VIBRATIONS(From: Theory of vib. with App. By W. T.

Thompson) • Degrees of Freedom: It is the number of coordinates required to

• All bodies possessing mass and elasticity are capable of vibrations. describe the motion of a system. A particle in space will have three
• Most Engineering Machines and Structures experience vibration degrees of freedom and a Rigid body will have six degrees of freedom
and their design requires consideration to oscillatory behavior. (three additional for rotation).
• Vibrating system can be linear or non-linear. Mathematical
techniques are well developed for linear. Nonlinear systems are HARMONIC MOTION
less well known. We will deal with linear systems. When a motion is repeated in equal intervals of time , it is called
• Free vibration: System oscillates without any external force. The periodic motion. The repetition time  is called the period of the
system vibrates at its natural frequency. oscillation, and its reciprocal,
• Forced vibration: System oscillates under external force. The system f=1/  is called the frequency
vibrates at forced/excitation frequency. (When excitation frequency The frequency is the number of times the oscillation is repeated in one
equals natural frequency resonance occurs. Resonance should be second.
avoided in designing machine componentes) The simplest form of periodic motion is harmonic motion expressed by
• Damping: Damping Force opposes the oscillatory motion. When x=Asin2(t/  )
damping force is small, system vibrates at its natural frequency. When
Where A is the amplitude, is the period. Motion is repeated when t= 
damping force is large, vibration stops.

Harmonic motion as a projection of a point moving in a circle

x  A cos t  A sin(t   / 2)
x  2 A sin t  2 A sin(t  )   2 x (2)
It is clear from the equations above that the velocity and acceleration
leads the displacement by /2 and  respectively.

x  A sin t (1)
2
Where   2f

Now the velocity and acceleration of harmonic motion can be
determined by differenting Eq (1).
x  A cos t  A sin(t   / 2)
x  2 A sin t  2 A sin(t  )   2 x (2)

Free vibration: This is the vibration that takes place in absence of

the external force and the system vibrates at its natural frequency.
mx   kx
k
Defining the frequency as  n2 
m
1-deg of
We get x   n2 x  0
 (1)
freedom,
moves Equation (1) is homogeneous second-order linear differential
only along equation, and has the following general solution.
vertical
direction x  A sin  n t  B cos n t
At static equilibrium k=w=mg A and B are constants evaluated from initial conditions
If “x” is the displacement from the static equilibrium, we can write: x ( 0) and x( 0)
mx   F  w  k  kx Hence x
x( 0)
sin  n t  x ( 0) cos n t
mx   kx ( k  w at static equilibrium) n
 Example-2: Determine the natural frequency of the mass “M”
We know   2 on the end of a cantilever beam of negligible mass. The
m deflection of the cantilever beam is given by x=Ml3/(3EI)
Hence   2
k
1 1 k x=Fl3/(3EI)= F/k
and fn  
 2 m
Thus the stiffness of the
1 w/ 1 mg /  1 g beam is given by k=3EI/l3
  
2 m 2 m 2 
Hence the natural frequency is fn=1/23EI/Ml3)
The natural frequency can be determined from static deflection
Example-3: An automobile wheel and tire are suspended by a
Example-1: A 0.25 Kg mass is suspended by a spring having a steel rod 0.5cm in dia and 2 m long. The wheel is given an
stiffness of 0.1533 N/mm. Determine its natural frequency, angular displacement and released, it makes 10 oscillations in
determine its static deflection. Soln: fn=1/2k/m)=3.94 Hz 30.2 sec. Determine the polar moment of inertia of the wheel
and tire.
w=mg=k Therefore =mg/k=16mm.

Let Jo is the
J    k 
moment of inertia
10
 n  2  2.08rad / s of the bar about
30.2 O. Taking
k  GI p / l moment about O
where I p   d 4 / 32
and G  8 * 10 9 N / m 2 M O   P1  ka a  mgc   P2  kb b  J O
therefore k  2.455 N . m / rad However Pa  mgc  P2b  0 in equilibrium position
1

Hence   ka  kb 2   J O
2

Polar moment of inertia is J  k /  n2  0.567 kg . m 2

k a 2  b 2 
or   0
Example-4: A bar pivoted about O, springs has equal stiffness JO
k a 2  b 2 
k. The bar is horizontal in equilibrium with spring forces P2 &
P2 . Determine the equation of motion & frequency. Therefore  n 
JO

Energy method: The differential equation of motion can be Example-6: Determine the differential equation of motion for
established by considering the conservation of kinetic and potential the cylinder shown below. For no slip condition, r=R
energy. i.e. total energy is constant.
d The translational velocity of the
T  U  cons tan t therefore T  U   0 center of the cylinder is ( R  r )
dt
or T1  U 1  T2  U 2 or Tmax  U max the rotational velocity is
(  )  R / r  1
Example-5: Determine the natural frequency of the system.
because r  R
1 1 
Tmax   J 2  m ( r1) 2 
2 2  max Hence   ( R / r )


