STiCM

Select / Special Topics in Classical Mechanics

P. C. Deshmukh

Department of Physics
Indian Institute of Technology Madras
Chennai 600036

pcd@physics.iitm.ac.in

STiCM Lecture 07: Unit 2 Oscillators, Resonances, Waves

PCD_STiCM 1
 Oscillation:
Repetitive
Physical
phenomenon
Dynamics of
spring–mass systems,
 pendulum,
 oscillatory electromagnetic circuits, Ganga
 bio rhythms,
 radiation oscillators,
 share market fluctuations molecular vibrations,
atomic, molecular, solid
state, ……..
PCD_STiCM 2
 Oscillations.
Small oscillations.
SHM.
Unit 2: Driven and
Oscillators damped
Resonances
Yamuna oscillator.
Waves
Resonance,
Quality factor.
Waves.

Flip-Flop Square
Wave oscillator
PCD_STiCM 3
 Learning Goals

• Recognize stable, unstable, neutral

equilibrium points and saddle points.

• Learn that in a region close enough to any

point of stable equilibrium, motion can be
described by the simple harmonic oscillator.

• Discover electro-mechanical analogies

and how they can be exploited in solving
problems in different branches of Physics.
Learn about effects of damping, and effects
of a periodic driving force.
PCD_STiCM 4
 Learning Goals

• Get introduced to resonances in

physical systems and the primary
indicators of the quality of
measurement techniques, such as
the ‘Quality Factor’.

• We shall also become familiar

with the ‘wave motion’ which is of
ubiquitous application in both
‘classical’ and ‘quantum
mechanics’.
PCD_STiCM 5
Galileo Newton

( q, q )
F  ma
Linear Response.
Principle PCD_STiCM
of causality. 6
 ( q, q )
F  ma
Linear Response.
Principle of causality.

Principle of
Galileo Newton
Variation
L ( q, q )
H ( q, p )
L d  L 
  0
q dt  q 
H H
q ,p
Lagrange
Hamilton
PCD_STiCM p qk 7
Galileo

( q, q )
Newton F  ma
Linear Response.
Principle of causality.
PCD_STiCM 8
 Kinds of equilibrium

unstable
stable
unstable

neutral
stable

stable
PCD_STiCM 9
 U
Un/stable STABLE N
equilibrium? S
T
+ +
A
B
L
Saddle point E

PCD_STiCM 10
meaning of small oscillations U(x)
‘Zero’, at equilibrium
x
U
U ( x)  U ( x0 )  ( x  x0 ) 
x x0

1  2U U(x)
+ ( x  x0 ) +
2

2! x 2 x0

1  3U
+ ( x  x0 )3  ...
3! x3 x0
x

Potential for a
Approximations, close to x0 Linear harmonic oscillator
dU
F   kx
1  2U 1 2 dx
U ( x)  U ( x0 )+ ( x  x0 ) = kx
2

2! x x
2
2 k
0 x x
by choosing U ( x0 )  0 and x0  0. m
PCD_STiCM 11
‘reference circle’
for
Simple Harmonic Oscillations

Intrinsic natural frequency

  2
Shadow of the red dot in uniform
PCD_STiCM 12
circular motion constitutes SHM
 Unlike what happens in
a resistor, the current
L C and voltage in an
inductance L, and in a
capacitor C, does not
dV
I is proportional to , peak together.
dt
not to V, as in the case of a resistor.
VL Voltage lags the current in a
capacitor by 900,
I
but
VC
leads the current in an inductor by
PCD_STiCM 13
the same amount.
 Q
VC 
C
dI d 2Q
VL   L   L 2   LQ
dt dt
d d dV
I  Q  Q  (CV )  C
dt dt dt
dV
I is proportional to ,
dt
not to V, as in the case of a resistor.
VL
I Voltage lags the current in a capacitor by 900,
VC but leads the current in an inductor by the same amount.
PCD_STiCM 14
PCD_STiCM 15
 k
x x
m Electro-mechanical analogues:
1 1
Q   Q Inductance mass, inertia
LC Capacitance 1/k, compliance
(1) q   q
(2) Most general solution: q  Aei0t  Bei0t
Substitute (2) in (1)  0  

A & B: determined by
INITIAL CONDITIONS

Question:

Could we have
associated L with
PCD_STiCM 1/k and C with m? 16
Any wonder that Feynman calls the above
relation as Newton’s law of electricity’ ?
Two initial conditions provide solutions to the
‘equation of motion’ in a linear response
formalism. PCD_STiCM 17
 q   q (t )
The most general solution is
q (t )  Aei0t +Be  i0t where 0 = 
the frequency is governed by  ;
A and B are determined by initial conditions
the solution at time t  0 is
q(t  0)  A  B; also, q(t  0)  i0 ( A  B)

solving for A an B from the two equations,

1  q(t  0) 
A q(t  0)  i  ;
2  0 
B  A* (complex conjugate)

