B.N.

Bandodkar College of Science, Thane

T.Y.B.Sc

Classical Mechanics

Prepared by Dr M.N. Nyayate

Mechanics of Continuous Media

V.Imp.

Equation of motion of a vibrating string:

Consider a string of finite length „L‟ fixed at the rigid supports. If it is plucked
and released, it vibrates. Our aim is to study motion of this vibrating string. For
this purpose an equation governing the motion of the string is to be obtained
using Newton‟s laws of motion.
There are infinitely many particles on the string. Hence applying Newton‟s
laws to these particles, we get infinitely many equations. Instead of this, we
employ Euler‟s approach.
Consider an element „dl‟ of the string. Newton‟s law can be applied to
such an element. We introduce two co-ordinates, x which changes continuously
from zero until the end of the string. The other dependant variable is the
displacement of the element of the string at position x from the mean position
in the vertical direction. Variable u is a function of position and time t i.e. u(x,
t).

Following are the assumptions used in obtaining the equation of motion of the
vibrating string.
1) The string vibrates in the vertical plane only (No sidewise swing)
2) Each particle on the string vibrates along the vertical directions only (No
sidewise deflection).
3) The Tension in the string  is small and constant throughout the length.

Since the element dx vibrates along the vertical direction only, the net force
must be along the vertical direction.
The net force acting on the element dx along vertical direction can be calculated
as follows:
f net   sin  x  dx   sin  x
  small sin   tan
f net   tan  x  dx   tan  x
 u   u 
     
 x  x  dx  x  x
 u   u 
    
 x  x  dx  x  x
f net  lim dx
dx   0 dx
   u 
f net   dx
x  x 
 2u
f net   2  x, t dx (1)
x

According to Newton‟s 2 nd law of motion. This net force must be equal to the
product of mass and acceleration of the element in the vertical direction.

 Mass  Acceleration  f net

 2u  2u
 dx    dx
t 2 x 2
 2u    u
2
 x, t     2  x, t   0
x 2    t

Where  - linear mass density of a string i.e. mass per unit length of the string

 2u    2u
x, t    1  2  x, t   0
x    t
2

2
 2u    2u
x, t    1 
 t 2
 x, t   0
x 2
   
 2u 1  2u
 x , t    x, t   0
x 2 c 2 t 2

Where c    = is velocity of propagation of waves along the string

This is the required equation of motion of the vibrating string.

Normal Modes of Vibration: -

The 2nd order partial linear different equation with constant coefficient satisfied
by the vibrating string (equation of motion of the vibrating string), is given by

 2u 1  2u
 x , t    x, t   0
x 2 c 2 t 2

This partial different equation is subject to the following boundary condition as

the string is fixed at its two ends to the rigid support.
u 0, t   0 or u  x, t  x 0  0
u L, t   0 or u  x, t  x  L  0

In order to solve the above partial differential equation we employ the method of
separation of variables. The solution u(x,t) can be expressed as a product of two
functions such that each function is a function of only one independent
variable.

Let,
ux, t   X x t  (1)
 u
x, t   t  d X 2x 
2 2

x 2
dx
and
 2u d 2 t 
 x , t   X  x 
t 2 dt 2

Substituting this in equation of motion

 2 y 1  2u
 0
x 2 c 2 t 2

d2X 1 d 2
t   X  x  0
dx 2 c 2 dt 2
by ux, t   X x t 

1 d 2 X 1 1 d 2
 0
X  x  dx 2 c 2 t  dt 2

Multiply throughout by c2
c2 d 2 X 1 d 2
 0
X  x  dx 2 t  dt 2

c 2 d 2 x  1 d 2 t 

X  x  dx 2 t  dt 2

The L.H.S. of above equation involves only independent variable x where as

R.H.S. involves only independent variable t. Since x and t are independent
variables both the sides must be equal to some constant.

c 2 d 2 x  1 d 2 t 
 = constant
X  x  dx 2 t  dt 2

In order to have oscillatory solution, the constant must be negative.

c 2 d 2 x  1 d 2 t 
   2
X  x  dx 2
t  dt 2

If the R.H.S. is not a negative constant but a positive constant then the solution
will not be oscillatory but will be exponentially decreasing or increasing
function.

1 d 2 t 
   2 (2)
t  dt 2

c 2 d 2 x 
  2 (3)
X  x  dx 2

Equation (2) can be solved as follows

1 d 2 t 
  2
t  dt 2

d 2 t 
2
  2 t   0
dt

This is 2nd order, ordinary, linear different equation with constant coefficient.
 t   A cos t  B sin t

Equation no. (2) can also be solved similarly,

c 2 d 2 x 
  2
X  x  dx 2

d 2  x    
2

    0
dx 2 c

   
 X  x   c cos x   D sin  x 
c  c 

u x, t   X x t 
     
u x, t    A cost  B sin t  C cos x   D sin  x  
 c   c 

Applying the boundary conditions

i) ux, t  x0  0  X x x0

 0  c cos 0  D sin 0  c
c0
 
 X x   D sin  x 
c 
ii) ux, t  x L  0  X x xL
 L 
 D sin  0
 c 
case I) D=0.

This is not allowed as it gives the trivial solution that the string is at rest
forever.

OR

 L 
Case ii) sin  0
 c 
L
 case (a) 0
c
This is not allowed as   0 L  0
L
OR case (b)  n n=1,2,3,----
c
nc
 n 
L

 u x, t   X x Θt 
 x 
 D sin  n A cos n t  B sin  n t 
 c 
Without loss of any generality, the arbitrary constant D can be taken as unity.

 x   x 
 ux, t   A sin n  cos  n t  B sin n  sin  n t
 c   c 
 nx   nct   nx   nct 
 u n  x, t   An sin   cos   Bn sin   sin   n = 1,2,3,… n  N
 L   L   L   L 
Normal modes of oscillations:

The solution of the vibrating string are given by

 nx   nct   nx   nct 

u n  x, t   An sin   cos   Bn sin   sin  
 L   L   L   L 

If the value of n is fixed then the frequency of oscillation of the vibrating particle
on the string is also fixed. But the frequency is same for all the particles on the
string. Thus whenever, a string vibrates in such a way that each an d every
particle oscillates with the same frequency then this mode of vibration is called
normal mode of vibrations.
It is observed that the number of normal modes of system of many particles is
equal to the number of particles in the system. For e.g. two coupled harmonies
oscillators has two particles. Hence, it has only two normal modes of
oscillations namely symmetric and anti-symmetric modes of oscillation.
In the case of vibrating string, the number of particles is infinite. Therefore
there are infinitely many normal modes of vibrations which can be obtained by
permitting n = 1,2,3...

The most general solution of the vibrating string is given by,


u  x, t    u x  x, t 
n 1

  nx   nct   nx   nct 
   An sin   cos   Bn sin   sin  
n 1   L   L   L   L 

All the arbitrary constants in the above general solution of the vibrating strin g
can be obtained by using the following initial conditions.

i) ux, t  |  u 0 x  given
t 0

u x, t 
ii) |  v0 x  given
t t 0

By using the Fourier Series Analysis, the constant An is given by,

 n x 
An   u 0  x sin 
2 L
dx
L 0
 L 

The constants B n can be obtained similarly by differentially u with respect to

 nx 
time and then multiplying by sin  
 L 
 nx 
L
v0 x sin
2
Bn  
ncL 0  L 
dx

 2u 1  2u
Q. Show that if any function f x  ct  satisfies the wave equation

x 2 c 2 t 2
for small amplitude transverse waves on a stretched string then the wave
velocity is  c .