1
U max  k ( r2 ) 2max energy stored in spring
2 point of contact rotates at
At an int ermediate stage , summation of two energies is Cons tan t
 1 2 1  2 1 2
 2 J  2 m ( r1 )   2 k ( r2 )  Cons tan t
 
R  r  2  1 w r  R  1  3 w R  r 2 2
2 2
1w
T
1 1  1
Differenti ating  J 2  m 2 r12  k 2 r22  0 2g 2 g 2  r   4g
2 2  2
kr22
U  w( R  r )(1  cos  )
( J  mr12 )  kr22 therefore n 
J  mr12
 2

 
R  r  2  1 w r  R  1  3 w R  r 2 2
1w 2 Rayleigh’s effective mass method: This is an energy method
T applied to multi-mass system connected by links, levers or gears. The
2g 2 g 2  r   4g
multi-masse system is first converted to an equivalent lumped
U  w( R  r )(1  cos  ) /effective mass at a specific point. The kinetic energy of the effective
mass can then be written as: T  1 / 2m eff x 2
Adding the above two equations and differentiating: 1 
Tmax   m eff x 2 
1
kinetic energy and U max  kx 2max energy stored in spring
2  max
3 w  2

 2 g  R  r    w( R  r ) sin    0
2
 At an int ermediate stage , summation of two energies is cons tan t
  1 1
m eff x 2  kx 2  Cons tan t , hence Differenti ating
1 1
2 m eff x x  2 kx x  0
For small angles sin    therefore 2 2 2 2
k
m eff x  kx  0 therefore  n 
2g m eff
   0
3 R  r  Example-7: Determine the
2g effect of the mass of the
Hence the frequency is  n 
3 R  r  spring on the natural
frequency of the system.

The velocity of a spring element located at a The deflection of the beam at any point is:
distance “y” from the fixed end is: x y / l  3x x 
3
x 1
y  y max   4   when   
 l  l    l 2
The kinetic energy of the spring element is:
2
1  y  ms
l
1 ms 2
2 0  l  l
T add   x  dy  x The maximum kinetic
2 3
1
hence the total kinetic energy is : mx 2 
1 ms 2 1 m
x  (m  s ) x 2
energy of the beam is:
2 2 3 2 3
k m
Therefore  n  an effective increase of mass by s 2
  3x  x   
m 3 3
m s l/2
 4     dx  0 . 4857 m b y 2max
1 2m b 1
3 Tmax    y max 
  
2 0
l  l l   2
Example-8: A simply supported beam of mass mb has a The effective mass is then m eff  M  0 .4857 m b
concentrated load of M at midspan. Determine the effective Now the stiffness times the deflection is load :
mass of the system at midspan and find its natural frequency k * Pl 3 /( 48 EI )  P ( where deflection  Pl 3 /( 48 EI ))
Now the stiffness is given by : k  48 EI / l 3
The deflection at midspan
due to concentrated load P is: 48 EI
Hence the natural frequency is :  n 
Pl3/48EI l 3 M  0 .4857 m b 

Virtual work method: In this method a system in equilibrium is Viscously damped free vibration: Damping force Fd  cx
given a virtual displacement. Summation of virtual work is zero.
From free body diagram
Example-9: Determine the the equation of motion is
equation of motion for the
mx  cx  kx  F ( t )
rigid beam of mass M
The solution of the
 Ml 2 
Inertia forceW     equation has two
 3  parts. For F(t)=0, we
 l l
Spring forceW   k    have a solution for free-
 2 2 For free damped vibration
 
Damper forceW   cl l
mx  cx  kx  0 ( A)
damped-vibration. For F(t)
not equal to zero we obtain
l
l2
UniformloadW   ( p0 f t dx) x  p0 f (t )  Assume a solution of the form
a particular solution
0
2 irrespective of the free-
 Ml 2  
Sum min gthevirtualto zero
l2
  cl 2  k   p0 f (t )
l2 x  e st where " s" is a cons tan t damped-vibration.
 3  4 2
 Substituting x  e st back int o Eq. ( A)
(ms  cs  k )e st  0
2
c
 t
c k The term e 2m
is an exp onential decay function of time
s2  s  0 (characterstic Eq ) 2

When   
m m c k
(overdamped ) no oscillation are possible
The roots of the characterstic Eq. are:  2m m
2

   
2

When   
c c k c k
s1,2   ( B) ( underdamped ) oscillation are possible
2m  2m m  2m m
2
The general solution of the Eq. ( A)
When   
c k
( critically damped ) oscillation are possible
x  Ae  Be s1t s2 t
(C)  2m m
2
cons tan ts " A" & " B" are evaluated from x ( 0) & x(0)  c  k  cc  2m
k
 2m n  2 km
 
Substituting Eq. ( B) int o Eq. (C ) we get  2m m m
   c   k  t  c  k
 c
2 2


c  
 2 m  m 
      t Damping ratio  
 Ae   2m m 

t
xe 2m
 Be cc
 
 

