Vibration-based Condition Monitoring – R.B.

Randall
Appendix – Exercises and Tutorial Questions – Selected Answers
Section A.1.1 Introduction and Background
1. See Sections 1.2, 1.3, 1.4.
2. See Section 1.3.2
3. See Sections 1.5.1, 1.5.2, 1.5.4.
4. See Section 1.5.4.1. Note that in some cases, such as with analogue-to-digital (AD)
converters, the value of the least significant bit defines both the smallest value that can
be measured, and also the smallest change in a signal that can be registered. However,
transducers with an analogue output often have a defined and possibly limited range
from minimum to maximum measurable value, but measurable changes in the value
within this range might be much smaller than the minimum measurable value. Spectral
components with very small amplitudes might still be detectable, when added to the total
signal, as long as the latter lies within the valid range of the transducer.
5. See Sections 1.5.1-1.5.4.
6. See Section 1.5.4.1.
7. See Section 1.5.5.
8. See Section 1.5.2.
9. See Sections 1.5.4, 1.5.4.1.
10. See Sections 1.5.3, 1.5.4, 1.5.4.1.
11. See Sections 1.5.6, 1.6.2.
12. See Sections 1.6.1, 3.6.1.
13. See Section 1.7
Section A.2.1 Vibration Signals from Machines
1. See Section 2.2.1.1.2.
2. See Section 2.2.5.2.
3. See Section 2.2.1.1.1.
4. See Section 2.2.1.4
5. See Section 2.2.1.4.3
6. See Section 2.2.1.2.
7. See Section 2.2.2.2.
8. See Section 2.2.2.
9. See Section 2.2.2.2. Note that the toothmesh frequency is generated by faults that are the
same for each toothmeshing event (ie the average over all toothmeshing events).
10. See Section 2.2.2.1.
11. See Section 2.2.2.2.
12. See Sections 2.2.3, 5.5.2.
13. See Section 2.2.3.
14. In Fig. A.2.1(a), there are a series of harmonics of four times motor shaft speed,
corresponding to the lobe meshing frequency in the compressor. These must be
mechanically transmitted back to the motor.
In Fig. A.2.1(b) the component at 2  mains frequency, typically from a stator anomaly,
 is five times (14 dB) stronger than the component at 2  shaft speed, which could result
from misalignment. See Sections 2.1.1.2 and 2.2.5.

Section A.3.1 Basic Signal Processing

A.3.1.1 Fourier Series
Example Fourier Series Calculation

 
1 
 

‐1  1  t   

 
‐1 
 

g (t )   t ,  1  t  1 

 
Sine Series (function is odd, so no cosine series)
T 2
2
T  T 2
bk  g ( t ) sin 2  f k t dt

1
 2 k 
 2
2   t sin  2
t  dt

1
1
   t sin  kt  dt
1
 
dv   sin  kt  cos  kt 
1
u  t , du  1 , , v
k
 
1 1
 t 
cos  kt  cos  kt  dt
1
   
 k  1 1  k
2
 1 

1
cos  k   cos   k     sin  kt 11
k  k 
 2 
  1 
k

  k 
2 2
 for k  1 bk   RMS 
 
1 1
for k  2 bk  RMS   
 2
2 2
for k  3 bk   RMS 
3 3
Complex Form
T 2
G fk  1
g ( t ) e  j 2  f k t dt
T  T 2
 
