BASIC MACHINERY VIBRATIONS

AN INTRODUCTION TO VIBRATION ANALYSIS

ENGR. TAMUNO KARIBO

Basic Machine Vibrations Overview

Measurement units
Physical Characteristics
Vibrations
Frequency
Period
Amplitude
phase
Basic Machine Vibrations Overview
Motions
Harmonic
Periodic
Vibration measures
Displacement
Velocity
Acceleration
Conversion of measures
Basic Machine Vibrations Overview
Vibration measurements
Phase measurement
Vibration analysis
Excitation
Natural frequencies, mode shapes and critical
speeds
 Vibration Units
The basic units used in this presentation to
describe vibratory forces and motions are:
Pound (lb.)
Inch (in.)
Second (sec.)
 Vibration Units (Cont.)
Amplitudes of vibration motion are commonly
described using the following measurement units:
Displacement, mils – peak to peak
(1,000mils – 1 inch)
Velocity, in/sec. – peak or rms (ips-peak or rms)
Acceleration, g’s – peak or rms (386.1in/sec² = 1g)
 Vibration Units (Cont.)
Frequencies are expressed in:
Cycles/minute (CPM)
Cycles/second (Hz, CPS) or
Orders (1x, 2x, ……10x)
Shaft speed is expressed in revolutions per minute (RPM)
Phase is expressed in degrees (deg.), in which one
revolution of a shafts or one period of vibration is 360°
Phase measurement is used in conjunction with a
vibration measurement to locate the position of a rotor
 System Properties
Mass = Volume x Density
Weight = Mass x g

or

Mass = weight/g
Mass = lb./in/sec² = lb. x sec²/in
g = gravitational constant
= 386.1in/sec²
 Harmonic Motion (Cont.)

The period is usually measured in milliseconds

(1,000 milliseconds per second)
Harmonic motion has only one frequency
 Natural Frequencies, Mode shapes and
Critical Speed
A natural frequency is a property of the system
design (mass and stiffness)
Resonance is a condition in which the excitation
frequency is at or close to a natural frequency
Critical speed is a resonance caused by a
rotating machine
A mode shape is a shape associated with a
system when it vibrates at a natural frequency
 Harmonic Motion
Frequency is defined as:
The number of cycles or event per unit time
Units of frequency are Hz (CPS), CPM and orders
Period is the time required for one cycle
τ – seconds/cycle
f = 1/τ = cycles/second
Harmonic Motion (Cont.)
Harmonic Motion (Cont.)
 Periodic Motion
Motion that repeats itself at equal time intervals
is periodic
Periodic motion can be broken down into a
number of harmonic motions
Example : (Next Slide)
16.67 mSec/cycle = 0.01667 sec/cycle
(1,000 mSec = 1sec)
f = 1cycle/0.01667 sec. = 60.00Hz = 3600CPM
Periodic Motion (Cont.)
 Amplitude
Amplitude is the level of vibration in a data
sample
Peak Amplitude is the largest value, plus or
minus, in a data sample
Peak to peak amplitude is the maximum
excursion of a vibration cycle measure on
adjacent positive and negative peaks
Rms = root mean square of spectrum
Rms = 0.707 x peak amplitude
 Vibration Measures (Cont.)
Measure Units Description
Displacement mils p-p Motion of machine, structure
or rotor, relates to stress

Velocity In/sec Time rate of motion, relates to

component fatigue

Accelerations g’s Relates to forces present in

components

1 mil = 0.001 inch; p-p = peak to peak

1g = 386.1 inches/sec²
Harmonic Conversion
 Harmonic Conversion (Cont.)

Velocity = 2π f D
Acceleration = 2π f V or (2πf)² D
D = peak displacement, inches
f = frequency, cycles /sec
V = velocity, in/sec
A = Acceleration, in/sec²
(divide by 386.1 in/sec²/g to obtain acceleration
in g’s)
 Basic Math

Procedure
Write formula
A = 2π f V

Write Known values

A = 2g
f = 5,000cpm
 Basic Math Procedure (Cont.)

Change to lb. –in. –sec. unit

A = 2 x 386.1 in/sec² = 772.2in/sec²
5,000 cycles
f= sec = 83.33Hz (cycles/sec)
60 min
min

Identity unknown
v = velocity, in/sec
 Basic Math Procedure (Cont.)