PCD_STiCM 18
 k Mean kinetic energy
x x spring-mass system 1
m KE = mx 2
x  A sin(0t   ) 2

k T

0   f (t )dt
Mean potential energy
m 1 2
Mean: f (t ) = t 0
T PE  kx
2

t 0
dt

1
<cos ( t)dt>   <sin 2 ( t)dt>
2

2
1 1 1 1 2
KE = m x 2  m( A0 )   A m0 2
2

2 2 2 4
1 2 1 2 1 1 2
PE  kx  kA   kA
2 2 2 4
k
Note : PE  KE , since 0 
PCD_STiCM
2
19
m
  1 
Q   Q
 LC 
(1) q   q
(2) Most general solution: q  Aei0t  Be i0t
Substitute (2) in (1)  0  

Graph plotting exercises

a) plot q and q as functions of t
b) sketch instantaneous V and I as functions of t
c) what is the phase difference between q and q ?
d) what is the phase difference between I and V ?
PCD_STiCM 20
(1) q   q

i0t  i0t
(2) Most general solution: q  Ae  Be

0  

SUPERPOSITION

Coupled oscillators
PCD_STiCM 21
 Longitudinal oscillations Reference: Berkeley’s Mechanics

1 2 Longitudinal Oscillations

Frictionless support

Principle of superposition Coupled Oscillators

Frequency of oscillations?

1 Longitudinal Displacement, to the

left or right, both make BOTH
THE SPRINGS apply a restoring
2 force on the mass in essentially
THE SAME DIRECTION.

Frictionless support ‘effective spring constant’ = ?

PCD_STiCM 22
ao: relaxed length of the springs Tension exerted by each string
a: instantaneous stretched length AT EQUILIBRIUM
T  k (a  a0 )

Transverse oscillations
l
x
θ
a a
View in the plane of vibration
The total restoring force along  x
is  2T sin 
x
mx  2T sin   2k (l  a0 )
l

PCD_STiCM 23
meaning of small oscillations U(x)

x
U
U ( x)  U ( x0 )  ( x  x0 ) 
x x0

1  2U U(x)
+ ( x  x0 ) +
2

2! x 2 x0

1  3U
+ ( x  x0 )3  ...
3! x3 x0
x

Potential for a
Approximations, close to x0 Linear harmonic oscillator
dU
F   kx
1  2U 1 2 dx
U ( x)  U ( x0 )+ ( x  x0 ) = kx
2

2! x x
2
2 k
0 x x
by choosing U ( x0 )  0 and x0  0. m
PCD_STiCM 24
 ao: relaxed length of the springs
a: instantaneous stretched length
T0=k(a-a0)

T=k(l-a0) Ref.: Berkeley, Vol.1/Mechanics

l
x
θ
x mx  2T sin   2k (l  a0 )
l
a a
View in the plane of vibration
SLINKY approximation
(l  a0 ) k
if a0  l i.e.
a0
 1 ;  1; x  2 x
l l m
SLINKY ~ SHO with effective spring constant (2k),
- for very large values of l without losing linear elasticity!
A typical slinky with a0 of only 3” can be stretched to
as much as ~15’ without loosing
PCD_STiCM the linear elasticity! 25
(1) q   q Displacement q(t  0)  A  B
(2) Gen. solution: q  Aei0t  Bei0t q(t  0)  i0 ( A  B)
Substitute (2) in (1)  0   We can find A and B in terms of
q(t  0) and q(t  0)

Displacement q  q( x, t )  q0 cos t
However, q0 cos  t   ( x) 
is also a solution
What is the functional form of  ( x)?

over one wavelength,  must change through 2

 2 2
 and  = x    kx  
x  
where  is some constant angle.
PCD_STiCM 26
 2
q(t )  q0 cos t   ( x)  = x    kx  

q(t )  q0 cos t  (kx  ) = q 0cos{ t  kx  }

phase :   t  kx  

On a surface of constant phase : d  0 i.e.  dt  kdx  0

i.e. kdx   dt
 + dx 
when
0 
dt k
dx
0 when
dt
 a wave travelling to the left dx
0
dt
 a wave travelling to the right
PCD_STiCM 27
f(x-vt) represents a pulse traveling to the right
dx dx
 0, i.e. as a positive quantity
dt dt
 a wave travelling to the right

g(x+vt) represents a pulse traveling to the left

dx dx
 0, i.e. as a negative quantity
dt dt
 a wave travelling to the left

The wave
wavelength
covers one
 in one period T,

The traveling speed of the wave is v=  
PCD_STiCM T 28


PCD_STiCM 29
We will take a break….

…… Any questions ?
pcd@physics.iitm.ac.in

Next: DAMPED oscillations

PCD_STiCM 30
STiCM

Select / Special Topics in Classical Mechanics

P. C. Deshmukh

Department of Physics
Indian Institute of Technology Madras
Chennai 600036

pcd@physics.iitm.ac.in

STiCM Lecture 08: Unit 2 Oscillators, Resonances, Waves

PCD_STiCM 31
Total energy E is constant: conservative forces
<KE> = <PE> : not true when friction is present

Damped harmonic oscillator

Is there only a restoring force in real situations?