Oscillatory motion (underdamped) 

   c    n
c c

   Xe
Now
 2m
2m x  e  t Ae
n
i  1 2 n t 
 Be
i  
1 2 n t
 1    t   
n t
sin 2

 
n
In terms of  Eq.( B ) is s1, 2       1  n 2

s1, 2
n

x  e  t C1 sin  1 2  t   C cos 1    t 
n 2
2
n
or    i 1   2 ( D)
n X ,  , C1 , C2 are cons tan ts to be det er min ed from x (0) & x(0)
 x(0)   n x ( 0) 
In terms of  Eq.( A) is x  2 n x   n2 x 

1
m
F (t ) x  e  t 
n

 n 1   2
sin   
1   2  n t  x ( 0) cos   
1   2  nt 

2
For   0
s1,2
 i Hence frequency of damped oscillation  d   n 1   2
n d
s1, 2
For 0    1    i 1   2
n
s1, 2
For   1 is real number
n

Non-oscillatory motion (overdamped)  Critically damped motion 

   
 1  n t
2
    1 n t
2

x  Ae  Be For   1 the two roots S1 & S2 are equal S1&2  n

 
putting the values in the equation below
x(0)      1  n x (0) 2

A
 x (0)   x(0)
x  e nt  n
sin  1    t  x(0) cos 1    t 
2 2

2 n   1   1 2 n n
2
 n 
zero sin  1    t  tends to  1    t
 
AS 1  2 tends to 2 2

 x(0)     2  1  n x (0)
n n

x  e nt x (0)  n x(0)t  x(0)

B
Therefore

2 n  2  1

The motion is aperiodic

and exponentially
decaying function of time
 Logarithmic Decrement Example-10 A vibrating system with viscous damping has
A way of measuring the decay of free oscillations due to damping. w=4.43kg, k=5250 N/m, and c=35N.s/m. Determine the
Consider the damped vibrations x  Xe  t sin n
 
1   2  nt    logarithmic decrement and the ratio of any two successive
amplitude.
Logarithmic Decrement is defined as:
k 5250
x e  t sin 1   2  n t1  
  ln 1  ln  ( t  )
n 1
 
  n 
m

4.43
 34rad / s

x2 e n 1 d

sin 1   2  n ( t1   d )     cc  2mn  2 * 4.53 * 34  308.4N .s / m
n t1 c 35.0
e    0.068
  ln n ( t1  d )
because the values of sin es are same at  d int ervals
e cc 308.4
e   n t1

2 2 * 0.068
  ln  ln e     n  d
   0.429
n d
  n ( t 1   d )
e
2 2  1 2 1  0.0682
  n 
n 1 2 12 x1
 e  1.536
an approxim ate equation for  x2
( w hen 1   2  1) is   2 

HERMONICALLY EXCITED VIBRATION

Remembering that the velocity
From free body diagram and acceleration leads the
the equation of motion is displacement by /2 and 
mx  cx  kx  F (t )  FO sin t.......Eq ( A) respectively.

The solution of the

equation has two parts. mx  cx  kx  FO sin t.......Eq( A1 )
•For F(t)=0, we have a
solution for free-damped- The above differential equation
vibration. can be represented by the
•For F(t)=F0sint we diagram.
Assume a particular solution of the form FO
obtain a particular solution X
x  X sin(t   ) irrespective of the free-
The amplitude and phase angle
can be obtained from the
k  m   c 
2 2 2

Where X is the amplitude and  is the phase damped-vibration. c

displacement angle with the excitation force diagram.   tan 1
k  m 2

The amplitude and phase angle written in non-dimensional form show that they are
function of frequency ratio n diagram and damping factor
 
2  
X
FO
;
Xk

1
; tan  
c
  n 
k  m   c  k  m 2 2
2 2 2 FO 2
    2      2  
1      2   1   
   n 
  n     n  For n<<1 For n=1 For n>>1
•The inertia and •The phase angle  •The phase angle 
LARGE INFLUNCE damping forces are =900 . approaches 1800 .
ON AMPLITUDE small.
•The inertia force is •The impressed
AND PHASE •This results in force balances the
balanced by the
ANGLE NEAR small phase angle . spring force inertia force
RESONANCE
(n=1) •The impressed •The impressed •The amplitude is:
force is nearly equal force balances the X=F0/cn
to the spring force damping force
•The amplitude is:
X=F0/cn
 ROTATING UNBALANCE A common source of vibration excitation
Now we can rewrite the equation:
mx  cx  kx  FO sin t Let x be the displacement of the non-
rotating mass and the displacement of the
rotating mass be x+esint
Now we can write the equation of motion:
In the new form:
F d 2 ( x  e sin t )
x  2 n x   n2 x  O sin t ( M  m) x  m  kx  cx
dt 2
m Re arranged to : M x  cx  kx  (me 2 ) sin t
Comparing with Eq(A1) the solution is:
2
   
   2  
And write the complete solution as:
X
me 2
;
MX
  n ; tan  
c
  n 
k  M   c  k  M 2 2
2 2 me 2
 
 
    2      2
2

FO sin t    1      2   1   

 n 
x(t )   X 1e  nt sin 1   2  nt  1 
  n     n 

k 2
   2     2
1      2 
   n     n 
The complete
solution can be x(t ) 
me 2 sin t   