Rewrite formula to isolate unknown

A = 2π f V
A
V=
2πf
772.2
V=
2 x 3.14 x 83.33
V = 1.48 in/sec
 Conversion Examples
Example 1.1
Convert a displacement of 2mils peak to peak
at 1775 CPM to velocity in in./sec. – peak
2mils peak to peak = 1 mil peak = 0.001inch
peak
f = 1,775 CPM; f = 1,775 CPM/60 = 29.58 Hz
Then V = 2π f D
V = (6.28) (29.58) (0.001) = 0.186 in/sec.
 Conversion Examples
Example 1.2
Convert a velocity of 0.15 IPS – peak at 6,000Hz
to acceleration in g’s rms
A = 2π f V
A = (6.28) (6,000Hz) (0.15 IPS-peak)
A = 5,652 in/sec²
A = 5,652/386.1 = 14.64g’s peak
A = (14.63) (0.707) = 10.35 g’s rms
 Conversion Examples
Example 1.3
Convert an acceleration of 0.5 g’s rms at 1,775
CPM to displacement in mils, peak-to-peak
f = 1,775 cycles/min; f = 1,775 CPM/60 = 29.58 Hz
D = acceleration / (2π f)²
D = 0.5 g (386.1 in./sec²/g)/ [(2π x 29.58)²]
D = 0.0056 (2) (1,000 mils/in.)
D = 11.2 mils peak to peak
Waveform Analysis
 Excitation
Source Frequency (Multiple of RPM)
Fault induced
Mass unbalance 1x
Misalignment 1x, 2x
Bent Shaft 1x
Mechanical looseness Odd orders of x (may be 1x)
Casing and foundation distortion 1x
Antifriction bearings Bearing frequencies, not integer ones
Impact mechanisms Multi frequency
Design induced
Universal joints 2x
Asymmetric shaft 2x
Gear mesh (n teeth) nx
Couplings (m jaws) mx
Fluid film bearings (oil whirl) 0.43x – 0.47x
Blades and vanes (m) mx
Reciprocating machines Half and full multiples of speed, on design
 Signal Magnitude Determination
 OVERALL PEAK
The largest positive or negative zero to peak value in the
velocity or acceleration time waveform
 OVERALL PEAK TO PEAK
The largest peak to peak value (measured on adjacent
positive and negative peaks) in a displacement time
waveform
 OVERALL RMS (ROOT-MEAN-SQUARE), averaged or not
averaged
ANALOG: magnitude of filtered or unfiltered signal
passed through an RMS detector
DIGITAL: averaged energy signal as calculated by the
spectrum FFT
 Signal Magnitude Determination
 PEAK COMPONENT
Peak value (rms x 1.414) a component (bin) in the
spectrum – averaged or not averaged
 RMS COMPONENT
Rms averaged valve of a component (bin) in the
spectrum
peak = 1.414 x RMS
 NOTE
 rms cannot be obtained from the time waveform unless it has
a single frequency
 PEAK cannot be obtained from the spectrum unless it has
only one component
Signal Magnitude Determination
 Summary of Basic Vibration
Three important characteristics of vibration are
Frequency
Amplitude
Phase
The frequency is
The number of cycles per unit of time
The period is
The time required for one cycle of vibration
It is the reciprocal of frequency
 Summary of Basic Vibration (Cont.)
 Amplitude is the maximum value of vibration at a given
location on a machine. It is expressed in
mils (displacement)
in./sec. (velocity), or
g’s (acceleration)
 The amplitude of a vibration is expressed in units of
Peak
Peak to peak
rms
 Peak and rms are used with velocity and acceleration
 Mils peak to peak is used with displacement
 Summary of Basic Vibration (Cont.)
The measure of vibration -
Displacement (stress)
Velocity (fatigue) and
Acceleration (force)
Can be converted one to the other if the
vibration is a single frequency (harmonic)
Phase is the time relationship between vibration
and/or forces of the same frequency
A force, or excitation, causes vibration.
Vibration always lags force in time
 Summary of Basic Vibration (Cont.)
Vibration forces arise from-
Process variables
Improper design
Bad installation and
defects
Vibrations are analyzed in the time waveform and
the frequency spectrum
Natural frequencies are
A property of a machine system
Depend on mass and stiffness
 Summary of Basic Vibration (Cont.)
Resonance occurs when a forcing frequency is
equal to or close to a natural frequency
A critical speed is a special resonance in rotating
machine
Vibration is amplified at resonance
 CHAPTER 2
DATA ACQUISTION
 Data Acquisition
Selecting a measure
Transducer Selection
Proximity probe
Velocity transducer
accelerometer
Transducer mounting
Transducer location
Frequency Span
 Selecting A Measure
Table 2.1
MEASURE USEFUL PHYSICAL APPLICATION
FREQUENCY PARAMETER
SPAN

Relative 0 – 1,000 Hz Stress/ motion Relative motions in

displacement bearings/ casings
Absolute 0 – 20 Hz Stress/motion Structural motion
displacement (ISO Range
0 – 10 Hz)