The function of the type f x  ct  represents transverse waves on a stretched

string as the waves travels along the string the phase at any point on the string
is given by x–ct.
In order to find the velocity of the waves we consider points on the string with
the same phase.

x – ct = constant phase

differentiating with respect to „t‟

dx
c  0
dt
 dx 
 c
 dt 

This is the velocity of the waves traveling in the positive x direction. The velocity
of the wave traveling in the opposite direction i.e. reflected waves is therefore „ -
c‟.
Hence the velocity of the waves can be 'c' .

Let x – ct = y
f f  x  ct 

x x
f  y  y

y x
2 f   f  y  y 
 
x 2
x  y x 
  f  y  y  y
 
y  y x  x
 2 f  y   y 
2

  
y 2  x 
 x  ct  y
y
1
x
2 f 2 f
 2 (1)
x 2 y

f f y

t y t
2 f   f y 
 
t 2
t  y t 
  f y  y

y  y t  t
 2 f  y 
2

 2  
y  t 
y
  c
t
2 f
 2 c2
y
1 2 f 2 f
 2 (2)
c 2 t 2 y

Comparing (1) and (2)

2 f 1 2 f

x 2 c 2 t 2
f satisfies the wave equation

2] Motion of string of length „L‟ stretched between two points is given by

  nx   nct   nx   nct 
ux, t     An sin  cos   Bn sin  sin 
n 1   L   L   L   L 
L
If initially string is given transverse displacement at midpoint from rest (Bn)
10
show that
i) An = 0 for n = 2,4,6,---
ii) B n = 0 for all values of n

u 0 x  
1 L
x; 0 x
5 2
  x  L ;
1 L
xL
5 2

 nx 
L
u 0 x sin
2
An 
L0  L 
dx
L
x  nx   nx 
2 L

0 5 sin L dx  L L 5 L  x sin L x

2 2 1

L
2
L
 nx   nx   nx 
2 L L
2 2 2

5L 0 x sin L dx  5 L sin L dx  5L L x sin L dx
2 2
Consider
 nx   nx    nx  d 
 x sin L dx  x sin L dx    sin L  dx x
 nx  L L nx
 x  cos

 
L  n  n  cos
L
dx

 nx  L L  nx  L
  x cos   sin  
 L  n n  L  n
 nx  L  L   nx 
2

  x cos    sin 
 L  n  n   L 
L L
  nx  2  L   nx 
2

 sin 
2
L
   x cos    
n   L  0  n   L 
 L  n   L   n 
2
L
  x cos 2    n  sin 2 
n       

Let n = 2m m = 1,2,3---
 nx 
L

L  x sin L dx
2
 L
  nx    nx 
2
L  L 
  x cos L    n  sin L 
n    L      L
2 2

n   L  n 
2
L  L 
   L cos n  cos     sin n  sin L 
n  2 2   n   
Let n = 2m m = 1,2,3
2
 m  L 
 L  2  1    n  0
L L
=
n    
 nx  L  m
L
 L   1 
L
=  sin  dx 
L  L  n  2 
2

nx 
L
L  L2 L2
=   cos  
n  L  L 2m m
2

L  n 
= cos n  cos
n  2 
=
L
n

1   1
m

=
2  L2
 
5L  4m
 
1 m 
 
2 L
1  1m
 2  L2
  
L2 
 1m 
 5 n 5L  2m 4m 

Kinematics of fluids
The fluid is supposed to be divided into very large number of fluid elements
called as fluid particles. The motion of fluid is characterised by certain
parameters such as pressure, density, velocity etc. These parameters are
functions of space and time co-ordinates. Laws of motion are applicable to
actual fluid particles and not to the space co-ordinates. Therefore, we follow the
fluid particle along its motion for small interval of time in order to find the total
role of change of parameter with time. Consider the parameter pressure of the
fluid.
dp  px  x, y  y, z  z, t  t   px, y, z, t 
p p p p
 x  y  z   t
x y z t
dp  p p p  p
  v x  v y  v x  
dt  x y x  t


p  ˆ p ˆ p ˆ p  ˆ
 i
t  x
j
y z 

 k  v x i  v y ˆj  v z kˆ 
p  
  p  v
t
dp p  
  v  p
dt t
d 
  v 
dt t

 

To determine
d
V  :
dt
Consider a small volume element V is the form of an cubicle with the edges x,
y, z
We have
V  xyz
d
V   d xyz 
dt dt
d  d  d 
  x yz   y xz   z xy (1)
 dt   dt   dt 
To find
d
x 
dt
 v x x x  v x x
 v   v x x 
 lim  x x δx δx
δx0
 δx 
d x  v x
 x (2)
dt x

Similarly
v
d
y    y y (3)
dt y
d
z    v z z (4)
dt z

Substituting (2),(3) and (4) in (1)

 v 
d
V    v x xyz   y y xz   v z z xy
dt  x   y   z 
v v y v
  x xyz  xyz  z xyz
x y z
 v v y v z 
  x   V
 x  y z 
 
d
V     v V
dt

Characteristics of fluid motion:

1) Fluid flow can be viscous are non-viscous

2) The fluid flow can be streamline or turbulent
3) The fluid flow can be rotational or irrotational
4) The fluid can be compressible or incompressible

If the fluid is incompressible then rate of change of its volume will vanish

d
V   0
dt
 
 v  0
The relation for the total time derivative for volume elements is given by

d
V     vV
dt

This relation is true for all volume elements in different co-ordinate system. It is
also valid for any finite volume of a fluid.

Consider the given finite volume V to be made up of large number of

infinitesimal volume elements.
i.e. V  V
Taking the summation of above relation
 
d  
 dt V      v V
d
L.H.S.   V
dt
d

dt
 V (Summation over volume element and time differentiation
can be interchanged, as they are independent)
dV

dt
 
 
R.H.S.      v V
 
    v V
 
     v dV
V
dV  
      v dV
dt V

Applying Gauss‟ Divergence Theorem,

dV  
   v  ds
dt s

   v  nˆds

Consider a small cap area „ds‟ on the surface of given volume V. it is imagined
that the surface element ds moves out of the volume in a normal direction. The

distance covered in time „dt‟ is v  nˆdt . It is the length of the cylinder. Therefore,

the volume of the fluid that goes out of this area in time „dt‟ is v  nˆ dt ds . Total
volume of a fluid going out of a closed surface in time „dt‟

  v  nˆdtds
s

But this must be equal to the change in a volume of the fluid in time dt. Let it
be „dV‟.

 v  nˆdtds  dV
s
 dV 
Therefore    v  nˆ ds
dt s

Physically it means that the rate of decrease of volume of a fluid must be the
same as the rate at which the fluid escapes the surface enclosing the given
volume.
Thus is in agreement with conservation of fluid matter.