 X 1e  nt sin 1   2  nt  1 
written as: k  M   c 2 2 2

WHIRLING OF ROTATING SHAFT Defined as the rotation of the plane    ;   r  r  0 ;   t  
made by a bent shaft and the line of
centers of bearings k 2 c
   r  e cos  ; r  e 2 sin 
2

m  m
The shaft i.e. the line e=SG is rotating

at a constant speed  and the line c
 2
m n
r=OS whirling at speed  that may or Hence tan    2
k
not be equal to    2 1    
m  
 n
The acceleration of the mass centre is:
aG  aS  aG / S aG/S is only radial From the vector triangle above:
acceleration k
(directed from G  2

      cos   m
aG  r  r 2  e 2 cost    i  r  2r  e 2 sin t    j
to S) as the shaft 2 2
k 2 c 
is rotating at       
constant speed m  m 
The equation of motion resolved in radial and tangential direction:
Putting in Eq. above
  
 kr  cr  m r  r 2   e 2 cost    ;  cr  m r  2r   e 2 sin t      
2

e 
For synchronous whirl:    ;   r  r  0 ;   t   me 2  n 
r 
k 2 c k  m   c 
2 2 2 2
    2      2
   r  e cos  ; r  e 2 sin  1      2  
2
 
m  m   n     n 

SUPPORT MOTION In many cases the dynamical system is excited by the motion of the support.
 k  ic  it
Let “y” be the motion of the support & “x” be the displacement of the mass m
The absolute displacement is: x Ye
 k  m  ic 
2
The equation of motion
mx   k ( x  y )  c( x  y )
let z  x  y then The steady state amplitude and
X

k 2  (c) 2
; tan  
mc 3
Y k  m   c 
2 2 2
 
k k  m 2  c 
2 2

mz  cz  kz   my  m 2Y sin t phase from this equation are:

where y  Y sin t
The above equation is comparable to the equation below:

Mx  cx  kx  (me 2 ) sin t The curves show that the ratio
X/Y=1, at the frequency  /  n  2
for all values of damping. Thus
Thus the solution is: it is obvious that the natural
frequency of the supported
system should be small
z  Z sin(t   ) compared to the disturbing
frequency. This can be
m 2Y c achieved by using soft
Z ; tan   springs(VIBRATION
k  m   c 
2 2 2 k  m 2 ISOLATION)
 VIBRATION ISOLATION Transmissibility (TR) of the disturbing force is defined as:

2
 2 
1   
TR 
FT
  n  
X
F0 1   /     2 /  
n
2 2
n
Y

With negligible damping the above equation reduces to:

The force to be isolated is transmitted through the spring & damper
1 1
shown in the figure above. Its equation can be given as: TR  
 /  n 2  1 2f 2  / g  1
2
 2  The amplitude “X” (see Eq for “X”) of the isolated mass “m” is reduced by placing it on a large mass “M”.
FT  (kX ) 2  (cX ) 2  kX 1    However, the stiffness “k” then must be increased to keep the ratio k/(m+M)=constant, so that the natural
 n  frequency and hence the TR is unchanged. THE AMPLITUDE IS HOWEVER REDUCED BECAUSE THE
SPRING CONSTANT “k” APPEARS IN THE DENOMINATOR OF EQUATION FOR “X” (see Eq for “X”)
with disturbing force F0 sin t
F0 / k
X
1   /     2 /  
n
2 2
n
2

ENERGY DISSIPATION BY DAMPING STRUCTURAL DAMPING

Structural Damping is proportional square of amplitude
•Damping is present in all oscillatory system.
Wd  X 2
•It removes energy which is dissipated into heat or is radiated away
 ceqX 2  X 2 ; ceq (equivalent damping )   /  
•In all cases force-displacement curve will enclose an area called
Differential equation for a system with structural damping
“Hystereses loop”

The energy lost per cycle can be given as: mx  x  kx  FO sin t
Wd   Fd dx 
where Fd  cx and x  X sin(t   ); x  X cos(t   ) VIBRATION MEASURING INSTRUMENT
2

Wd   cxdx   cx 2 dt  c 2 X 2  cos 2 (t   )dt   cX 2
0 mx  k ( x  y )  c( x  y )
The energy lost per cycle can be represented graphically:
let z  x  y then
x  X cos(t   )  X 1  sin 2 (t   )   X 2  x 2
Hence Fd  cx   c X 2  x 2
Displacement, velocity or mz  cz  kz  m 2Y sin t
acceleration is indicated by
2
 Fd   x 
2 the relative motion of the where y  Y sin t
    1 suspended mass
 cX   X 

mz  cz  kz  m 2Y sin t ( from last slide)

The above equation is comparable to the equation below: Response of vibration
Mx  cx  kx  (me 2 ) sin t measuring instrument

Thus the solution is:

The type of instrument is
z  Z sin(t   ) determined by the useful range
2 of frequencies with respect to
  the natural frequency n of the
Y  
m 2Y  n 
instrument
Z 
k  m   c 
2 2 2 2
   2     2
1      2
  n 
  n   