Velocity 10 – 1,000Hz Energy/Fatigue General machine

condition, medium
frequency vibrations

Acceleration > 1,000Hz Force General machine

condition, high-frequency
vibration
 Selecting A Measure
Default frequency spans for Data collectors
Table 2.2
Component Span
Shaft vibration 10 x RPM
Gearbox 3 x GM
Rolling element bearing 10 x BPFI
Pump 3 x VP
Motors/generators 3 x 2 LF
Fans 3 x BP
Sleeve bearings 10 x RPM
 Procedure: Measure Selection

Calculate default frequency spans (Fmax), or

spans based on experience
Compare to useful frequency ranges of measures
Select seismic measures for high and low
frequencies encountered on pedestals
Select relative displacement measure where non-
contacting displacement sensor measure effective
shaft vibration
 Example: Measure Selection
 Example 2.1: Select a measure for the 9Mw single-reduction
gearbox described in Table 2.3
 The gearbox uses fluid-film bearings and is sufficiently large
(greater than 500HP) to justify permanently-mounted
proximity probes for evaluating the position of the journal in
the bearing and ratio of the vibration to the bearing
clearance. The analyst is thus able to assess severity of the
shaft vibration of the journal.
 Because the gear-mesh frequency (3,00Hz) is greater than
1,000Hz (see table 2.1), casing acceleration should be
monitored and analyzed. Frequency spans of 10,000Hz (See
Table 2.2) should be monitored in acceleration and
75,000CPM (high-speed input) and 12,000 CPM (output)
respectively in shaft vibration from the proximity probes
 Example: Measure Selection
 Example 2.2: Select a measure for a low speed 300RPm
dryer roll (see Table 2.3)
 The multi-ton roll is mounted on large spherical roller
bearings with 26 elements per row. Mass unbalance is not a
major consideration due to the low speed which results in a
relatively small unbalance force. The highest rolling element
bearing frequency is the ball pass frequency of the inner
race. It can be estimated as
BPFI = (0.6) (RPM) (N)
BPFI = (0.6) (300) (26) = 4,680 CPM (78Hz)
 Therefore, the frequency span is 780Hz (See table 2.2). This
value is within the velocity range. (See table 2.1)
 Example: Measure Selection
 Example 2.3: Select a measure for a 200HP – four pole induction motor
with eight rolling elements in the bearings.
 The operating speed vibrations have a frequency of 1,800CPM (30Hz)
and a frequency span of 300hz, which is within the velocity range. The
bearing frequency span is
(BPFI) (10) = (0.6) (8) (1800) = 86,400 CPM (1,440Hz)
 Because the majority of the activity is in the velocity range, a velocity
transducer can be used even though some activity is above 1,000Hz.
The useful frequency spans of all measures overlap. Therefore, the
measure should be selected from the predominant portion of the
frequency activity of the component. For example, if the default
frequency span for the bearing had been 2,880Hz (16 rolling element
bearings), acceleration would have been selected as a measure for
the bearings. Unfortunately, the shaft vibration span of 300Hz remains
within the velocity range. Therefore two measures, velocity and
acceleration are required
 Selected Measure
Table 2.3. Measure for Selected machines
Machine Hp/Mw Speed (RPM) Bearing Type Measure(s) Transducers
Large 4,000hp 3,600rpm Fluid-film Displacement Proximity probe
induction -Shaft
Motor
Induction 200hp 1,800rpm 8 rolling Velocity – Integrated
Motor elements casing accelerometer
or velocity
Diesel Engine 400hp 1,800rpm Fluid film Velocity – Integrated
casing accelerometer
or velocity
High 18,000hp 5000rpm Fluid – film Displacement Proximity probe
Performance – Shaft Integrated
Centrifugal Velocity accelerometer
pump casing or velocity

Reciprocating 200hp 300rpm 15 rolling Velocity – Integrated

pump elements casing accelerometer
 Selected Measure
Table 2.3. Measure for Selected machines
Machine Hp/Mw Speed Bearing Measure(s) Transducers
(RPM) Type
Centrifugal 1,000hp 5,000rpm Fluid-film Displacement Proximity probe
Compressor -Shaft
Reciprocating 500hp 400rpm Fluid - Velocity – Integrated
compressor Film casing accelerometer
or velocity
Dryer Roll 300rpm 26 rolling Velocity – Integrated
elements casing accelerometer
or velocity
 Transducer Selection
Measure
Signal Strength
Sensitivity
Size
Frequency response
Signal processing (integration, amplification and
filtering)
 Proximity Probe
 Relative displacement sensor
 Sensitivity is 200MV/mil
 DC signal is proportional to distance from shaft
 AC signal is proportional to vibration
 Negative 24 volt power supply
 Range: 5 to 80 mils
 Flat Frequency response
 Sensitive to shaft imperfections
 Mechanical (scratches, nicks, etc.)
 Metallurgical (residual magnetic fields)
 Velocity Transducer
 Absolute vibration velocity
 Self-generating signal; no power supply
 Sensitivity is 100 – 1,000 MV/IPS
 Frequency response is flat from 10Hz to 2,000Hz (ISO limit is 1,000Hz)
 Absolute shaft vibration (with fishtail)
 Accelerometer
 Absolute vibration acceleration
 24 volt power supply
 Sensitivity is 5 – 1,000 MV/g
 Size controls frequency response
 Signal integrated for velocity, double integrated for
displacement
 Subject to cable noise
 Low frequency integration noise can produce
“ski-slope” in spectrum plot
 Triggering Devices
Provide 1x or Nx signal for phase, synchronous time
averaging or orbits
Optical pickup
Proximity probe
Magnetic pickup
Strobe light
OPTICAL PICKUP MAGNETIC PICKUP
 Transducer Mounting