Imp.
Continuity equation for fluid flow:
Consider infinitesimal volume element  V of a given fluid. Let  be the density of
a fluid.
Mass of the fluid element is given by
m  V
According to conservation of mass principle, the net rate of change of mass
should be zero.
i.e.
d
m   0
dt
   V   0
d
dt
d d V 
 V   0
dt dt

Using identities
i)
d   
dt

t

 v   

ii)
d
V     v  V
dt
 
    

 
 
 v    V      v V  0
 t 
    
  v      v  0
t
  
   v   0
t
This is required continuity equation.

This is in agreement with principle of conservation of mass. This can be

explained as follows.
  
   v 
t

Integrating over a volume V

  
V t dV  V   v dV
 
 dV    v   nˆ dS (using Gauss‟ divergence theorem)
t V s

 m  
     v   nˆ ds
t S
The L.H.S. of the above equation represents the rate of change of mass of the
fluid in the given volume V whereas R.H.S. represents the rate of flow of mass
of the fluid through the closed surface „S‟ bounding the given volume V

Imp.

Physical Meaning of curl of V :   V 
 

Consider a fluid motion containing a vortex. Let V be the fluid velocity around
 
  
the vortex. Consider the surface integration of curl of v (i.e.   v )

   v  nˆds

sur

Using Stoke‟s Theorem we get

   v  nˆd   v  dl  0
  
sur C

Fluid motion with vortex

The closed path i.e. the boundary of surface is a circular path. At each and
every point on this path the velocity of the fluid and the time element „ dt’ have
 
the same direction. Therefore v  dl is never zero. This implies that the
integration not equal to zero. In turn, it implies that the integrand of the first
 
integration is not zero. Thus   v  0

In other words, whenever there is a rotational motion of the fluid, curl of v i.e.
 
  
  v is non-zero. Thus   v is a measure of rotational motion or curling of the
fluid or angular velocity of the fluid motion.

Consider a general case in which apparently, there is no vortex in the fluid flow
 
but there is velocity gradient. In such a situation, it is observed that   v  0 .
 
 
Even in this case it can be shown that   v represents the angular velocity of
the fluid motion.

A rotating co-ordinate system rotating with constant angular velocity  . The
velocity of the fluid in inertial frame and that in the rotating frame are related
   
by v  v *    r

Where v  fluid velocity in inertial frame

v *  fluid velocity in rotating frame
 


 angular velocity of rotating frame
r  radius vector.

Taking curl of the above relation

      
  v    v *      r 

Using product rule we get,

       
               
  v    v *     r  r    r        r

Since ' ' is constant we get,

   
         
v  v*   r    r
 
  r  3
 
   
  r  
     
  v    v *  3  
  
   v *  2

Since  is angular velocity of a rotating frame that we have introduced, we
 1 
select,     v
  2  *  
 v  v  v
 
 v*  0

Thus this means that in rotating frame fluid has no angular velocity. In other
 
words, fluid has angular velocity in the inertial frame. Thus,   v is still the
measure of angular velocity of the fluid motion.

Equation of motion for an ideal fluid: -

The equation of motion is applicable to only ideal fluid motion. The fluid motion
is called ideal if it does not support any shearing stress. In other words, the
fluid motion is due to pressure difference only or any other external body force.
If fluid does not support any shearing stress, it means the fluid has zero or
negligible viscosity.
Consider small volume element  V in the form of parallelepiped.
We calculate net force acting on the volume element due to the pressure
only. It is assumed that fluid moves along +ve Xdirection. Then the pressure of
x  x is less than that at x. Hence the net force acting on the volume element
along the X-direction due to pressure difference is given by

Fx   force x   force x x

  p y z  x   p y z  x   x
  p y z x x   p y z x 
 - lim  x
x 0
 x 
 p 
δFx   x y z (1)
 x 
Similarly,
  p 
Fy   δx δy δz (2)
 y 
 p 
Fz   δx δy δz (3)
 z 
 Net force acting on the volume element due to pressure difference is given by,

δF  F x iˆ  Fy ˆj  Fz kˆ
 p p ˆ p ˆ 
  iˆ  j  k x y z
 x y z 

F  hiˆ  f y ˆj  Fz kˆ
 p p ˆ p ˆ 
  iˆ  j  k xyz
 x y z 
 
F  pV (4)
In addition to this force due to pressure, there may be additional external force
acting on the body of a fluid element.

Let f this external body force acting on the fluid per unit volume. Therefore, the

body force acting on the volume element V will be fV
According to Newton‟s 2nd law of motion the product of mass and acceleration of
the fluid element must be equal to net force acting on the element

(mass)x(acceleration) = Net force

  
 dv 
V    PV  fV
 dt 
 V is an arbitrary volume element  0
  
dv
  P  f
dt
(f may be gravitational force)

 
 v      
    v   v   P  f
 t 
 
 
v    1 
  v   v  P 
t 
f

This is required equation of motion for the ideal fluid motion. It is known as
Euler‟s equation of motion for the ideal fluid.

Conservation Laws in Fluid Motion:

There are several conservation laws applicable to the system of particles such
as conservation of mass, conservation of linear momentum, conservation of
angular momentum and conservation of energy. All these conservation laws are
because of Newton‟s laws of motion.

Similar conservation laws are applicable even for fluid motion. For e.g.
Continuity equation for fluid motion leads to conservation at mass principle.
  
   v   0
t
This continuity equation can be generalised which will take into account a
source of fluid. Let Q be the rate of generation of fluid per unit volume. In this
  
case, the continuity equation modifies to    v   Q
t
 
Multiplying by dV and integrating over volume dV;  dV    v   nˆ ds   QdV
t V sur vol
The above equation implies that rate of increase of fluid mass plus rate of
escape of mass through the surface bounding the volume must be the same as
the rate at which the mass of fluid is generated in the volume.

Conservation of linear momentum:


In fluid motion, the product v plays the same role as the linear momentum in
particle dynamics. It is termed as linear momentum density.
Consider Euler‟s equation:
 
dv 
V   pV  fV
dt
m  V &
d
m  0
dt
  
 Vv   pV  fV
d
dt
 
d   
 v V   f  p V
dt
Integrate over volume V

  
d
 
dt V   dV
 v dV    pdV  f
V V

Using Gauss‟s theorem

 
   p  nˆds   fdV
s V
According to linear momentum conservation principle for particle dynamics,
rate of charge of momentum is always zero or momentum is conserved if no
external force acts on the system.
 
In case of fluid motion external force is f . We assume f =0.
d
 v dV    p  nˆds
dt V s

The first term on R.H.S. of the above equation is the force acting on the
bounding surface due to the pressure exerted by the fluid outside the volume.
Thus within the volume the rate of change of linear momentum of the fluid is
always zero. This is the conservation of linear momentum.
Similarly, conservation of angular momentum for the fluid can be established.