1
1
 t e  j  kt dt
2 1


  jkt e  jkt 
u  t , du  1 , dv   e , v 
 jk 
1
1  t e  jkt  1 e  jkt
1

2 1 jk
    dt
2  jk  1

1  1  jk 1 jk  2

1  jk
e  e jk 
  e  e  
2  jk jk  j 22k 2
j sin k 

1
cos k  
jk 2k 2
 j
 cosk 
k

 for k  1 G  f k  
j 2
RMS 
 
 j
for k  2 G  f k  
1
RMS 
2 2

for k  3 G  f k  
j 2
RMS   
3 3

No. k Ak ak bk
(1) 0 0 0
1 -2j/ 0 4/
2 0 0 0
3 2j/3 0 -4/3
(2) 0 0 0 0
1 2/ 4/ 0
2 0 0 0
3 -2/3 -4/3 0
(3) 0 1 1 0
1 -2j/ 0 4/
2 0 0 0
3 2j/3 0 -4/3
No. k Ak ak bk
(4) 0 0 0
1 -j/ 0 2/
2 j/2 0 -1/
3 -j/3 0 2/3
(5) 0 0 0 0
1 j/ 0 -2/
2 j/2 0 -1/
3 j/3 0 -2/3
(6) 0 1 1 0
1 j/ 0 -2/
2 j/2 0 -1/
3 j/3 0 -2/3
(7) 0 0 0
1 4/2 8/2 0
2 0 0 0
3 4/92 8/92 0
(8) 0 0 0 0
1 -4j/2 0 8/2
2 0 0 0
2
3 4j/9 0 -8/92
(9) 0 1 1 0
1 4/2 8/2 0
2 0 0 0
3 4/92 8/92 0
(10) 0 1 1 0
1 -4/2 -8/2 0
2 0 0 0
3 -4/92 -8/92 0

(11) for T/3. A0 = 1; Ak 

9
4 2 k 2
exp j2k 3  1 , k  0
2. Fouriier Series
1 T /2
G  fk    g (t ) e  j 2 f k t dt
T  T / 2

2 ktt
1 T /2 j
  g (t ) e T dt
T T / 2
(1)

2 kt
1 T 4 j
G( fk ) 
T 
T 4
Ae T
dt
T 4
 A   j 2Tkt  A  j 2k j
k

 e   e  e
2

j 2 k   T 4 j 2 k  
A k 
 sin 
 k  2 

k 
sin  
k 0 
0
buut for k  0  2 
( 
 1  G (0)
A
0 k  2
 
 2 
A   A
k 1  G (1)  sin   
 2 
A
k2  G (2)  sin    0
2
A  3   A
k 3  G (3)  sin   
3  2  3
(2)

1 T 4  A  j 2Tkt 1 T 2  A  j 2Tkt 1 A  j 2Tkt

T 4
G( fk )  T 2
T 2
e T 4
dt 
T 2
e T 4 2 e dt
dt 
T
 A    j 2 kt  2 kt T 4 
T 4 2 kt T 2
2  j   j  
   e T    e T    e T  
j 2 k    T 2  T 4   T 4 

 A  j k
2 e  e  j k   e j 2  e  j 2   e j 2  e  j 2  
k k k k
      
j 2 k       
A k  A
 sin   siin  k 
 k  2  2 k

A  sin(0) sin(0) 
k 0   0
2  (0) (0) 
but for k  0, sinn( k )  0
A k 
 G  fk  
 k  2 
sinn  ass for (1)

G1(0) = A/2
G(1) = A//
Sin(x))/x
 

G(5) = A/5
5
G(2) = 0 G(44) = 0
G2(0) = 0
G(3)
G = –A/3
(3)

2 kt
1 T 16 j
G( fk ) 
T T 16
Aee T
dt
T 16
A   j 2Tkt  A  j 8k j
k

 e    e  e 8

j 2 k   T 16 j 2 k  
A k 

 k  8 
sin 

A sin(0) A
G (0)  
8 (0) 8
A  
G (1)  sin  
 8
A  
G (2)  sin  
2 4
A  3 
G (3)  sin  
3  8 
A   A
G (4)  sin   
4  2  4
A//8
A/4

Sin(x)/x

0 1 2 3 4 5 6 7 8
A.3.1.2 Fourier Transforms
(1)

B A B
G ( f )   B2 Ae  j 2 ft dt  e  j 2 ft   B2
2  j 2 f 2

A A
 e j 2 f  e  j 2 f   sin  fB 
j 2 f f
sin  ffB 
 AB
 fBB 

AB
Sin(x)/x
G( f ) where
w x = fB

3 2 1 1 2 3 f
B B B B B B

G( f )


G ( f )

f
NB
N ± is the same
s angle, bbut the
ph
hase should be
b depicted aas odd
-
(2)