Method Frequency limit

Hand Held 500 Hz
Magnet 2,000 Hz
Adhesive 2,500 – 4,000 Hz
Bees Wax 5,000Hz
Stud 6,100 – 10,000 Hz
Transducer Mounting
 Measurement Points
Close to the bearing
Horizontal/Vertical bearing centerline
Axial in load zone
Signal transmission sensitivity
Machine design
Measurement Points
 Transducer Location
 Radial Direction: operating speed (1x) faults such as mass
unbalance, looseness and misalignment
 Axial Direction: gear and rolling element faults

Machine vibration
 displacement
 velocity
Transducer Location
Time and Frequency Relationship
Data Display
 Summary of Data Acquisition
Rolling element bearing and gear-excited
vibrations are measured in the axial direction
Select proper frequency spans and lines so that all
vibration activity is captured with adequate
resolution
Set up the data collector for acquisition to provide
data displays that enhance analysis
Measures – Displacement, velocity or acceleration
– to evaluate machine condition are machine
specific
 Summary of Data Acquisition
Vibration transducers should be selected for
frequency response, signal strength, size, machine
type and bearing type
Signal strength depends on the measure selected
and the frequency of interest
Acceleration signals are
Small in magnitude at low frequencies
As are displacement signals at high frequencies
Integration of acceleration signals can cause
large-magnitude low frequency noise
 Summary of Data Acquisition
Frequency response is
The ability of a transducer to reproduce the
magnitude of vibration
Within a given frequency range
Vibration transducers should be:
Located close to the bearing and
Mounted to acquire data at the frequency of interest
Vibration at operating speed (1x) such as mass
unbalance, misalignment and looseness are
Monitored in the radial direction
Analyzed in the axial and radial direction
 CHAPTER 3
DATA PROCESSING
 Data Processing
Oscilloscope
FFT Spectrum Analyzer
Electronic data Collector
Data Sampling
Data Display
Windowing
Window Selection
Resolution
Data Collector Setup resolution
 Oscilloscopes
Triggering Controls display
External intensity input Z-axis
Vertical amplifier Voltage scale
Horizontal time base sec or msec
Time base can be switched to amplifier, i.e.
voltage scale
Hor amp + Ver amp + z-axis = ORBIT
Oscilloscopes
 Oscilloscopes

Example 3.1: Find the amplitude and period of the

time waveform shown in Figure 3.2
Time base setting = 10 msec/div
Amplifier setting = 0.2 volt/div
Transducer calibration = 1,000 mv/in./sec.
Oscilloscope Analysis
 Oscilloscope Analysis

Period = (4div) (10msec/div) = 40msec = 0.04 sec

Frequency = 1/0.04 sec = 25 Hz = 1,500 CPM
Amplitude = (1div) 0.2 volt/div = 0.2volt = 200mv
200mv/1,000mv/in./sec. = 0.2 in/sec
 FFT Analysis
Fast Fourier Transform (FFT)
Computer-based instrument
Analog to digital converter
Digitized data block
Number of samples is power of 2
Sampling rate up to 256,000 samples /sec
Number of lines limited 100,200,400… up to 6400
May have zoom – maintain the number of lines
but decrease fmax
 Electronic Data Collector
FFT based analysis – similar features
Large data storage capability
Portable, less than 10 lbs
Route friendly
Overall or band monitoring
Downloadable to computer
Contains programs to simulate diagnostic
instrument
WINDROCK VIBRATION ANALYZER
 Data Sampling
1,024 samples for 400 lines
Ts = data acquisition time = number of lines/Fmax
Lowest resolvable frequency = 1/Ts
Data Sampling
 Data Display
Spectrum plot and Time wave
 Window Selection
Window Purpose Amplitude Window Factor
uncertainty (WF)
Uniform Impact test 56.50% 1.0
Hanning Fault Analysis 18.80% 1.5

Flat top Condition 1.00% 3.8

Monitoring

Frequency Span
Bandwidth = (WF)
Number of lines
Resolution = 2 x Bandwidth
 Dynamic Range