Oct.2000
 
Consider a fluid flow in which velocity v x, t   iˆ; x  0 Find acceleration a x, t 
at
x
of fluid element at position x and time „t‟. Is the fluid flow incompressible?
Explain.


a  x, t  
dv
dt
d  at 
  iˆ 
dt  x 

Using
d
dt

  
[]  []  v   []
t

  

a x, t   v   v   v
t

  

  at   at        at 
  iˆ   iˆ    iˆ  ˆj  kˆ   iˆ
t  x   x   x y z   x 
aiˆ  at    at ˆ
   i
x  x  x  x 
 a ˆ  at  at ˆ 
 i     2 i 
x  x  x 
 a a 2t 2 
   3 iˆ
x x 

The fluid is said to be incompressible if   v  0
        at 
Consider   v   iˆ  ˆj  kˆ    iˆ 
 x y y   x 
  at 
  
x  x 
at
 2
x
as x > 0 and t  0
 
   v  0 The fluid is compressible

V.Imp.
String as a limiting case of system of particles:
Consider finite number of particles connected between two fixed ends. The first
particle and the last particle are at rest always. We assume that these particles
move only in vertical direction and there is absolutely no sidewise deflection.
The amplitude of vibration is assumed small for all the particles. There is
constant separation between the particles (They are equally spaced.)

Let  be the constant tension with which the particles are acted upon by
adjacent particles. Let (N+2) be the total number of particles with first and the
last at rest. The distance between the first and last particle i.e. the length of the
string is, (n+1)h = L.

Let mass of each particle be same say „m‟ thus the equation of motion for the
system of N particles is given by
 2u j  u j 1  u j   u j  u j 1 
m 2        
t  h   h 

Where uj is the displacement of jth particle along vertical direction i.e. u(jh,t) =
uj(t).

In this case linear mass density is

mn  1 m

n  1h h
The above equation of motion can be simplified as
 2 u j    1  u j 1  u j u j  u j 1 
    
t 2   h  h h 
We have in the limit h becoming very small
 u   u j 1  u j 
 x   1   lim  
h 0
 j h  h 
 2

 u   u j  u j 1 
 x   1   lim  
h 0
 j  h  h 
 2
 2u j  

 1  u   u  

      
t 
 h  x   j  1  h  x   j  1  h 
2

  2  2 

  
  1  u   u  
 lim     
 h  0  h  x   j   h  x   j   h 
 1   1 
   2  2   
 u j   u 
2 2
  2
t 2   x  jh
 2u 2  u
2
 c
t 2 x 2
In the limiting case when h 0 or the number of particles tends to infinity, we
get the same differential equation with continuous distribution of mass alon g
the string.

Conservation of energy
We have Euler‟s equation
  
dv
 V  pV  fV
dt

d
dVv   pV  fV  0
dt 
But V  m (constant) taking dot product with fluid velocity v we get,
 d     
v  Vv   v  pV  v  fdV  0
dt
d 1     
  Vv 2   v  pV  v  fV  0 (1)
dt  2 
Consider
d
 pV   dp V  p d V 
dt dt dt
p    
 V  v  pV  p  v V
t
  p  
 v  pV   pV   V  p  v V
d
(2)
dt t

Generally the body force acting on the fluid is gravitational force. Therefore,
 
f V  m g

  V g

Any vector field can always be expressed in the form of a scalar potential
function.
For e.g. electric field is given by,
 
E  V
Where V – scalar potential function

Similarly the intensity of gravitational field can be expressed in terms of

gravitational potential function G.
 
 g  G
Where, G is gravitational potential function

Consider
 
 v  fv
 
 v  VG
 

  v  G V
We have, by definition

dG G  
  v  G
dt t
  G dG
 v  G  
t dt
    G dG 
  v  f V     V (3)
 t dt 
Substituting form equation (2) and (3) in equation (1) we get,

d 1 2 p    G dG 
 Vv    pV   V  p  v V  
d
  V  0
dt  2  dt t  t dt 

d 1  p   G
 Vv  pV  pVG   V  p  v V  V
2

dt  2  t t

Now,
G
a)  0 Because, the gravitational potential function does not depend
t
explicitly on time
p
b)  0 Generally, we consider steady flow, therefore the fluid pressure may
t
change from place to place but at a given point pressure is independent of time.
 
c)   v  0 For incompressible fluids
1
  Vv 2 is the K.E. of the fluid element V , V is the P.E. of the element due
2
to pressure and Gδδ is the gravitational potential energy of the fluid.
Thus, for steady and incompressible fluid flow we have,

d 1 2 
 v  p  G  SV  0
dt  2 
1
 v 2  p  G  cons tan t
2
This is the conservation of energy principle. This itself is known as Bernoulli‟s
equation for ideal fluid.

Ex. 2001

Moving Co-ordinate System

1) Moving origin of co-ordinates:-

By triangle law of vector

  
r  r*  R

Differentiating with respect to „t‟

   
dr dr * dR
 
dt dt dt

Differentiating with respect to „t‟

  
d 2r d 2r * d 2 R
  2
dt 2 dt 2 dt
  
a  a   aR

If starred frame is with uniform (constant) velocity with respect to unstarred

frame then,

aR  0
 
a  a
 
ma  ma 

ma  F Newton‟s second law of motion

 ma  F

An inertial frame is defined as the frame in which Newton‟s laws of motion are
valid. All the frames moving with constant velocity with respect to each other
are all equivalent inertial frames. This is Newtonian principle of relativity.

If a R  0
Then in (S) frame
 
ma  F Newton‟s laws of motion

But in (S*) frame

 
ma   F Newton‟s laws not applicable

S* is called as non-inertial frame of reference because it has acceleration with

respect to S frame.


 
ma  m a *  a R
  
ma   ma  ma R
  
ma   F  ma R

Imp.
1) Rotating co-ordinate system

In (S) frame,

Position vector of p is r  xiˆ  yˆj  zkˆ
In (S*) frame,

r  x  iˆ   y  ˆj   z  kˆ  (Position vector need not be r  since it is distance from
origin and O and O’ always coincide)

iˆ  , ˆj  , kˆ  are functions of iˆ,ˆj, kˆ (since they are rotating is any arbitrary direction)
Our aim is to relate the velocities of the particle as measured in two frames S
and S*

By definition,
  
dB  Bt  Δt   Bt  distance coveredin time Δt
 lim   =
dt t 0 Δt  Time taken Δt

But in same time t, vector will travel a different distance in in two different
frames.

d
   velocity with respect to S
 dt 
 d 
   velocity with respect to S*
 dt 


dr d ˆ ˆ

dt dt

xi  yj  zkˆ 
 dx   dy   dz 
  iˆ    ˆj   kˆ
 dr   dt   dt 

In S* frame,


dt

dt

d r d   ˆ

x i  y  ˆj   z  kˆ 

 d   d  y  ˆ  d  z   ˆ
  x  iˆ     j   k
 dt   dt   dt 

Consider, velocity of a particle of position vector r with respect to (S) frame.