2 Seismometer: When the natural frequency of the instrument is low,
c n resulting in large /n , The relative displacement Z approaches Y
tan    The relative motion
k  m 2  
2
regardless of the damping factor . Hence, the mass remains Z is converted to
1    stationary and support case moves with the vibrating body. (see
electric voltage by

 n 
making the seismic
fig.). Useful in frequency range of 10 to 2000 Hz. Both mass a magnet
displacement and acceleration are available from the velocity-type relative to coil fixed
in the case.
transducer, by means of the signal integrator or differentiator.
Accelerometer: When the natural frequency of the instrument is high, resulting in
low /n , The relative displacement Z becomes proportional to acceleration (see Accelerometer: The useful range is limited, can be seen below. For =0.7 the useful
below) frequency range (/n) is 0 to 0.2

2
 
Y  
 n 
2
  acceleration
Z  Y   
 n   n2
2
   2   
2

1      2 
   n    n 
 1
(as  0.0 or  1.0)
n 2
   2     2
1      2 
   n     n 

Im pulse Excitation
TRANSIENT VIBRATION 
Fˆ    (t   )dt  1.0 0    
Dynamical system excited by a suddenly 0
applied non-periodic force results in 

transient response. System vibrates at natural  f (t ) (t   )dt  f ( ) 0  

frequency with amplitude varying according Because Fdt  mdv 0

to type of excitation. Fˆ  mdv •This change in velocity will happen in a

very short period of time
Fˆ
 dv  change in velocity •Also the time of application of the force is
m very small. Hence the displacement will also
Im pulse Excitation
 velocity if starting from rest be very small (x(0)=0)

Fˆ  F (t )dt  Initial velocity  x (0)
Under free un-damped condition a spring mass system will behave as (we know):
When Fˆ  1.0 and   zero x ( 0 )
x sin  n t  x (0 ) cos  n t
Fˆ n
 unit impulse or delta function A spring mass system at rest under impulse will behave as:

Fˆ
delta function   (t   )  0 ( for t   ) x sin  n t  Fˆ h (t )
m n
 very l arg e value ( for t   ) 1
where h (t )  sin  n t  response to unit impulse
m n

ARBITARY EXCITATION
When damping is present a spring mass system will behave as:

x ( 0 ) e   n t
x sin 1   2  n t
n 1   2 Fˆ  f ( )   f ( )  h (t   )
Contribution to the response at time “t”
x ( 0 )  Fˆ / m The strength of the impulse
Where h(t-) is the response to an unit
Fˆ e  n t impulse started at t= . By combining
x sin 1    n t  Fˆ h (t )
2

m n 1   2 all such contributions, the response to

the arbitrary excitation f(t) is
where h (t )  response to unit impulse represented by the integral:
t
x (t )   f ( ) h (t   ) d 
0

This integral is called convolution integral or superposition integral

 Example4.2.1
Fo
x (t )  Xe  n t sin( 1   2  n t   ) 
The response to the step excitation shown m  n2
For x ( 0 )  x ( 0 )  0
F  e  n t 
x  o 1  cos( 1   2  n t   ) 
Considering undamped system: k  1 2 
t
x (t )   f ( ) h (t   ) d  
tan  
0 1 2

1
where h (t   )  sin  n (t   ); f ( )  Fo
m n
t Figure below shows that the peak response is less than 2xk/F0 when damping is present
Fo F
m  n 0
x (t )  sin  n (t   ) d   o (1  cos  n t )
k

For a damped system we can repeat the procedure with:

e  n ( t  )
h (t   )  sin 1   2  n (t   )
m n 1   2

LAPLACE TRANSFORM FORMULATION FOR A DAMPED SPRING

MASS SYSTEM WITH INITIAL CONDITIONS

mx  cx  kx  F (t ) u sin g appendix  B :

mx  kx  mg
 
m s x( s )  sx(0)  x (0)  csx( s)  x(0)  kx( s )  F ( s )
2

Laplace transform
F ( s) (ms  c) x(0)  mx (0)
x(s)  2 
ms  cs  k ms 2  cs  k 
m s 2 x( s )  sx(0)  x (0)  kx( s)   mg
s
The solution x(t) is found from the inverse of the above equation. This is explained by
the following example with x (0)  2 gh and x(0)  0
Example 4.3.2
Consider the spring mass system shown, dropped through a height of “h”. If “x” is  mg
m s 2 x( s )  2 gh  kx( s) 
s

measured from the position of m at the instant t=0 when spring contacts the floor.
The differential equation of motion as long as spring is in contact with the floor is: k g
s x( s )  2 gh  x( s ) 
2

m s
mx  kx  mg g
x( s )( s   n )   2 gh
2 2

g
x( s )( s 2  n2 )   2 gh (repeated from last slide) x(t ) 
2 gh
sin n t 
g
(1  cos nt ) (repeated from last slide)
s n n2
2 gh g 2 gh  g 
2