Energy dissipation
Breaking, damping in automobiles,
galvanometer

PCD_STiCM 32
 S.H.O. mx  kx where 0 2k
m
Damped Ffriction  cv  cx
Oscillator:

mx  kx  cx c

 x  2 x  0 x  0
2
2m
If EM & Gravitational forces are conservative,
and all forces are made up of fundamental forces,

Then, why is friction dissipative?

Just what is ‘lost’, PCD_STiCM
and why? 33
All ‘net’ interactions in nature:

superpositions of fundamental interactions,

- nuclear (‘strong’ interaction),
- electro-weak
(electromagnetic/nuclear ‘weak’),
-and gravity.

So, what is the origin of dissipation?

PCD_STiCM 34
Cause of ‘Friction’: Often, we track the evolution of
the state of some pre-specified mechanical system
without keeping track of everything else that this
system interacts with.

There are thus unspecified degrees of freedom !

Dissipation: result of our neglect of these

unspecified degrees of freedom,
even as the component interactions individually
conserve energy.

PCD_STiCM 35
 The equation of motion: mx  kx  bx
 x  2 x  0 x  0 2 Eq.[a]

k c
where 0  2

m 2m
We seek a solution in the form: x (t )  Ae qt
Eq.[b]

Why seek this and inquire what conditions would result on q if

form? Eq.[b] is to be admitted as a solution of Eq.[a]

Substitute [b] in [a]: x(t )  Aqeqt , x(t )  Aq 2eqt

Aq 2eqt  2 Aqeqt  0 2 Aeqt  0
q  2 q  0  0 PCD_STiCM
2 2
36
 c
q + q +   0
2 2
m0 2  k
m quadratic equation
mq 2 + cq  k = 0.
c
c  c 2  4mk

q= . 2m
2m

q1 =   +    , q2 =      
2 2 2 2

x(t )  A1e q1 t
 A2e ;
q2 t

A1 and A2 are constants

determined by initial conditions,
at t  0, on x(t ), x(t )
PCD_STiCM 37
 q1 =   +  2   2 , k c
0  2

m 2m
q2 =     2   2

CASE 1:

When   0 ,
OVERDAMPED OSCILLATOR
 2   2 is a real number whose value/magnitude is   ,
so both q1 and q 2 become 'real' and essentially 'negative'

Since : x(t )  A1e q1 t

 A2e ;
q2 t

both the terms approach zero as t  , asymptotically

PCD_STiCM 38
 x(t )  A1eq1 t  A2eq2 t k
q1,2 =       ,
2 2
0 2

When   0 ,
m
c
   is a real number whose value is   ,
2 2

so both q1 and q 2 become 'real' and essentially 'negative'
2m
x(t )  A1e  A2e
q1 t q2 t
hence
q1 x(t  0)  x(t  0)
x(t )  q1 A1e q1 t
 q2 A2e q2 t
q1  q2
 A2

Hence, q1 x(t  0)  x(t  0)

x(t  0)   A1
x(t  0)  A1  A2 q1  q2
x(t  0)  q1 A1  q2 A2 ‘Overshoot’ : not possible. Oscillations
q1 x(t  0)  q1 A1  q1 A2 being completely killed, this oscillator
x(t  0)  q1 A1  q2 A2 isPCD_STiCM
called ‘OVERDAMPED’. 39
 k
0 
2
CASE 2 UNDERDAMPED OSCILLATOR
m


c  0     is an imaginary number
2 2

2m
q1,2 =     2   2 ,

q1 =   + i 0 2   2    i where   0 2   2
q2 =   - i 0 2   2    i i.e.,   0
by an amount determined by 
  i t    i t
x(t )  A1e  A2e
x(t )  e  t
A e
1
 i t
 A2e
 it


x(t )  e t ( A1  A2 ) cos(t )  i( A1  A2 )sin(t ) 
PCD_STiCM 40
CASE 2 k
0 
UNDERDAMPED OSCILLATOR 2

When   0 ,  2   2 is an imaginary number m

c

x  e t A1e i t  A2e
 it

 
2m
x  e t ( A1  A2 ) cos(t )  i( A1  A2 )sin(t ) 
Introduce two new parameters B & θ instead of A1 and A2.
→ insight in the nature of the solutions
A1  A2  B sin 
i( A1  A2 )  B cos  iBe i iBei
A1   , A2  
2 2

x(t )  Be t sin  cos(t )  cos  sin(t ) 

x(t )  Be  t
sin(t PCD_STiCM
) 41
UNDERDAMPED OSCILLATOR
  0 ,  2   2 :imaginary  2  k
0

x(t )  Be t sin  cos(t )  cos  sin(t ) 

m
c

x(t )  Be t sin(t   ) 2m

  0 2   2
i.e.,   0 by an amount determined by 

• Solution: sinusoidal, at circular frequency ω

determined by the two parameters 0 and  .