But the position vector r is expressed in terms of co-ordinates of (S*)frame. In
this case, even the unit vectors iˆ  ˆj  etc. are not at rest with respect to (S)
frame.
Therefore

dr d  ˆ 

dt dt

x i  y  ˆj   z  kˆ  
d  d  dy ˆ 

 dˆj   dz  ˆ   dk  
 x iˆ   x   iˆ    j  y     k  z   
dt  dt  dt  dt  dt  dt 
 dx     dy     dz   ˆ     diˆ  ˆ
 dj  dk 

  ˆ
i   ˆ
 j   k    x y z 
 dt   dt   dt    dt dt dt 

d r diˆ  diˆ  dk 
 x  y  z
dt dt dt dt
The time derivatives of unit vectors of S* frame with respect to (S) frame can be
obtained as follows:

Let vector B be any vector at rest in the frame S*. It‟s time derivative with
respect to (S) frame is as usual given by
  
dB  Bt  t   Bt 
 lim  
dt t 0 t 
  
B  Bt  t   Bt 
 
B  B sin  t

B
 B sin 
t

Even in direction it becomes


B  
B
t  
B dB  
lim  B
t 0 t dt

B is supposed to be at rest in S* frame. Similarly, unit vectors iˆ  , ˆj  , kˆ  are at
rest in S* frame. Therefore, we have,
 
d ˆ
dt

i    iˆ 

 
d ˆ
dt

j    ˆj 

 
d ˆ
dt

k    kˆ 

Using this result, we get

 
dr d  r
dt

dt
    
  
 x    iˆ   y    ˆj   z    kˆ 


d r 
dt
 
   x  ˆj   y  ˆj   z  kˆ 
 
dr d  r  
  r
dt dt
Velocity of particle with respect to (S) frame is usual to the sum of the velocity
 
of the same particle with respect to S* frame and   r

The acceleration of a particle with respect to the frames (S) and S* frame and
relation between them can be obtained as follows:
 
d 2 r d  dr 
  
dt 2 dt  dt 

d  d r  
   r using (1)
dt  dt 

d  d r  d  
      r 
dt  dt  dt
   
 d   d  r   d  r  d   dr
        r  
 dt  dt  dt  dt dt
   
d  r  d  r  d        d  r   
 2        r      r
dt dt  dt   dt 
    
d 2 r  d  r d   d r   
    r        r 
dt 2 dt dt dt
   
d 2 r d 2 r  d  r    d  
  2       r   r
dt 2 dt 2 dt dt

This equation gives a relation between accelerations of a particle with respect to

the frame (S) and acceleration with respect to frame S* where S* rotates with
angular velocity ' ' with respect to (S) frame. The above relation is called
coriolis theorem‟
  
    r  is called centripetal acceleration

d r
2  is called coriolis acceleration
dt

d 
 r is called has no specific name
dt

It is non-zero when even rotating frame has acceleration.

Remarks:
  
1) C.P. acceleration      r 

    
    r      r sin 90
  r sin  
  2 r sin 

d r
2) Coriolis acceleration. = 2 
dt

Carioles acceleration is present only when particle has non-zero velocity and
not parallel to axis of rotation with respect to S* frame.

d 
3) The term r
dt
It is present only if S* has angular acceleration with respect to S frame.

Equation of motion of a particle near the surface of earth [Is an

earth‟s frame is inertial frame:]
Co-ordinate system fixed with respect to earth is non inertial frame because
earth rotates about an axis passing through south and north poles with
constant angular velocity.
Consider a particle of mass „m‟ experiencing force F in addition to the
gravitational force acting on the particle. If S is an inertial frame fixed with
respect to star then according to Newton‟s law of motion we have,
 
d 2r 
m 2  F  mg (1)
dt
According to Coriolis Theorem the relation between acceleration of a particle in
(S) and (S*) frame is given by,
   
d 2 r d 2 r     dr d 
      r   2   r
dt 2 dt 2 dt dt

Where S* is a rotating frame fixed with respect to earth and origin

corresponding with centre of the earth. It is rotating with constant angular

d
velocity .  0
dt
  
d 2 r d 2 r     dr
 2       r   2 
dt dt 2 dt

Multiplying both sides by

  
d 2r d 2 r     dr
m 2  m 2  m    r   2m 
dt dt dt

Substituting in (1) we have,

  
d 2 r     dr 
m 2  m    r   2m   F  mg
dt dt
2  
      d r
m 2  F  mg  m    r   2m 
d r
dt dt
2 
d r
Where is acceleration of the particle with respect to the S* frame is earth‟s
dt 2
frame.
This is the required equation of motion near the surface of the earth.
  
m    r  is called centrifugal force

  d r 
2m    is called carioles force
 dt 
[diagram]
  
Centrifugal acceleration =      r    2 r sin 
At equator,  = 90
 2  2
2 2
 
Centrifugal acceleration is maximum   r    r 
2
 R  3.37cm / s
2

 T   24  3600 
At poles  = 0 Centrifugal acceleration =0 and no centrifugal force
  
d 2 r      d r
m 2  F  mg      r   2m 
dt dt
  
  d r
 F  mg eff  2m 
dt

At poles g eff max  g

 

At Equator g eff max  g      r   g  ω2 R

    

This is the reason why the earth if flattened at the poles and bulging at the
equator.
Let  be the angle between real vertical and apparent vertical.

tan   
Ηorizontal component of centri fugal acceleration

Vertical component of accelera tion
 2 r sin  cos

g -  2 r sin 2 
 2 r sin  cos
  2nd term is vanishingly small
g
 2 r sin 2

2g

 m ax   
2
r
 = 45
2g
 0 6‟

 d r
The coriolis force 2m  is effective whenever particle on the surface of the
dt
earth has non-zero velocity with respect to the earth.
There are three observable effects of coriolis force:
1) A particle dropped vertically down is observed to shift towards the east
side due to coriolis force.
2) Air mass rushes to an area of low pressure. This air mass is observed to
shift towards right due to coriolis force. This leads to formation of
Cyclones. They are anticlockwise is northern hemisphere and clockwise
in southern hemisphere.

3) Plane of oscillation of a simple pendulum is observed to rotate due to

coriolis force e.g. Foucalt‟s Pendulum

Problems:

1) Calculate the magnitude of Corilolis Acceleration of a particle of moving

with velocity 200m/s at an angle of 60 with the angular velocity w of
rotation of Earth.

 d r
Corilolis Acceleration  2 
dt

d r
 2 sin 60
dt
  2 
  7.273  105 rad/dec   
 T 
Coriolis Acceleration  2  7.273  10 5  200  sin 60
 2.52cm / s 2

2) Calculate the centripetal acceleration of a particle at a place where co-

latitude is 30 and radius of earth 6400km

Centripetal Acceleration
  
    r 
  2 r sin 
  
2
 7.273105  6.4 106 sin 30 
 1.69cm / s 2

V.Imp. 3) A Co-ordinate system S* is rotating with respect to an inertial co-

ordinate system S. The two systems have the same origin and the angular
velocity of S* relative to S is given by

  2iˆ  t 2 ˆj  tkˆ

The position vector of a particle in the system S is given by r  tiˆ  ˆj  t 2 kˆ
Find
(1) The centrifugal acceleration of the particle in S* at time t = 1sec.
(2) Find the corilolis acceleration of a particle in S* at t = 1 sec.
All quantities are in S.I. system of units
  
(1) Centrifugal acceleration      r 
iˆ ˆj k
 
 r  2 t2 t
t 1 t2
    
 iˆ t 4  t  ˆj 2t 2  t 2  kˆ 2  t 3 
At t = 1 sec.
  ˆj2  1  kˆ2  1
  ˆj  kˆ
i ˆj kˆ
  
Centrifugal acceleration     r   2 t 2 t
0  1 t1
 iˆ1  1  ˆj2  0  kˆ 2  0
 2iˆ  2 ˆj  2kˆ