Capability of analyzer to resolve small-amplitude

frequency components in the presence of large
amplitude ones (usually acceleration)
Auto Ranging
dB = 20 LOG V/VREF
V/VREF 10dB/20
 Data Collector Setup Resolution
Example 3.2: Calculate the lines of resolution and Data
Acquisition time
 Data are being acquired on a two-pole electric motor
with a suspected air-gap problem (vibration is occurring
at 120Hz or 7200CPM)
 The motor operates at 3,580 RPM. Two times operating
speed is 7,160CPM
 The difference between two times line frequency and two
times operating speed is 40 CPM
 If the frequency span of 500 Hz (30,000CPM) were chosen,
what number of lines of resolution will be required if a
Hanning window is used?
 What will be the data acquisition time
 Data Collector Setup Resolution
Resolution = 2 x (Fmax/N) x WF, or rearranging
the equation
No. of lines = 2 x (Fmax/resolution) x WF
No. of lines = 2 x (30,000/40) x 1.5 = 2,250
Therefore 3,200 lines would be required
Data Acquisition time = no of lines/ Frequency
Span
3,200/500Hz = 6.4 sec
 Data Collector Setup Resolution
Example 3.3: Lines of resolution
Data are being acquired from a fan operating at
956RPM
The fan is mounted close to a second fan that
operates at 970RPM
How many lines of resolution are required to
resolve shaft frequency using a Hanning window
and a minimum frequency span of 10 times the
operating speed (9700CPM)
Since 12,000 CPM is an available frequency span
on the data collector, it is selected
 Data Collector Setup Resolution
Resolution required = 970 CPM – 956 CPM = 14
CPM
The 14 CPM = [2 x (12,000CPM) x 1.5)]/ no. of lines
No. of lines = [2 x (12,000 CPM) x 1.5] /14 CPM =
2,572 lines
Therefore 3,200 lines, the next highest number of
lines available on the data collector, would
provide sufficient resolution
Data Collector Setup Resolution
Summary of Data Processing
 Summary of Data Processing
The oscilloscope can display orbits (x-y) of a
shaft or pedestal motion by changing the time
base to the horizontal amplifier
Shaft rotational speed and number of revolutions
of the shaft per cycle of vibration can be
displayed using z-axis intensification of the
oscilloscope
Triggering signal obtained from a proximity
probe or an optical pickup can be used to
initiate and/or control data acquisition
 Summary of Data Processing
The FFT analyzer displays a time waveform and a
spectrum from a digitized block of data
FFT data are displayed in discrete lines called
bins. Frequencies between these lines cannot be
distinguished
The lowest resolvable frequency is 1/Ts
The data acquisition time is second for a block
of data used by an FFT analyzer to produce a
spectrum is equal to the number of lines divided
by the frequency span in Hertz
 Summary of Data Processing
 The Hanning window is used in general data collection
because it is a good compromise between amplitude
accuracy and frequency resolution
 The uniform window is used for impact because it starts
and ends at zero without compromising the initial data
sample
 The amount of resolution determines whether or not
closely spaced frequencies can be viewed in the
spectrum
 The resolution of an FFT analyzer setup is calculated by
multiplying 2 times the window factor by the ratio of the
frequency span to the number of lines
 CHAPTER 4
FAULT DIAGNOSIS
 Fault Diagnosis
Fault diagnosis techniques
Operating speed faults
Rolling element bearings
Gearboxes
Electric Motors
Centrifugal and Axial machines
Pumps
Fans
Compressors
 Fault Analysis Technique

Fault analysis related to frequencies

Known machine forcing frequencies matched to
measured spectrum
 Fault Diagnosis Technique
Technique Use Description Instrument
Time waveform Modulation, Amplitude vs. Analog and
analysis pulses, phase, Time Digital
truncation, Glitch Oscilloscope, FFT
Spectrum
Analyzer
Orbital analysis Shaft motion, sub- Relative Digital vector
synchronous whirl Displacement of filter, Oscilloscope
Rotor in bearing in
xy direction
Phase analysis Force/Motion Relative time Strobe light,
Relationships between force digital vector
and vibration filter,
signals or Oscilloscope, FFT
between 2 or Spectrum
more vib. Signals Analyzer
Spectrum Analysis Direct Amplitude vs. FFT Spectrum
frequencies, Frequency Analyzer
Natural freq. etc.
 Format Identification
A: MAG – the vertical axis of the spectrum in root
mean square (rms) units. Peak is calculated by
multiplying the amplitude of the spectral
component by 1.414
RMS: 10 – ten is the number of rms averages in
the spectrum
mIPS – milli-inches per second. For example
200mIPS = 0.20 inches per second
mIPS/DIV – number of milli-IPS per division –
there are ten divisions on the spectral scale
 Format Identification (Cont.)
STOP – 1,250Hz – the Fmax is 1,250Hz or 75,000CPM
B: STORED – the time waveform. It is typically
indicated by B: TIME
BW: 11.936Hz – the bandwidth. BW = (Fmax/no. of
lines) (window factor)
This analyzer has a fixed number of lines (400). A
flat top window was used when the data were
acquired
Therefor, BBW = 1,250Hz/400 lines(3.82) = 11.94 Hz
STOP : 80msec – the time waveform has an
80msec span, or 0.008seconds per division
 Format Identification (Cont.)
X: 84.375 Hz – Horizontal coordinate (frequency in
Hz) in the spectrum at the cursor marker
Y: 73.48mIPS – vertical coordinate (amplitude in
rms) at the cursor marker