• Frequency ω < ω0

• Amplitude decreases exponentially with time

• shifted by θ
Oscillation is phasePCD_STiCM 42
 k
0 
2

UNDERDAMPED OSCILLATOR m
c

2m
When   0 ,  2   2 is an imaginary number

x(t )  Be  t
sin(t   )

This solution is *NOT* “periodic”; *NOT* repetitive.

One may regard the oscillatory sinusoidal term to

have an exponentially diminishing amplitude.

But the *ZEROES* are repetitive; strictly

periodic; occur at a time period of T=2π/ω,
called “period of the damped oscillator”.
PCD_STiCM 43
UNDERDAMPED OSCILLATOR   0 ,  2   2 :imaginary
k c
0  &  
2
x(t )  Be sin(t   )
 t
m 2m
*ZEROES* are repetitive; strictly periodic; occur at a time
period of T=2π/ω, called “period of the damped oscillator”.

The number of oscillations in

a small time interval  t In two successive periods ‘T’,
t  t the amplitude falls according to
N (in  t)= = t=
T 2 the following ratio:

Bn1 Be (t T ) 
  e T
 e 
1
  ; frequency Bn Be t
T Logarithmic decrement factor
PCD_STiCM 44
UNDERDAMPED OSCILLATOR
Logarithmic
  0 ,  2   2 :imaginary decrement factor
B2 Be  ( t T )
x(t )  Be t sin(t   ) 
B1 Be  t
In two successive periods ‘T’, the B2  T  
amplitude falls according to the following ratio:  e  e
B1
Question: By what amount does the amplitude
diminish over a time  t  NT ?
BN 1  NT
Now, =e  e  N ,
B1
1
hence, when  = ,
NT
the 'amplitude decrease factor' would be 1 .
PCD_STiCM 45
e
 UNDERDAMPED OSCILLATOR

  0 ,  2   2 is an imaginary number

x(t )  Be  t
sin(t   )
Unlike the ‘overdamped oscillator’ (no oscillations),
we do have oscillations that are ‘damped’, not
‘killed’; hence called UNDERDAMPED
OSCILLATIONS

PCD_STiCM 46
Case 3: ‘CRITICAL DAMPING’ q1,2 =     2   2

 =0 , q1  q2  q : the two roots are equal

x(t )  Ae qt

Can we get the 2nd linearly independent  t

solution by considering the following x(t )  Bte
simplest departure from the previous one?
 t  t  t
x(t )  Ae  Bte  ( A  Bt )e
A
At t   , the system reaches the equilibium position,
B
and then, after the overshoot,
the next attainment of equilibrium can be
only after infinite time.
PCD_STiCM 47
Let us recapitulate main results!

Overdamped Oscillator
When   0 , i.e. c 2  4mk ,
 2   2 is a real number whose value is   ,
so both q1 and q 2 become 'real' and essentially 'negative'

Amplitude versus time No oscillation !

No overshoot off
equilibrium

PCD_STiCM 48
Underdamped Oscillator x  A1e q1 t
 A2e q2 t

When   0 ,  2   2 is an imaginary number; c 2  4mk ,

q1 =    i; q2 =    i; where   0 2   2
i.e.,   0 by an amount determined by  c

x  e t ( A1  A2 ) cos(t )  i( A1  A2 )sin(t ) 
2m

x(t )  Be  t sin  cos(t )  cos  sin(t ) 

x(t )  Be   t sin( t   )
Exponential fall of amplitude
‘zero/equilibrium crossings’ do occur
Oscillations damped, not killed !
Amplitude versus time
Amplitude diminishes PCD_STiCM
more rapidly for larger values of c 49
Critically damped oscillator
 =0 , c2  4mk , q1  q2  q : the two roots are equal
x(t )  Ae t  Bte t  ( A  Bt )e t
The equilibrium position x=0 is reached
A
in 'finite' time interval, t   .
B
After the overshoot, the next attainment
of equilibrium can be only after 'infinite' time.

Amplitude versus time

PCD_STiCM 50
Amplitude versus time: all the three cases

Overdamped: oscillations
are `killed’

Underdamped:
oscillations are damped

critically damped:
overshoots equilibrium
in finite time
PCD_STiCM 51
 We will take a break ……

…… Any questions ?
pcd@physics.iitm.ac.in

Next:
Forced oscillations
Restoring force, damping force
and driving force…..
RESONANCES…..
PCD_STiCM Waves…… 52
STiCM

Select / Special Topics in Classical Mechanics

P. C. Deshmukh

Department of Physics
Indian Institute of Technology Madras
Chennai 600036

pcd@physics.iitm.ac.in

STiCM Lecture 09: Unit 2 Oscillators,

PCD_STiCM
Resonances, Waves 53
 Forced oscillations
Restoring force, damping force and driving force