 2 iˆ  ˆj  kˆ 
 
     r   22   22  22  12  2 3m / s 2

2) Corilolis Acceleration

 d r
2  as  4iˆ  6 ˆj  2kˆ
dt
 ˆ ˆ 2ˆ
r  ti  j  t k
  
dr d  r  
  r
dt dt
 
d  r dr  
   r
dt dt

dr d ˆ ˆ 2 ˆ
dt dt
 ti  j  t k 
 iˆ  2tkˆ

d r
dt
  
 iˆ  2tkˆ   ˆj  kˆ 
 iˆ  ˆj  2t-1kˆ

d r ˆ ˆ
 i  j  2t-1kˆ
dt

 iˆ ˆj kˆ
 d r
ω  2 1 1
dt
1 1 1

 iˆ1-1  ˆj2  1  kˆ2  1

 -ˆj  kˆ

 d r
 2 
Corilolis force dt
 2 ˆj  2kˆ

Foucault‟s Pendulum: -

Consider a simple pendulum with a heavy bob and infinitely long string.
Equation of motion of this single particle with respect to earth is given by
 
d r    d r
m 2  T  mg eff  2m  2
dt dt
T is the tension in the string.

Tension and g eff are always in vertical plane but the horizontal component of

 d r
Coriolis force  2m  2 is non-zero even though very small. There is no other
dt
force is horizontal direction. Therefore, its effect is always seen as slow rotation
of plane of oscillation of the simple pendulum.

In order to find the angular velocity of the plane of oscillation, we

introduce new rotating co-ordinate system S ‟ with respect to the earth such that
S‟ rotates with constant angular velocity (  k̂ ) about vertical axis passing
through point of support. The angular velocity  is so adjusted that all the
forces acting on the bob are in the vertical plane. Therefore, plane of oscillation
in S ‟ will remain fixed. In this can then the angular velocity of plane of
oscillation with respect to earth will be the same as the angular velocity of S ‟
with respect to earth.
We have,
 
d  r d r 
  kˆ  r
dt dt
2   
d r d 2 r
dt 2
 2  2k 
dt
ˆ d r
dt


 kˆ  kˆ  r 
Third term is zero since S ‟ frame is rotating with constant angular velocity.
Substituting,
  
md  2 r
dt 2
ˆ
 2mk 
d r
dt
2 ˆ
ˆ 
  
 m k  k  r  T  mg eff  2m  
  d r
 dt

 kˆ  r 

2  
d r  
m 2  T  mg eff  2mkˆ 
dt
d r
dt
 

 m 2 kˆ  kˆ  r  2m 
 d r
dt



 2m  kˆ  r



 

 T  mg eff  2m kˆ    
 d r
dt
    
 
    
 
 m 2 kˆ kˆ  r  r kˆ  kˆ  2m kˆ  r   r   kˆ
 
In the above equation T and m g eff , last two terms are also is the vertical plane

as k̂ is in vertical plane and position vector r is in the vertical plane. For small

amplitudes
d r
dt
is in the horizontal plane. If the bracketed term k̂   is also  

d r
in the horizontal plane, then the cross product of this term with will be in
dt
the vertical direction.

Thus, we have k̂    in the horizontal plane. This is possible if the dot
product of this term with k̂ is zero.
 

kˆ    kˆ  0
   cos  0
   cos

At the North Pole, co-latitude   0

    cos

In other words, plane of oscillation with simple pendulum will rotate with
period of 24 hours whereas at a place on equator  = 90    0 , therefore
plane of oscillation will not rotate. Period will be infinite.

Problem:
A particle is released near the earth‟s surface in Northern hemisphere. Obtain
the expression for eastward deflection of the particle.

Solution:
[Diagram]

We select Z* along the vertical direction and Y* is the horizontal plane pointing
North. Therefore X* will be towards east.  has components only along Y* and Z*
direction ( lies in Y*Z* plane)

  0,  sin  ,  cos 

A particle of mass is released along Z* axis near the surface of the earth.
Approximately, it has velocity along vertical direction only. But due to coriolis
force, the particle is deflected along the X* i.e. along eastward direction.

 d r
Coriolis force =  2m 
dt
   d  z ˆ
 2m  sin ˆj   coskˆ   k
 dt 
 d  z ˆ
 2m sin   i
 dt 
d 2 z
For particle falling down  g
dt 2
dz 
Integrating,   gt  c
dt
dz
But at t=0 0c0
dt
dz 
  0  gt
dt

Carioles force = 2m sin  (-geff t) iˆ

Thus net Coriolis force is towards east and therefore particle is deflected
eastward.

By Newton‟s Laws of motion

d 2 x
 2 sin g eff t
dt 2
dx  t2
Integrally  2 sin geff c
dt 2
dx 
At t=0 0c0
dt
dx 
  eff sin  t 2 
dt
Integrals
1
x   g eff t 3 sin   c
3

At t=0; x*=0  c=0

1
 x   g eff t 3 sin 
3
Where, t is the time taken by the particle to reach the ground (surface of earth).
It is called time of fall.

To find t:
dz 
We have   g eff t
dt
1
Integrally z    g eff t 2  c
2
At t=0, z = h height through which the particle is released
* c  h
1
z    g eff t 2  h
2
1
 z   h  g eff t 2
2
At t = time of fall when the particle reaches the ground z*=0
1
 0  h  g eff t 2
2
1
 2h  2

t   
g 
 eff 
Substituting this is expresser of x * we get,
3
1  2h  2
x    g eff sin 
3  g eff 


If - is latitude the co-latitude  = 90 – 

3
1  2h  2
x  
  g eff cos
3  g eff 


Larmor‟s Theorem:
1) There is similarity between Coriolis force and magnetic force on a moving
charge particle.
 
dr  d r
q  B and 2m 
dt dt
Both these forces depends on velocity of the particle. Therefore, the effect of
magnetic field can be simulated by introducing rotating frame.

2) Lorentz magnetic force on a moving charge particle is equivalent to rotating

co-ordinate system. This equivalence is called Larmor‟s Theorem.

Statement:
Consider a system of charged particles having the same specific charge (i.e.
q/m same for all the particle). It is acted upon by there mutual forces and by a
common central force towards a common centre. In addition to these forces, the

system is subject to weak uniform magnetic field B . It‟s possible motions will
be the same as the motions it could perform is absences of magnetic field
superimposed upon slow processional motion of the entire system about the
 q 
centre of force with angular velocity given by    B
2m
This is known as Larmor‟s Theorem.