73.48mIPSrms = 0.073mIPSrms = 0.104 IPS peak

BND: 218.6 mIPS – the overall rms value of the

component in the spectrum over the frequency
span selected. Thus rms = 0.219 IPS
 Spectrum Analysis
Step 1

Identify the operating speed(s) and its multiples

(orders)

The data in Fig. 4.1 are in a frequency and order

format
Operating Speed Frequencies
 Spectrum Analysis
Step 2

Identify dominant frequencies that are multiples of

operating speed

These include blade pass in fans, vane pass in

pumps, and gear mesh in gears (figure 4.2)
Gear Mesh
 Spectrum Analysis
Step 3

Identify nonsynchronous multiples of operating

speed such as bearing frequencies (Figure 4.3)
Nonsynchronous Multiples (8.12 x RPS)
 Spectrum Analysis
Step 4

Identify beat frequencies

Two frequency component close to each other,

their amplitude add and subtract during the beat
cycle (Figure 4.4)
Beats
 Spectrum Analysis
Step 5

Identify frequencies that do not depend directly

on operating speed such as natural frequencies or
frequencies from adjacent machines (Figure 4.5)
Natural frequencies or other frequencies
unrelated to Operating Speed
 Spectrum Analysis
Step 6
 Identify sidebands (Figure 4.6) that are related to a low-
frequency component of vibration that modulates
(changes) the amplitude of a high-frequency vibration
 Sidebands are frequency components that appear in a
spectrum in addition to a dominant frequency such as
gear mesh
 Modulation of the gear mesh vibration of a gearbox by
pinion or gear eccentricity (figure 4.6) is a good example
 A Sideband identifies the location of a fault if its frequency
matches the speed of a machine component
Sidebands
 Operating Speed Faults
Table 4.2 Identification and Correction of malfunctions of Rotating machinery
Fault Frequency Figure # Spectrum, Time or orbit Correction
Critical speed 1x, 2x, 3x, etc. 4.7 Amplified vibration due Tune natural
to proximity to operating frequency
speed
Mass 1x, Low 4.8 Distinct 1x with much Field or shop
Unbalance Amplitude lower values of 2x, balancing
orders 3x,etc (circular orbit
shape)
Misalignment 1x. 2x 4.9, 4.10 Distinct 1x with equal or Perform hot
occasionally 3x higher 2x, 3x (8-orbit and/or cold
shape) alignment
Shaft Bow 1x 4.11 Dropout of vibration Heating or
around critical speed pruning to
straighten
rotor
Fluid file 1x, sub 4.12 High 1x, high 1/2x Replace
bearing wear, harmonics, .Cannot be balanced bearing
excessive orders
clearance
 Operating Speed Faults
Table 4.2 Identification and Correction of malfunctions of Rotating machinery
Fault Frequency Figure # Spectrum, Time or orbit Correction
Resonance 1x, 2x, 3x, etc 4.6 High balance sensitivity, Change
high amplitude vibration structural
at order of natural freq. natural freq.
Looseness 1x plus large 4.13 High 1x with lower level Shim and
number of orders. Large 1.2order, tighten bot
order, 1/2x low axial vibration to obtain
may show rigidity
Eccentricity 1x 4.14 High 1x Machine
journal for
eccentricity
Coupling 1x, 2x, 3x, etc 4.1 Similar to misalignment Replace
lockup coupling or
remove
sludge
Distortion 1x and orders 1x from preload of Check soft
bearings, 2x line foot
frequency.
Mass Unbalance
Excessive Bearing Clearance
Misalignment
Misalignment and coupling lockup
Resonance
Looseness: Displacement Data Display
Looseness: Velocity Data Display
Eccentricity
 Distortion
May cause preload on bearings: dominant 1x
vibration
Soft foot condition: gear mesh vibration
Distorted motor casings: 2x line frequency
Phase analysis may assist in evaluation of data
 Rolling Element Bearings
Bearing frequencies
Measurement techniques
Analysis techniques
Defects
High Frequency detection (HFD)
Bearing Nomenclature
 Bearing Frequencies
Ball Pass Frequency of the outer race (BPFO);
generated by balls or rollers passing over
defectives races
Ball Pass Frequency of the inner race (BPFI;
generated by balls or rollers passing over
defectives races
Ball Spin Frequency (BSF); generated by ball or
roller defects
Fundamental train frequency (FTF); generated by
cage defects or improper movemens
 Bearing Frequencies (Cont.)
Φ = contact angle; angle between lines
perpendicular to the shaft and from the center of
the ball to the point where the arc of the ball and
the race make contact
N = Number of rolling elements (Balls or Rollers)
P = pitch diameter; use average value if exact
value is not known, inches
B = Ball or roller diameter, average value for
tapered bearings, inches
Hz = speed of rotating unit in revolutions per second
Bearing Frequencies (Cont.)
Bearing Frequencies Example
 Bearing Frequencies Example (Cont.)
BPFI = (29/2) 3.55 [1 + 0.0883] = 56 Hz (3,360cpm)