"The Hand That Rocks The

Cradle,
Is The Hand That Rules The
World"

-William Ross Wallace

This poem was first published in

1865 under the title "What Rules
The World".
PCD_STiCM 54
 Forced oscillations
Restoring force, damping force and driving force
F  mx  kx  cx  Fdr
i.e.,
c k Fdr
x x x 
m m m
For a simple pendulum with damping ,
c g Fdr
  
For an LCR oscillator, ml l ml
R 1 Vdr
Q Q Q
L LC L
1
or, LQ  RQ  Q  Vdr PCD_STiCM 55
C
 c k Fdr
F  mx  kx  cx  Fdr or x  x  x 
m m m
Actual form of the solution depends on the functional form of Fdr
Let Fdr =F0ei ( t  ) , a periodic force, with frequency 
 is a phase angle - depends on 'when' we 'start' the driving force
k c F0 i ( t  )
0  &  
2
x  2 x   x  e
2
m 2m 0
m
Special case: No damping
 F i (t   ) 
x   x   e
2
 .
m 
Complex amplitude which i it it
includes time-independent
F dr  F e e = Fe
phase eiθ PCD_STiCM where F  F e
i 56

  F i (t   ) 
x   x   e Fdr  Feit where F  F ei
2
 .
m 
x(t )  xeit , (where x includes the phase factor) is a solution
of the differential equation for damped, forced vibrations
The exponential form allows us to
x  ix, interpret the effect of differentiation
with respect to time through the
x  (i) x 2
operator (d/dt) to be equivalent to
Using above relations in multiplication by (iΩ )
i (t   )
x   2 x  [ F e ] / m, Note !
we get (02   2 ) x  F / m, , the driving frequency becomes
equal to 0 , the natural frequency,
F
x
 o )
m ( 2
  2 the amplitude blows up to infinity.
PCD_STiCM 57
General case, including damping:
Note! There are two angles
F0 to keep track of!
x  2 x  0 x  exp i (t   )
2

m
‘Timing’
 : – when
Try a solution x(t )  Aeit
exactly do
with A  A0 ei (  )  : phase lag of
you start
applying the

x  A(i)eit  ix oscillation w.r.t. driving force

it the driving force

x  A(i) e 2
  x
2

Fdr  F ei eit =Feit

Substituting for x and x : where F  F ei
[2  i(2)  02 ] x(t )  ( F / m)eit
it
 ( F / m)eit
i.e.[(02  2 )  i 2 ] A ePCD_STiCM 58
General case, including damping:
i it it
F0 Fdr  F e e =Fe
x  2 x  0 x  exp i (t   )
2

m where F  F ei
x(t )  Aeit
i (  )
with A  A0e

[(02   2 )  i 2 ] A eit  ( F / m)eit

A
F / m
     i2
2
0
2

A0 e i (  )

 F0 ei / m

 0   i 2
 2
  2

ei
as F  F0PCD_STiCM 59
 cancel

A0e i (  )

 F0ei / m  ; A0e i

F0 / m
     i2
2
0
2
 0   i2
 2
  2

e i

F0 /(mA0 )
 0   i2
 2
  2

Separate now the real and imaginary parts

by multiplying both numerator and denominator
by the complex conjugate of the denominator

0  0  0
     F /  mA  2
2 2
F / mA
cos   and sin  
0 0

  
2
0
2 2
  4 22  0 
 2

2 2
 4 22
PCD_STiCM 60
 0  0  0
     F /  mA  2
2 2
F / mA
cos   and sin  
0 0

  
2
0 
2 2
 4 22  0 
 2

2 2
 4 22

 2  2
and   tan  2 1
2
; tan  = 2

 0    0   2

Squaring and adding F0

A () 
sin  & cos 
 
2 2
2 2
m 2
0   4 22
Recall that our solution is: Phase factor  changes
x(t )  Aeit markedly with the frequency
with A  A0ei (  )  of the driving force.

( F0 / m)
x(t )  ei (t   ) .
(02  2 )2  4 2 2
PCD_STiCM 61
 F0 i (t  )
Thus the solution for x  2 x   x  e
2
0 becomes
m
( F0 / m)
x(t )  ei (t   ) .
(02   2 ) 2  4 2 2

Physical features of the steady state solution:

The oscillation is out of step with Fdriving through the angle .
The amplitude of the oscillation is governed by the amplitude
of the driving force, modulated further by the factor
1 , and also by the inertia m
(0   )  4 
2 2 2 2 2

Fascinating
Nature of the solution depends on  applications in
and on the proximity of  to 0 . mechanical, electrical
and many other
PCD_STiCM physical systems. 62
x(t )  Aeit x(t ) 
( F0 / m)
ei (t   ) .
with A  A0e i (  ) (02  2 )2  4 22

F0
A () 
m   
2
0 
2 2
 4 22

A0  A0 ()

As a function of the frequency of the driving

force, when will the amplitude of oscillation
be a maximum?
PCD_STiCM 63
Condition for Resonance
dA0 F0
when is  0? A () 
d
m   
2
0 
2 2
 4 22

Two frequencies
are of interest 0 Intrinsic, natural frequency.
 External, under our control!