Proof:
For simplicity, consider a system of particles having the same charge and same
mass for the entire particle. The equation of motion for the system of particles
is

d 2 rj  
m 2  F ji  F jc j =1,2, …,N (1)
dt

Where F ji is the internal mutual interaction force acting on the j th particle due

to other particles in the system and F jc is a common central force acting
towards common centre.
When magnetic field is applied, the equation of motion becomes,
 
d 2 rj i c dr 
m 2  Fj  Fj  q  B (2)
dt dt
[Remark (3) we introduce a rotating frame S* rotating with constant angular
velocity such that the equation number 2 reduces to

d 2 r j  i  c
m  Fj  Fj (3)
dt 2

Thus, the possible motions of the particles as described by the equation 2 and
3 are the same. Thus, the possible motions in presence of magnetic field would
be the same as in the absence of magnetic field but in the rotating co-ordinate
system]
We have
  
dr d  r  
  r
dt dt
  
d 2 r d 2 r     d r
      r   2 
dt 2 dt 2 dt

substituting in 1
  
d 2 r j i c  d rj    
 m    r j   F j  F j  q
 d rj   
m  2m     rj   B
dt 2 dt  dt 
 
2   
i c d rj  
m
d rj
 F j  F j  2 m
 d rj
  m   
  
  r j   q  B  q 
 
 r j   B
dt 2 dt dt


 
     
 2m  qB  m    r j   qB    r j 
d rj    
 F ji  F jc 
dt
The constant angular velocity of rotating frame is so adjusted, that the
bracketed term on R.H.S. becomes zero.
 
2m  qB  0
 
 2m  qB
 q 
   B
2m

d 2 ri i c   
m  F j  F j  m  
  
   r j   qB    r j 
dt 2
   q    q       q   
 F ji  F jc  m B    B  r j   qB    B  r j 
 2m    2m     2m  
2 

   
d rj   q   
2
q   
2
m 2
 F ji  F jc  B  B  rj  B  B  rj
dt 4m 2m

d 2 r j
 
 i  c q2   
m  F j  Fj  B  B  rj
dt 2 4m

Since the applied magnetic field is weak

 
  
B  B  r j  B 2 r j sin   small
B2 is negligible in comparison with the other forces.
Thus, we get

d 2 r j i c
m  F j  Fj (3)
dt 2
Hence, the Larmor‟s theorem is proved.

Degree of freedom
The minimum number of independent parameters or co-ordinates required to
describe the system completely is known as the degree of freedom of the
system.
1) A particle free to move in free space: the degree of freedom is three.
2) A rigid body: the degree of freedom is six – three co-ordinates are required to
specify the C.M. of the rigid body and three more co-ordinates are required to
specify the orientation of rigid body.

The degree of freedom n of a system of N particles is given by; n  3N  k where,

k is number of constraints.

A constraint is certain restriction on the motion of particle. We consider only

those constraints, which can be expressed in the forms of relation between the
co-ordinates of system. Such constraints are known „Holonomic constraints‟
Consider e.g. degree of freedom of a system consisting of 3 particles with the
constraint that the distance between two particles is always constant. The
degree of freedom of such a system is six.

r12 = constant, r23 = constant, r31 = constant

 n  33  3  9  3
n6

Extending this to a rigid body consisting of very large number of particles with
the same constraints that the distance between any two particles is always
constant, the degree of freedom of the rigid body is therefore six (rij = constant)

Generalised velocity
Time derivative of generalised co-ordinate is called the corresponding
generalised velocity. Generalised velocities are denoted by
q1 , q 2 ,    q n 
The generalised velocities of a particle freely moving in three dimension is
   
r,,  or  , , z

Generalised momentum
In Cartesian co-ordinates, the kinetic energy of a particle in 1-dimension is
given by
1
K  mx 2
2
The ordinary momentum of a particle is defined as
p  mx
This can be expressed as the derivative of K.E. K with respect to x
K
 p  mx 
x
On this line, generalised momentum corresponding to the generalised co-
ordinate q can be defined as:
K q , t 
p
q
The Cartesian co-ordinates and the generalised co-ordinates of a particle in one
dimension is related by the equation:
x = x[q,t]
& q = q [x,t]

e.g.
x = r sin  cos
y = r sin  sin
z = r cos
i.e.
x = x [r,,  ]
Also, r  x 2  y 2  z 2 , r = r[x, y, z]
 K q , t 
P
q
(K is a function of x and x is a function of q )
K  x,t  x

x q
x
P p
q
But x  xqt , t  (explicit dependence on time)
x x
 x   q  (*)
q t

Differentiating with respect to q

x x

q q

We always treat q and q as independent variables

A) F  xt , t 

dF F F
 x 
dt x t
x
P  p
q
Differentiating with respect to „t‟
x d  x 
 P  p  p  
q dt  q 
We have,
d  x    x    x 
 qt , t      q    (using *)
dt  q  q  q  t  q 

  x    x 
  q    (A)
q  q  q  t 

  x x 
  q  
q  q t 

 x 
q
x x
 P  p p
q q
Using Newton‟s 2nd law of motion (rate of change of momentum of a particle is
directly proportional to impressed force)

P  F
x x
 P  F p
q q
x  K  x 
F   
q  x  q 

x k
P  F  (Composite rule)
q q
 x
Generalized force in 1-dimension is defined as F
q

[Conservative force field: If the work done by the force is independent of path
followed by the particle then the force is said to be conservative force.]

If the force is conservative then it can always be expressed as gradient of a

scalar function. In one dimension, the one can write
x V
Let F   Q
q q
Where Q’ is non-conservative generalized force. It cannot be expressed as
gradient of a Scalar potential. [e.g. Friction is non-conservative force]
V K
P     Q
q q

 K  V   Q
q
d
P    K  V   Q 
dt q
d  K  
   K  V   Q 
dt  q  q
Generally in most of the cases potential is independent of velocities. It depends
only on co-ordinates. In that case,
Consider,

K  V   K  V
q q q
K
 0
q
K

q
d 
 K  V    K  V   Q
dt  q  q

Let L  K  V
L is called as Lagrangian function defined as the difference of K.E. and P.E.

d  L  L
   Q
dt  q  q
It is the Langrangian equation in 1-dimension. In most of the physical
situations, the particle does not experience any non-conservative forces.
Therefore in absence of non-conservative forces Q’ = 0, the Lagrange‟s equation
in one dimension becomes
d  L  L
  0
dt  q  q

N particle no constraints
xi  xi q1 t q 2 t     q3 N t , t 
y i  y i q1 t q 2 t     q3 N t , t 
z i  z i q1 t q 2 t     q3 N t , t 

xi  xi q k t , t  k = 1,2,3…3N
y i  y i q k t , t  k = 1,2,3…3N
z i  z i q k t , t  k = 1,2,3…3N
q k  q k xi t , y i t , z i t , t 

xi x
x i   q h  i Where we have N equations
k q k t

K
1

 mi xi2  y i2  zi2
2 i

K
p xi 
x i
K
Pq k
q k
K x i K y i
 
i x i q k i y
 i q k

Lagrange‟s equation in several dimensions

Generalized {q1, q2 … q3N ,t}  {qk, t} k =1,2…3N

Cartesian {x1, y1, z1, x2, y2, z2…xN, yN, zN, t}  {xi, yi, zi, t}; i =1,2…N

1

K   mi x i2  y i2  z i2 
i 2

K 1
  mi 2 x i  mi xi  pix
x i 2

We have,
xi  xi q k t , t 

Similarly for yi and zi

q k  q k xi t , y i t , z i t , t 

K  K x i , y i , zi , t 
K  K q k , q k , t 

Definition
K
Pk  x i , y i , zi 
q k
 K x i K y i K zi 
Pk      
 x i q k y i q k zi q k 
x i x
Use  i (Similarly for yi and zi)
q k q k