B A
G ( f )   Ae  j 2 ft dt 
B
e  j 2 ft 
0  j 2 f 0

A jA
 1  e  j 2 fB   cos  2 fB   1  j sin  2 fB  
j 2 f 2 f 

sinn  2 fB   j cos  2 fB B   1

A

2 f 
A
G( f )  sinn 2  2 fB   cos 2  2 fB B   1  2 coss  2 fB 
2 f
A A
 2  2 cos  2 fB   sinn  fB 
2 f f
AB sin  fB 
 ass for (1)
 fB
 2sin 2  fBB 
G(f)= =tan 1     fB
 2sinn  fB  cos  fB  
but signn of real andd imaginary y parts shouuld be taken into accoun
nt, which m
may give jum
mps of
 at ssome frequeencies

Amplituude is the saame as for (1)

(
Phase iss as depicteed below:

1 2 3
B B B f
3 2 1
B B B
-
(3)

0  C  B C 
G ( f )    C  t  e  j 2 ft dt    C  t  e  j 2 ft dtt
B
 B  0
 B 
B C 
 2   C  t  cos  2 ft f  dt byy symmetry y
0
 B 
2C B
=22C  cos  2 ft  dt  t cos  2 ft  dt
B

0 B 0
 t 1 
 uusing  t coss  at  dt  a sin(at )  a 2 cos(at ) 
B
2C 2C  t  2C
sin  2 ft   0  sinn  2 ft    cos  2 ft   0
B B
  2 
2 f B  2 f  0 B  2 f 
2C C
 1  cos  2 fB
f    sin 2  fB  
B  2 f 
2
B 2 f 2 

 sin  ffB  
2

 CB  
  fBB  
A.3.1.3 Convolutiion Theorem
m and Oth
her Methods
(1)

t 
= +
-A/2

g 2 (t ) = g1 (t ) + g3 (t )
Throughh linearity of
o Fourier trransform

G2 ( f ) = G1 ( f ) + G3 ( f )

A/2

f f -A/2

i.e. G2 ( f )  G1 ( f ) without the DC com

mponent

(2) 1
T

F.S. = F
F.T. ×
1
t
T

Questioon 2.(1), T  2B
AB
A sin  fB 
G  fk  
AB
 fB  1
2B
× 
1 2 t
B B 1
2B
A/2

=
1
2B
1
8B
Questioon 2.(2), T  8B  mulltiplying funnction =
1
8B

Peak vaalue is A/8, and there arre eight harm

monics to th
he first zero
o crossing
 (3)

g1 (t ) g 2 (t )


g 2  t   g1 t  B
2 
 G2  f   G1  f   e  G1  f   e j ffB
 j 2 f B
2

Same aamplitude ass G1  f  butt additive phhase   fB

B

G1 ( f ) G2 ( f )


-

(4) If
g1 (t )  rectangulaar function and
a g 2 (t )  triangular function
f
  B
B
g 2 (t )  g1 (t )  g1 (t )   g1 ( ) g1 (t   )d  g 2 (0)   g1 ( ) g1( )d   2
A2 d  A2 B
   B
2

= A2B

 =

g1 (t ) g1 (t ) g 2 (t )

sin  fB  AB A 2B2 = CB
AB
 fB 
×
=
G1 ( f ) G1 ( f ) G2 ( f )
2 2
 siin( fB)   sin( fB) 
G2 ( f )  A B 2 2
  CB  
 ( fB)   ( fB) 
A.4 Fault Detection
A.4.1
1. See Sections 4.2.1, 4.2.2.
2. See Sections 4.2.2, 4.2.3.
3. See Section 4.2.2.
4. (a) See Sections 4.2.1, 4.2.2.
(b) See Section 2.2.3.
(c) See Sections 4.2.2, 6.2.1.
5. See Sections 4.3.3, 5.6.2.2.
6. See Section 4.3.3. Note that there is an indication that both combustion pressure and
compression pressure are deficient.
A.5 Diagnostic Techniques
A.5.1
1. (a) See Section 5.4.3.1.
(b) See Sections 5.4.3.1, 2.2.2.2.
(c) See Sections 5.4.3, 6.2.1.
2. See Section 5.4.3.
3. See Section 3.4.3.
A.6 Prognostics
A.6.1
1. See Sections 6.2.1, 2.2.1, 4.2.1.
2. See Section 6.2.1.
3. See Section 6.2.1.
4. See Section 6.2.2.
5. See Sections 6.2.2, 5.5.2.3.
6. See Section 6.4.
7. See Section 6.4.3.