speed has

been located
What type of energy is
present?

Los - Example 3
-P2V Pump Outboard Vertical
Analyze Spectrum
15-Nov-95 10:00:16
RMS = 1.27
LOAD = 100.0
RPM = 737.
RPS = 12.28

0.8

RMS Velocity in mm/Sec

0.6

P2V

Synchronous
Energy

0.4

0.2

0
0

6000

12000

18000

Frequency in CPM

24000

30000

 Divide spectrum in frequency bands based on the

types of mechanical faults that might appear on the machine

1X
2X
3X- 6X
BEARING BAND 1

BEARING BAND 2

9-30X RPM
30-50X RPM

 Divide spectrum in frequency bands based on the

types of mechanical faults that might appear on the machine

1X

Imbalance

2X
3X- 6X
BEARING BAND 1

BEARING BAND 2

9-30X RPM
30-50X RPM

 Divide spectrum in frequency bands based on the

types of mechanical faults that might appear on the machine

1X

Imbalance

2X

Misalignment

3X- 6X
BEARING BAND 1

BEARING BAND 2

9-30X RPM
30-50X RPM

 Divide spectrum in frequency bands based on the

types of mechanical faults that might appear on the machine

1X

Imbalance

2X

Misalignment

3X-Looseness
6X
BEARING
BAND
Bearing Band
1 1

BEARING
BAND
Bearing Band
2 2

9-30X RPM
30-50X RPM

 Divide spectrum in frequency bands based on the

types of mechanical faults that might appear on the machine

1X

Imbalance

2X

Misalignment

3X-Looseness
6X
BEARING
BAND
Bearing Band
1 1

BEARING BAND 2

9-30X RPM
30-50X RPM

 Divide spectrum in frequency bands based on the

types of mechanical faults that might appear on the machine

1X

Imbalance

2X

Misalignment

3X-Looseness
6X
BEARING
BAND
Bearing Band
1 1

BEARING
BAND
Bearing Band
2 2

9-30X RPM
30-50X RPM

Data Collection
Techniques

Although there are many different types of transducers

available, the most common type used for day to day
data collection are Accelerometers.

These transducers provide an electrical charge

proportional to acceleration by stressing piezoelectric
crystals typically 100mV/g sensors are used.

Imagine a doctor who listened to your heart through thick clothing and placed the
stethoscope closer to your kidney than to your heart. You would likely doubt his
diagnosis as he would be basing it on sounds distorted by undue obstruction
and noise from other organs.
We measure signals from the
heart

Heart

Vibration

We measure signals from

the bearings.

Doc, Ive been having

heart pain!

What!!

I have bad news.

You have no heartbeat!

When measuring vibration we must always attach the accelerometer

as close as possible to the bearing. More specifically, we must
attach it as close as possible to the centerline of the bearing to
avoid picking up distorted signals.

Protective
cover
Bearing
Location

Sensitivity

Stud
Mounting Mount
Pad

Dual Rail
Magnet

Flat
Magnet

Hand
Probe

Freq
1.5KHz

10KHz

32KHz

Whether it is your job to collect the data and/or analyse the

data it is important to understand that the technologies will not
give you the answer to a machines problem unless you have
collected meaningful, quality data

There are certain considerations that must be taken prior to any

data being collected, these are:
A good understanding of the internal make up of the machine, in order
to understand the best transmission path for data collection - bearing
locations, load zones etc.
Ensure data is collected in a repeatable manner so we can compare
two or more readings to each other - trending purposes
Variable speed machines - it is very important to collect data with the
correct running speed enter into the analyser

Route-based data collection is commonly used for walk

around vibration analysis

A maximum of 50 routes can be stored to each area

And one route can only contain 1044 measurement points

A route includes information from one area only

A route does not have to include all the equipment defined in that
area

The order of the equipment in the route can differ from that of the
database

Equipment can appear in more than one route BUT can not appear in
the same route twice

Route measurement points may not include all the points configured
on the equipment