x( s )   x(t ) 
g
   sin(nt   )  2 for x(t )  0
( s  n2 ) s( s 2  n2 )
2
 n   n2  n
2
Inverse Transform (Table B.1) 2 gh  g 
x (t )  n    cos(n t   )
1 1  n  n2 
 sin at
(s  a )
2 2
a 2 gh  g 
2

x(t )   n2    sin(n t   )
1 1  n  n2 
 2 (1  cos at )
s( s 2  a 2 ) a Re cognizing
g
Hence the inverse transform of x( s) is :   st & max imum accel & displacement occurs at sin(n t   )  1
n2
2 gh g x 2h
x(t )  sin n t  (1  cos n t ) 
 st
1
n n2 g
 PULSE EXCITATION OF UNDAMPED SPRING MASS SYSTEM TO THREE By superimposing the two equations: the response for t>t1
DIFFERENT EXCITATIONS FOR PERIODS t>t1 and t<t1
Fo  sin  n t sin  n (t  t1 ) 
x (t )  1    t  t1
k   n t1  n t1 
Constant with rise time
Constant with rise time Half sine pulse
Constant with rise time Rectangular pulse

The input can be considered to be the sum of two ramp functions: the first ramp function: Rectangular pulse
t 
f (t )  Fo   The input signal is considered as sum of two step functions shown
 t1  Response already known for the 1st one from Eq 4.2.2 given below
1  kx (t )
h (t )  sin  n t  n sin  n t ;  (1  cos  n t ) t  t1
m n k Fo
Rectangular pulse

n t
 2
x (t ) 
k F
0
o
t1
sin  n (t   ) d Constant with rise time Peak value  2 at  nt 

t 

F t sin  n t kx (t )
  (1  cos  n (t  t1 )) t  t1
 o(  ) t  t1 Similarly, response to the second step
k t1  n t1 Fo
For the second ramp function, inspecting the first one above:
 (1  cos  n (t ))  (1  cos  n (t  t1 )) 
kx (t )
Fo t  t1 sin  n (t  t1 ) Combining: Fo
x (t )   (  )
k t1  n t1   cos  n t  cos  n (t  t1 ) t  t1

t SHOCK RESPONSE CURVE/SPECTRUM

Half-sine pulse
F (t )  Fo sin for t  t1
t1 When the duration of the pulse of impulse force is short it is called shock
0 for t  t1 SHOCK RESPONSE SPECTRUM (SRS): It is a plot of maximum peak response of the
single degrees of freedom (SDF) oscillator as a function of natural period of the oscillator.
Differenti al Equation
The maximum of the peaks is called maximax.
F
x   n2 x  o sin t / t1 for t  t1 For an arbitrary excitation f (t )
m t

General solution max i max  x (t ) max  

0
f ( ) h ( t   ) d 

sin  t / t1
max
F
x  A sin  n t  B cos  n t  o 2 for t  t1 h (t   ) 
1
sin  n (t   )
m  n  ( / t1 ) 2 where
m n
A and B obtained from x ( 0 )  x (0 )  0
 Let us select the example of rectangular pulse
where the response is given as follows:Eq4.4.6
For t>t1 excitation force is zero so the solution is given as:
 (1  cos  n (t ))  (1  cos  n (t  t1 )) 
kx (t )
x (t1 ) Fo
x (t )  sin  n t  x (t1 ) cos  n t for t  t1
n   cos  n t  cos  n (t  t1 ) t  t1

x (t1 ) and x (t1 ) obtained from Equation for t  t1

Now, let us plot this equation as a function of time for different duration/length of shock t1

Plot the response equation as a function of time for different duration/length of shock t1 •The peak value “maximax” depends on the duration of the pulse/shock.
•It is laborious to find all “maximax” for all the possible values of shock duration t1
The dark or the dotted curve is the difference of two the functions shown in Equation4.4.6 below
• Hence we differentiate the following equation with respect to time to obtain the value
 (1  cos  n (t ))  (1  cos  n (t  t1 )) 
kx (t ) of the time for which response will be maximum
Fo  kx (t ) 
  cos  n t  cos  n (t  t1 ) t  t1
d
 
   n sin  n t p  sin  n (t p  t1 )  0 
dt  Fo 
•The peak response is (xk/Fo)max=0.8 where t p  time for peak response
sin  n t1
•happens at tm=0.32 tan  n t p 
 (1  cos  n t1 )
•when duration of shock is t1=0.125 From the graph
•This is one point on the SRS curve sin  n t1
sin  n t p 
2(1  cos  n t1 )
 (1  cos  n t1 ) 1
•The peak response is (xk/Fo)max=1.82 cos  n t p   (1  cos  n t1 )
2 (1  cos  n t1 ) 2
•happens at tm=0.45
Putting in original Eq 4 .4 .6 we get
•when duration of shock is t1=0.4 Eqation for peak response
•This is second point on the SRS curve  kx (t )  t
 F   2 (1  cos  n t1 )  2 sin 1

for t  t1
 o  max
  kx (t )  t Shock Response Spectrum (SRS) for the half sin pulse
   2 (1  cos  n t1 )  2 sin 1 for t  t1
 Fo  max 