In the absence of damping, the condition

that the amplitude is maximum is
that   0 ….. but what when damping is
present?

PCD_STiCM 64
Reference: Fowles ‘Analytical Mechanics’; Our notation is slightly different!
Condition for Resonance F0
A () 
when is
dA0
 0? m   
2
0 
2 2
 4 22
d
0 Intrinsic, natural frequency.
 External, under our control!

dA0

1 F0
2(  2
0   2
)(-2)+ 8  2

when is  2m  0?
d (02  2 )2  4 22 
3/ 2

The N r is zero when 2 = 02  2 2 i.e.  r  02  2 2

  0  ( 2 / 0 )   r , resonance frequency
condition for resonance for a damped driven pendulum
PCD_STiCM 65
 r  02  2 2 Recall that the frequency of the
unforced (underdamped) oscillator is:
r : resonance frequency
  0 2   2
 2  2

 r  0  1  2 
2

 0   2  02   2 ; 02   2   2

 0
 2 2 
1  2 
r    2
  2
  2 2

 0 
1/2
  
2  2     2
 0  1  2  r    2 2
  1  2  
 0      
2    2 1/2
 0    1  2 
0
  
 2 
  1  2 
2 
 PCD_STiCM 66
 Amplitude at Resonance A ()  F0


0 Intrinsic, natural frequency. m   

2
0 
2 2
 4 22
 External, under our control!
r  02  2 2
r : resonance frequency

F0
A () MAXIMUM  m
     2   4 2 02  2 2 
2
2 2 2
0 0

F0 i.e.
A0 () MAXIMUM  m F0
 2 A0 () MAXIMUM 0 2   2
2 0  
2 2 m
PCD_STiCM 67
 F0
Using:  2 A0 () MAXIMUM 0 2   2
m

F0
A () 
in
m  2
0  2 2
  4 
2 2

we get:
2 A0 () MAXIMUM 0   2 2
A () 
 2
0  
2 2
 4 
2 2

PCD_STiCM 68
 2 A0 () MAXIMUM 0 2   2
A () 
 2
0  
2 2
 4 22

Approximation
02  2  0   0      0
 0    20 
 r  0
A0 () MAXIMUM 20
A () 
 0  0 
2
   2  4 2
0
2
Cancelling
20 A0 () MAXIMUM 
in Numerator A () 
0      2
2
& Denominator PCD_STiCM 69
 F0 i (t  )
Thus the solution for x  2 x   x  e
2
0 becomes
m
( F0 / m)
x(t )  ei (t   ) . ‘particular’ solution
(02   2 ) 2  4 2 2

We must add the solution of the corresponding

homogeneous equation (that of ‘unforced’ damped
oscillator) as well.

This part is a transient solution consisting of

oscillations of decreasing amplitude for
under-damped oscillator.
PCD_STiCM 70
The GENERAL solution for the
damped driven oscillator will be
 t
x(t )  Be sin(t   ) 
( F0 / m) i (t   )
e
(0   )
2 2

Damping ignored in the

steady state part , but not in the transient.
( F0 / m)
ei (t   ) .
(02  2 )2  4 22
Why?
PCD_STiCM 71
  t ( F0 / m) i (t   )
x(t )  Be sin(t   )  e
(0   )
2 2

The three circular frequencies involved :

0 , the natural frequency;
 , the frequnecy of the damped oscillator
and , the driving frequency

Remember!   02   2 , where   k / m for mass-spring oscillator,

1
0 
g and   , for LC -circuit
l LC
for simple pendulum
PCD_STiCM 72
 
A0 () MAXIMUM  0   0 0  
A () 
0      2
2
Energy is proportional to
the square of the
when   0   ,
A0,max  A0,max amplitude,
A ()  =
 
2 2
2 and for frequencies
1 separated by 2
A0 () 2  Amax 2
2 about the resonance
frequency, the energy
reduces by a factor of 2.
2 “RESONANCE WIDTH”
 0
Define: Q  (for the case of weak damping)
2 2
Quality Factor
PCD_STiCM 73
A0 ()
Less Damping

More Damping

r
PCD_STiCM
 74
Frequency of the Driving Force
We will take a break ……

…… Any questions ?
pcd@physics.iitm.ac.in

Next:

….. Waves……
PCD_STiCM 75
STiCM

Select / Special Topics in Classical Mechanics

P. C. Deshmukh

Department of Physics
Indian Institute of Technology Madras
Chennai 600036

pcd@physics.iitm.ac.in

STiCM Lecture 10: Unit 2 Oscillators,

PCD_STiCM
Resonances, Waves 76
The Tacoma Narrows Bridge in Washington state, was with
1.9 km length one of the largest suspended bridges built at
the time. The bridge connecting the Tacoma Narrows
channel collapsed in a dramatic way on Thursday
November 7, 1940. Winds at about 50-70 km/hr produced
an oscillation which eventually broke the construction.