BPFO = (29/2) 3.55 [1 - 0.0883] = 46.9 Hz (2,814cpm)

BPFI = [(29.11/(2) (2.598)] 3.55 [1 - 0.0883²] = 19.7 Hz

 (1,182cpm)

FTF = 3.55/2 [1 + 0.0883] = 1.6 Hz (96cpm)

 Bearing Frequencies Sample Bearing #1
6224 Single Row , Radial

Number of balls (Nb) = 8

Ball Diameter (Bd) = 1.187”
Pitch diameter (Pd) = 6.594”
Contact angle = 0°
Unit Rotating Speed = 1775rpm
Bearing Frequencies Sample Bearing
Calculation #1
Bearing Frequencies Sample Bearing
Calculation #1 (Cont.)
 Frequency Zones
(Areas of Spectral response)
High- Frequency Detection (HFD) Methods (Cont.)

Measure of impulse/impact/noise from any

machine component
Trended for failure initiation
Trended for lubrication
Also responsive to
Steam leaks
Flow turbulence
 Measurement Techniques

Close to the bearing

Axial and/or Radial
Load zone
 Analysis Techniques

NARROW BAND ANALYSIS

Calculating bearing frequencies
Measuring and analyzing vibration signals
Identifying sidebands and frequencies in the
spectrum
Evaluating the spectrum and the time waveform
for the shape, energy and amplitude
Low amplitudes common
 Rolling Element Bearing Defects
Table 4.3. Analysis of Rolling Element Bearing Defects
Defect Frequency Time waveform/ Comment figure
Spectrum shape
Outer race BPFO Multiples of BPFO Shallow Flaking 4.3
defect increasing in
severity after 1
year
Inner race BPFI and Decreasing size of Shallow Flaking 4.16
defect multiples harmonics
Inner Race BPFI and Decreasing size of Bearing lasted 4.17
Defect multiples harmonics. Modulate 14 days
natural frequency
Ball Defect FTF By Operating Speed. Balls rattle 4.18
Modulated by FTF bearing at MNF
Excessive RPS Multiples of RPS Bearing showed 4.19
internal excessive wear
clearance
Nonsynchronous Multiples (8.12 x Rps)
Inner Race Defect (IRD)
Extensive Inner Race Defect (IRD)
Ball Defect
Internal Clearance
(Worn out ring in Load zone)
 Gearbox Analysis

Measurements
Analysis
Type of faults
Example of faults
 Measurements

Close to the bearings

Axial for gear mesh
Radial for 1x
 Gearbox Analysis
Gearmesh problems – use the spectrum
Scoring
Pitting
Uneven wear
Poor quality
Low contact ratio
Cracked, chipped or broken teeth – use time
waveform
Use Gearmesh frequencies and sideband of
operating speeds
Gearbox Calculations
 Gear Frequency Calculations
Example 4.2: Gear Frequency Calculations
The gearbox in Figure 4.20 is driven by a two-pole motor at
3,585 RPM. What are the gear mesh frequencies and shaft
speeds:
Input shaft speed = 3,585 RPM (59.75Hz)
Intermediate shaft speed = (3,585 RPM [26T/101T]) = 923 RPM
Outer shaft speed = (923 RPM) [31T/97T] = 295 RPM
High speed gear mesh = (3,585 RPM) (26T) = 93,210 CPM or
1,553.5 Hz
 Gearbox Faults
Table 4.4. Identifications of malfunctions in Gears and Gearboxes
Fault Frequency Time waveform/ Spectrum shape figure

Eccentric Gears Gear mesh Gear Mesh with Sidebands at 4.21

frequency of eccentric gear
Gear-mesh wear Gear Mesh Gear mesh with sideband at 4.2
frequency of worn, scored or
pitted gears; sometimes ½, 1/3 ¼
harmonics of gear mesh
Improper Gear mesh Gear Mesh with orders and 4.6
backlash or end sidebands at frequency of pinion
float or gear