Equation for peak response repeated from last slide and plotted below is the
Shock Response Spectrum (SRS) for rectangular pulse

Shock Response Spectrum (SRS) for triangular pulse

•The peak response is (xk/Fo)max=2.0

•when duration of shock is t1=0.5
•Note this is also the peak value for t<t1

SHOCK ISOLATION FINITE DIFFERENCE NUMERICAL COMPUTATION

For shock isolation, the maximum peak response or the transmissibility must be less than •When the differential equations can not be integrated in closed form
unity
t1 •When the system is non-linear or can not be expressed by simple analytical functions
2 sin  1 .0
 x  f ( x , t )
Let us represent an undamped system with
t1  known initial displacement and velocity as: x1  x ( 0 )
Thus for rectangular pulse the  30  0

 6 x1  x ( 0 )
shock isolation requires:
t1 1 h2 h3
 or 6t1   xi 1  xi  hx i  xi  xi  ....... (Taylor series )
The natural period of isolated  6 2 6
system must be greater than six
 h2 h3
times the pulse duration 6t1< n  xi 1  xi  h x i  xi  xi  ....... (Taylor series )
3t1 2 6
where h  t
For shock isolation from arbitrary pulses shown below:
Substracti ng Taylor series and ignoring higher order terms
It is safe to assume a rectangular pulse enclosing the shape of the pulse. 1
x i  ( xi 1  xi 1 )
2h
Adding Taylor series and ignoring higher order terms
1
xi  2 ( xi 1  2 xi  xi 1 )
h
For small t1/ the peak response occurs in the region t>t1 and approaches that of Re arranging above gives the diff Equn xi 1  2 xi  xi 1  h 2 f ( x1 , t1 )
a system excited by an impulse.

Problem with ZERO initial conditions:

Re peated from last slide xi 1  2 xi  xi 1  h 2 f ( xi , t i ) for i2
To start from x3 (equation below) we need to know x2 which we get from the 1st
To start from x3 (equation above) we need to know x2 which we get from the 1st
Taylor series equation given in last slide and below. BUT WITH ZERO INITIAL
Taylor series equation given in last slide and below CONDITIONS x2=0 and hence the solution for x3 can not be started
h2 h3 h2 h3
xi 1  xi  hx i  xi  xi  ....... (Taylor series ) xi 1  xi  h x i  xi  xi  ....... (Taylor series )
2 6 2 6
h2 h2
x 2  x1  h x1  f ( x1 , t ) x 2  x1  h x1  f ( x1 , t )  0
2 2

Re peated from last slide xi 1  2 xi  xi 1  h 2 f ( x1 , t1 ) for i2

x3  2 x 2  x1  h 2 f ( x 2 , t 2 ) for i2
Problem with ZERO initial conditions can be avoided by assuming linear variation of acceleration between
acceleration at t=0 and t=t1 as shown below
assu min g linear var iation between x1 and x2
x2  x1   t  0    t   h
x2

h
 x2 2
int egrating x  h2  h  x2 h / 2
2 2h
x
int egrating again x  2 h 3  x2 h 2 / 6
6h
 EXAMPLE 4.7.2 For damped system equation contains additional term x i
Solve the following differential equation of a spring mass system with triangular pulse xi 1  2 xi  xi 1  h 2 f ( x1 , x1 , t1 ) for i2

0 .5 x  8 2 x  F (t ) To start from x3 (equation above) we need to know x2 which we get from the 1st
Taylor series equation given in last slide and below
x  f ( x , t )  2 F (t )  16 2 x
initial conditions x1  x1  0 h2 h3
xi 1  xi  h x i  xi  xi  .......
22 2 6
From Equation above     0 .5
 4 (Taylor series )
h2
x 2  x1  h x1  f ( x1 , x1 , t1 )
2

x 2 obtained from 2 nd
h2 h3
Taylor series xi 1  xi  h x i  xi  xi  .......
2 6
for with i2
h2
x1  x 2  h x 2  f ( x 2 , x 2 , t 2 )
2

SYSTEMS WITH TWO OR MORE DEGREES OF FREEDOM

For “N” degrees of freedom there Matrix form of the Equations

will be “N” natural frequencies
corresponding to what is known 2 k   m  k A  0 2
1
as “N” natural modes of
 k 2 k  2  m  A 2
2
vibrations.
Example-11. Find the natural 2 k   m * 2 k  2  m   (  k ) * (  k )  0
2 2

modes of vibrations of the

following system
4k 2
 4 k  2 m  2 k  2 m  2 4 m 2
 k 2
 0
m x1   kx 1  k  x 2  x 1  2 m 4 2
 6 k m  3k 2 2
 0
2 m x 2   k ( x 2  x 1 )  kx (I )
3 km 
2 2 2
x 1  A 1 sin  t 3k
Let
 4