Forced/Driven
Damped
Oscillator

See video of this `Disaster

at Resonance’ at the
internet link given below!

http://www.math.harvard.edu/archive/21b_fall_03/tacoma/index.html
PCD_STiCM 77
 Resonances Enrico Caruso
Enrico Caruso - could shatter a 1873 - 1921
crystal goblet by singing a note of
just the right frequency.

In 2005, Discovery TV
Channel recruited
rock singer and vocal
coach Jamie Vendera
to hit some crystal
ware.

http://www.youtube.com/watch?v=Jy8js2FmGiY
PCD_STiCM 78
 Chalo,

HAMMER
Our se hi kaam
own
BAIJU chala lete
of
MANDI
hein!

PCD_STiCM 79
Google: MRI picture
PCD_STiCM 80
Solutions of the
oscillator problem play
a fundamental, crucial
role in DSP, information
transmission, etc.

PCD_STiCM 81
A wave packet, Wave / Pulse propagation
or a wave pulse,
is made up by y  f ( x)
y  f ( x  vt )
superposing a
large number of Unit 2:
Oscillators
sinusoidal Resonances
waves. Waves
Each X
component Wave pulse at t=0 vt
wave has a Peak is at x=0 Wave pulse at t > 0
Wavelength  Peak is at x=vt
The component
travels at its own Shape: same
phase velocity v   PCD_STiCM
Medium: Non-dispersive
82
Pulse shapes ---- Fourier Analysis

Fourier :

Any periodic function can be Jean Baptiste Joseph

written as a sum of simple Fourier
oscillating functions March 21, 1768
May 16, 1830

- sine and cosine functions

PCD_STiCM 83
  x  x 
Plot the function: f ( x)  2  H    H   1   1
 L  L 

when x  0, 2 L 
H (x) Unit
0 for
2: x<0 Heaviside step function

Oscillators
1 for x>0 “Unit step function”
Resonances
Waves

2L
x
x 0
X=0 L

-1

 x  x 
Square Wave: f ( x)  2  H    H   1   1
L
PCD_STiCM  L  84
 
4 1
f ( )  
 n1,3,5,.... n
sin  n 

PCD_STiCM 85
 8  1 1 1 1 
f ( )  sin   sin(3 )  sin(5 )  sin(7 )  sin(9 )  ......
 2  32 52 72 92 

Applications:

Digital Signal Processing

for example!

PCD_STiCM 86
We worked with the function f  f ( x)
 x  x 
Square Wave: f ( x)  2  H    H   1   1
 L  L 

also, we examined the saw-tooth triangular waves

Unit 2:
Oscillators
In general, in wave/pulse propogations, we
Resonances have
function of both spaceWaves
and time: f ( x, t ),
or, more generally, f (r , t )
often called the wavefunction ( r , t ).

 ( x, t )  f ( x  vt )
PCD_STiCM 87
If all the components of the wave-packet travel at
the same speed, the ‘shape’ of the wave-packet
propagates without distortion.

This is the property of a non-dispersive medium.

In a dispersive medium, the wave packet ‘spreads’.

PCD_STiCM 88
 
 z 
 
 ( z, t )  A cos  t    A cos(t  kz ) where k 
v

 v 
 2 
phase velocity v     
k k T

Note:
At fixed z, this represents a harmonic oscillation in time.
At fixed t, this represents a harmonic oscillation in space.

Important parameters: frequency, period, wavelength,

PCD_STiCM 89
amplitude, phase
 ‘phase’
The wavefunction  ( z, t )  A cos(t  kz )
where (t  kz )   ( z, t ), the phase function

At a given z, the phase varies linearly with time

At given t, the phase varies linearly with the space coordinate

In a medium, surface of constant phase is given by:

0  d   dt  kdz
dz 
  v , phase velocity.
dt k

‘phase velocity’ is the speed at which a wave-front defined by a surface

at a certain fixed phase ( e.g. a crest) advances with time.
PCD_STiCM 90
 
Phase velocity v  for a nondispersive group of waves. 
k

NON-DISPERSIVEWAVES: is constant.
k
k
In general, for dispersive waves,
v has a much more
complicated dependence on  (i.e. k ).
 is a function of k , given as  (k ), v  v (k ),
the functional form is different for different systems
Actually, it is the MEDIUM that is non-dispersive.