Broken, cracked Natural Pulses in the time waveform. 4.22

or chipped gear frequencies
teeth
Gear distortion Gear mesh Gear Mesh with orders in 4.23
Nat. Freq spectrum
Eccentric Pinion
Gear Mesh
Gear Mesh Backlash
Broken Gear Teeth
Misaligned or Distorted Gearbox
 Electric Motor Basics
Line frequency = 60Hz (USA) and 50Hz (Nigeria)
except in variable frequency motors
Nominal motor speed (NMS) = 2x Line Frequency/
Number of poles
Slip Frequency = NMS – Actual Speed
Poles – Electric Magnets
Rotor Bars) conductors of rotor
Slots (Winding containers of Stator)
 Motor: Synchronous Speed
Example 4.3: Motor: Synchronous Speed Calculation

What is the synchronous speed of an 8-pole induction

motor operating on 60Hz power

NMS = (2) (60)/8 = (15 cycles/sec) (60sec/min) = 900 RPM

 Calculation of Slip Frequency for an
induction motor

Example 4.4: Calculation of Slip Frequency for an

induction motor
A 4-pole induction motor operates at 1,774 RPM
What is the slip frequency in Hz
NMS = (2) (60)/4 = 30 Hz
Slip Frequency = 30 Hz – (1,774/60) = 0.433 Hz (26 CM)
Motor construction
 Motor Electrical Faults
Table 4.4. Identifications of malfunctions in Gears and Gearboxes
Fault Frequency Time waveform/ Spectrum figure
shape
Air-Gap variation 120Hz 120Hz plus sidebands, beating 4.25
2x with 120Hz
Broken rotor bars 1x 1x, Sidebands equal to 4.26
(Number of poles x Slip
Frequency)
Eccentric rotor 1x 1x, 2x, 120Hz beats possible 4.27
Stator flexibility 120Hz 2x 120Hz Beats
Off magnetic 1x RPS, 2X Impacting in Axial direction 4.28
center RPS
2X LF
Stator shorts 120Hz and 120Hz and Harmonics 4.29
harmonics
Air Gap Variation
Broken Rotor Bar
Eccentric Rotor
Magnetic Center
Stator Shorts
 Centrifugal and Axial Machines

Pumps and fans transport fluids by converting

mechanical work into energy of the fluid in the form of
pressure and velocity (Flow)
Compressors increase energy of the compressed fluid
as pressure
Mechanical energy from steam and gas turbines,
motors and IC engines
 Centrifugal and Axial Machines (Cont.)

Type of flow depends on impeller/ blading

arrangement
Performance characteristics related to head (pressure),
efficiency, flow and horsepower
Machine must operate on negative slope of head vs
flow curve (fig. 4.30) or it will be unstable
Flow Characteristics
 Common Pump Faults
Critical Speeds
Structural looseness (Principally vertical pumps
Acoustical resonance (Piping design)
Impeller eccentricity (nonconcentric machining,
deflection of impeller shaft due to head)
Impeller balance
Impeller/diffuser clearance (gaps)
Recirculation (low flow)
Cavitation (low suction head)
Oil Whirl (Bearing design and excessive clearance
 Flow Characteristics

At best efficiency design point, fluid discharge angle matches

angle of diffuser and flow is smooth with minimal disturbance
If flow is decreased (too much back pressure) or is increased (too
little back pressure), the fluid angle no longer matches the
discharge angle, resulting in high vibration and loss in efficiency
Flow Characteristics
 Recirculation
Back pressure too high
Visible axial shuttling
Rolling element bearing failure
Flow noise
Recirculation line 1” diameter instead of 3” diameter
Recirculation (cont.)
 Cavitation
Flow too high for head
 Fans
Mass Unbalance (Figure Rolling element bearings
4.36) Soft foot
Misalignment Impeller eccentricity
Critical Speeds Belts and pulleys
Resonance
Looseness ( figure 4.37)
Aerodynamic problems
Isolator problems
Oil whirl
Fan Characteristics
 Fan Calculations
Example 4.5: Fan Drive belts
A fan is belt-driven by a four-pole motor at 1,779
RPM with a nine inch pulley
If the fan has an 11.75inch pulley attached, what
is the fan speed?
Fan speed = [1,779 RPM (9 in.) /11.75 in.)]
=1,362.6 RPM
 Fan Calculations (cont.)
Example 4.6: Fan Blade pass frequency
What is the blade pass frequency of an ID fan that
operates at 896 RPM and has 12 blades
BPF = [896 RPM (12 BLADES)]
=10,752 RPM (179.2Hz)
Fan Mass Unbalance
Looseness in Fan bearing
Aerodynamically induced Fan vibration
 Centrifugal Compressors
Massive casings ( high pressure)
Lightweight rotors
Minimum flow point (surge limit) – flow reversal occurs
High speeds
Synchronous and sub-synchronous operating speed
faults mostly related to bearings