  0
x 2  A 2 sin  t m 2
2m 2
Substitute in above Equations
2 k A 3k 3k 2
   kA  0 2    0 ( III )
2
m 1 2

 kA 1  2 k  2 2
m A 2  0 ( II ) m 2m 2

( fo r 1 )
Solutions of above quadratic Equation ( III )  A1  k
    0 .7 3 1
3
1   
1  k
3  0 . 634
k  A2  2 k   12 m
2 2  m m ( fo r  2 )
 A1  k
3 1  k k      2 .7 3
2    3  2 . 366  A2  2 k   22 m
 2 2  m m
The normalized amplitude ratios also known as modes are:
k
Therefore 1  1  0 . 634  0 .7 3 1    2 .7 3 
m  1x     and  2 x    
k  1 .0 0   1 .0 0 
& 2  2  2 . 366
m
From equation ( II )
A1 k 2 k  2 2 m
 
A2 2k   m
2
k
The normal modes of oscillations can now be written as: Example-1.3 The weak spring k, is unstrained in vertical
1  position. Determine the normal mode vibrations.
 x1   0 .7 3 1 
   A1   s i n  1 t   1  ml 2  1   m lg  1  ka 2  1   2 
 2
x  1 .0 0  ml 2  2   m lg  2  ka 2  1   2 
2  
1 0   1  ka 2  mgl   ka 1  0 
  2 .7 3 
2

 x1  ml 2  
      
   A2   s i n  2 t   2  0 1   2   ka 2  ka 2  mgl 2  0 
 x2   1.0 0  Assuming solutions  1  A 1 cos  t and 2  A 2 cos  t
The equations for the rotating system following Eq. (I)
2
g g ka
The natural 1  and 2   2
J 11   k 1 1  k 2  2   1  l l ml 2

frequencies and  A1 
1
 A1 
1
J 2 2   k 2 (  2   1 )  k 3 2 ( I V )    1 . 0 and     1 . 0
shapes are A2  A2 
J 1 0  1 
  
0 J 2  2 
The equations for the rotating system following Eq. (I)
k 1  k 2   k 2  1   0 
    
 k 2 k 2  k 3   2  0

Coordinate Coupling: In most general cases the The following matrix equation shows zero dynamic and
differential equations are coupled. For undamped cases: static coupling but the coordinates are coupled by the
m 11 x 1  m 12 x 2  k 11 x 1  k 12 x 2  0 damping matrix.
m 21 x 1  m 22 x 2  k 21 x 1  k 22 x 2  0 m 11 0  x 1  c 11 c 12  x 1  k 11 0  x 1   0 
        
Matrix form of the Equations 0 m 22  x 2  c 21 
c 22  x 2  0 k 22  x 2   0 
m 11 m 12  x 1  k 11 k 12  x 1   0  System will be uncoupled if c 12  c 21  0
      
m 21 m 22  x 2  k 21 k 22  x 2   0  Example-11 The system shown
represents a 2 DOF because two
Dynamic coupling exists if mass matrix is non-diagonal coordinates are necessary to
Static coupling exists if stiffness matrix is non-diagonal describe its motion.
It is always possible to decouple the equations for un- Coordinates leading to static
damped system by choosing proper coordinate system. coupling
This is not always the case for damped system

m 0  x  ( k 1  k 2 ) ( k 2 l 2  k 1 l1 )  x   0  Coordinates leading to static coupling & Dynamic coupling

      
0 J    ( k 2 l 2  k 1 l 1 ) ( k 1 l 12  k 2 l 22 )     0 
For k 1 l 1  k 2 l 2
coupling disappears , giving un  coupled system
Coordinates If we choose x=x1 at the end of the bar, we get following
leading to static m ml 1  x 1  ( k 1  k 2 ) k 2l  x 1  0 
coupling       
ml 1 J 1    k 2l k 2l2   0 
There is some point “C” along the bar where a force applied Example-12 Determine normal modes of vibration for 2DOF
normal to the bar produces pure translation, i.e. k1l3  k 2 l 4 m=1463 kg, l1=1.37m, k1=35000 N/m, l2=1.68m, k2=37900 N/m,
The equations of motion are given below r=1.2m, l=3.05m, Jc= mr2
m x  k 1 ( x  l 1  )  k 2 x  l 2    0
m me  x c  ( k 1  k 2 ) 0  x c  0
       J c   k 1 ( x  l 1  ) l 1  k 2 x  l 2  l 2  0
me J c    0 ( k 1 l 32  k 2 l 24 )   0
Static coupling disappeared and dynamic coupling introduced
 Assuming harmonic motion we get
k  k 2   2m   k 1 l 1  k 2 l 2  x  0
    
1

 k 1 l 1  k 2 l 2  k 
l  k 2 l 22   2 J c     0 
2
1 1

From the determinant of the matrix, the two natural frequencies and
the amplitude ratios for two frequencies are
 1  6 . 9 rad / s  1 . 1 cps
 2  9 . 06 rad / s  1 . 44 cps
 x 
    4 . 45 m / rad
   1
 x 
   0 . 33 m / rad
  2

The two modes are shown in the figs.

One is translational with little
rotation, the other is pure rotational.