Properties of the MEDIUM are central to the phenomenology of

NON-DISPERSIVE WAVES. PCD_STiCM 91
 Superposition :AMPLITUDE-MODULATED TRAVELING WAVE

Superposition:
 ( z, t )  A cos(1t  k1 z )  A cos(2t  k2 z )

Then, we get
 ( z, t )  Amod ( z, t ) cos(avet  kave z )
where Amod ( z, t )  2 A cos(mod t  kmod z )
1 1
mod  1  2  ; kmod   k1  k2 
Ref: Berkeley/ 2 2
1 1
1  2  ; kave   k1  k2 
Vol.3/
Page270 also, ave  PCD_STiCM 92
2 2
 ( z, t )  Amod ( z, t ) cos(avet  kave z ) At what speed does
where Amod ( z, t )  2 A cos(mod t  kmod z ) the modulation
propagate?
1 1
mod  1  2  ; kmod   k1  k2 
2 2
1 1
also, ave  1  2  ; kave   k1  k2 
2 2

To follow a given modulation wave crest of the

modulation amplitude Amod ( z, t ),
we need to maintain a constant value of (mod t  kmod z )

i.e., in time dt , z must increase by dz in such a way that

d (mod t  kmod z ) PCD_STiCM
 (mod dt  kmod dz )  0 93
 ( z, t )  Amod ( z, t ) cos(avet  kave z ) At what speed does
the modulation
propagate?

In time dt , z must increase by dz in such a way that

d (mod t  kmod z )  (mod dt  kmod dz )  0

To satisfy this, the modulation must propogate at:

dz mod 1  2
 v mod  
dt kmod k1  k2
 d
   v g  'group velocity'
 k dk
PCD_STiCM 94
If all the components of the wave-packet travel at
the same speed, the ‘shape’ of the wave-packet
propagates without distortion.

This is the property of a non-dispersive medium.

In a dispersive medium, the wave packet ‘spreads’.

PCD_STiCM 95
 x
A Refraction
‘1’ 1 a Why does the light
B’ B ray go along the path
A B C,
‘2’ 2 b and not along
C A B’ C
d
Time taken for light
   
1/ 2 1/ 2
a 2
x 2
b 2
 (d  x ) 2
to travel the path
t  A B C:
v1 v2

   
dt 1 1 2 1/ 2 1 1 2 1/ 2
 a  x2 (2 x)  b  (d  x) 2 (2(d  x))
dx v1 2 v2 2
dt 1 x 1 (d  x) sin 1 sin  2
  0 
   
1/ 2 1/ 2
dx v1 a 2  x 2 v 2 b 2  (d  x) 2 v1 v2
sin 1 v1
Zero (Fermat’s principle)   nRefractive Index
PCD_STiCM sin  2 v 2 96
Refractive index, n:
Ratio of phase velocity of light in
vacuum to that in the medium

c vac vac
n  
v  

Different colors refract

through
nr  nr  
different
angles Refractive Index
Red
Normal depends on
dispersion FREQUENCY in a
Blue dispersive medium
PCD_STiCM 97
 c  vs. k graph:
  2  2  ck
 constant slope, speed of light
v vacuum c vacuum kmedium
n   
vmedium vmedium medium kvacuum
In the medium:
v medium c 1 c
  2  2  2  2
medium n medium nmedium
c
 kmedium Refractive Index depends
n( )
on frequency in a dispersive medium
 vs. k graph: not linear  Dispersion relation
PCD_STiCM 98
 Control speed of light ! Bring it to a halt !
Jan 18, 2001
Playing stop and go with light
http://physicsworld.com/cws/article/news/2729

2001 PRL 86:5 783

REVIEWS OF MODERN PHYSICS, VOL. 77,

APRIL 2005
Electromagnetically Induced Transparency:
Optics in coherent media
PCD_STiCM 99
Laser smashes light-speed record
http://physicsworld.com/cws/article/news/2810

In a recent (2000) experiment at Princeton, L.J.Wang et al.

managed to get a laser pulse travels at more than 300
times the speed of light !
L J Wang et al. 2000 Nature 406 277
Laws of physics: intact!
‘Normal dispersion’: group velocity < phase velocity.
‘Anomalous dispersion’:
R.I. decreases as frequency increases; vgr > vph.

 
>c
normal Red Blue
Normal dispersion Anomolous
Blue Red
PCD_STiCM 100
 c 
 n  vac nr  nr   My heart leaps up when I behold
v  A rainbow in the sky:
So was it when my life began;
So is it now I am a man;
So be it when I shall grow old,
Or let me die!…
- William Wordsworth
R.I. of
water for
red is Questions:
~1.331
1. Why is the red outside and blue inside?
R.I. of 2. Which part of this picture is the brightest,
water for
blue is and why?
~1.343

Rainbow, seen from the ‘Maid of the Mist’ ride101

PCD_STiCM at the
th
Niagara Falls, U.S.A., 18 July, 2009. - pcd
 We Will tAke A breAk…
…… Any questions ?
pcd@physics.iitm.ac.in

Next: Unit 3 Dynamical Symmetry

of the Kepler Problem
Plane polar
Cylindrical polar
Spherical polar coordinate systems
PCD_STiCM 102