 K  xi  K  y i  K  z i 
Pk           
 i  k  i  k  i  q k 
x q y q z
 x y z 
Pk    pix i  piy i  piz i 
i  q k q k q k 

Differentiating with respect to „t‟

  x  d  x   y  d  y   z  d  z 
Pk    p ix  i   pix  i   p iy  i   piy  i   p iz  i   piz  i 
i   q k  dt  q k   q k  dt  q k   q k  dt  q k 
Using Newton‟s law of motion
mi xi  Fix  mi x i   p ix
d
dt
  x   y   z   d  x  d  y  d  y 
Pk    Fix  i   Fiy  i   Fiz  i     pix  i   piy  i   piz  i 
i   q k   q k   q k  i  dt  q k  dt  q k  dt  q k 
d  xi x i
Use   
dt  q k
 q k
  x   y   z   x y z 
Pk    Fix  i   Fiy  i   Fiz  i     pix i  piy i  piz i 
i   q k   q k   q k  i  q k q k q k 

  x   y   z   K  xi   K  y i   K  zi 

 Pk    Fix  i   Fiy  i   Fiz  i              
i   q k   q k   q k  i  x i  q k   y i  q k   zi  q k 
  x   y   z  K
Pk    Fix  i   Fiy  i   Fiz  i  
i   q k   q k   q k  q k

Recall
K  K x i q k , q k , t , y i q k , q k , t , zi q k , q k , t 

K  K   x i   K   y i   K  zi 

              
q k  i   q k
i  x   y i   q k   zi  q k 

The forces acting on the i th particle can be conservative or non-conservative.

Conservative forces can be expressed as negative of gradient of Scalar potential.
The non-conservative part is left as it is.

We have,
V
Fix    Fix (Similar Fiy and Fiz)
xi
 V  xi   V  yi   V  z i    x   y   z  K
Pk                    Fix  i   Fiy  i   Fiz  i  
 xi  q k   yi  q k   z i  q k  i   q k   q k   q k  q k

V K
  Qk 
q k q k
Where Q’k is a non-conservative generalized force corresponding to generalized
co-ordinate qk
It is defined as:
 x y z 
Qk    Fix i  Fiy i  Fiz i 
i  q k q k q k 
K V
Pk    Qk
q k q k

 K  V   Qk
q k

 Pk   K  V   Qk
d
dt q k
d  K  
  K  V   Qk
dt  q k  q k

Potential V is a function of only qk

V=V[qk(t),t]

In most of the situations potential is independent of generalized velocities

V
 0
q k
K 
  K  V 
q k q k
d    
   K  V   K  V   Qk
dt  q k  q k

Let L = K–V  Lagrangian function

L  Lq k t , q k t , t   K  V
d  L  L
   Qk k = 1,2…3N
dt  q k  q k

If the particles do not experience any non-conservative force i.e. Q’k = 0, then
the Lagrange‟s equations become
d  L  L
   0 k = 1,2…3N
dt  q k  qk

Imp.
To prove,
x i xi

q k q k
Proof we have,
xi  xi q k t , t  k = 1, 2…3N

d
xi   x i   xi  d q k   xi
dt k q k  dt  t
 x   x 
x i    i q k   i 
k  q k   t 
Differentiating with respect to q k

(Generalised co-ordinates and generalised velocities are always treate d as

independent variables)
qk = k=1,2,…3N

In other words when partially differentiating with respect to q k , q k is kept

constant.

 x   xi 
  i    1  0
 q k   q k 

x i x
  i
q k q k

d  xi  x i
2)  
dt  q k  q k
We have,
xi  xi q k t , t 
 xi x
  i qk t , t 
qk qk

Differentiating with respect to time,

d  xi    xi     x 
      ql    i 
dt  q k  ql  q k  t t  q k 
  xi    x 
  q l   i 
l q l  q k  t  q k 
  xi x 
  q l  i 
q k  l ql t 
x
 i
q k

1) Atwood‟s Machine: -
Obtain the Lagrange function for Atwood‟s Machine and hence write the
equation of motion: -
(Obtain the acceleration of the system)

Constraints: -
1) y1 = 0
2) z1 = 0
3) y2 = 0
4) z2 = 0
5) l is constant (x1+x2 = l)
n  3(2)  5  1
 Degree of freedom of Atwood‟s machine is one. Hence one generalized co-
ordinate is required to describe the motion.
qk = x

Kinetic Energy:
1 1
K  m1 x 2  m2 l  x
2 2
 
2

Potential Energy:
V  m1 gx  m2 g l  x 

L  K V

2
1
2
2
 
 m1 x 2  m2 l  x  m1 gx  m2 g l  x 
1

Lagrange‟s equation:
d  L  L
 0
dt  x  x
L 1

 m1 2 x   m2 2 l  x  1
x 2
1
2


 m1 x  m2 l  x 
 m x  m x
1 2  l  0  
 m1  m2 x
L
 m1 g  m2 g
x
 m1  m2 x   m1 g  m2 g   0
d
dt
d
m1  m2 x   m1 g  m2 g  0
dt
m1  m2 x  m1  m2 g
x 
m1  m2 
g
m1  m2 
Note:
In Lagrange‟s Equation the force of constraints does not appear. Hence directly
one can not find the force of constraint. In order to determine the force of
constraints we assume that the constraints on motion is violated for small
displacements. This way, it is possible to determine the force of constraints by
writing the Lagrange‟s equation corresponding to the generalised co-ordinates
violating the constraints.

2) Obtain the Lagrange function for simple pendulum and write equation of
motion.

Number of constraints:
1) It can oscillate only in XY plane  z = 0
2) Length is constant. l  x 2  y 2
Degree of freedom = 3(1) – 2 = 1

Generalized co-ordinate q k  
 x  l sin 
x  l cos
y  l cos
y  l sin 

Kinetic Energy:
1

K  m x 2  y 2
2

1

 ml 2 θ 2 cos2 θ  sin 2 θ
2

1
 ml 2 θ
2

Potential Energy:
V  mgh
 mg OA-OB
 mg l  l cosθ 
V  mgl1  cosθ 

L  K V

ml   mgl 1  cos 
1 2 2

2

Lagrange‟s Equation
d  L  L
 0
dt    
L
 ml 2
 
L
  mgl sin 

d
 
 ml 2  mgl sin  0
dt
ml 2  mgl sin  0
g
  sin   0
l
g
    sin 
l
Since  is very small sin   
g
   
l

Constraints:-
A force of constraint puts some restrictions on the motion of the
particle. In the usual Lagrange‟s method, the constraint on the motion is taken
into account and accordingly only the necessary co-ordinates are considered.
By this method, it is not possible to determine the force of constraint. In order
to find the force of constraints, Lagrange‟s function is modified to take into
account that co-ordinate which is affected by the force of constraint. By writing
the Lagrange‟s equation for this co-ordinate, it is possible to obtain
corresponding force of constraint. This method is illustrated with the help of an
example of simple pendulum.

The degree of freedom of simple pendulum, with the constraint that the
distance of bob from the fixed support is always constant, is one. To determine
the force of constraint we introduce extra co-ordinate „r‟ which takes constant
value r = l when constraint is imposed.
x  r sin θ
x  r cosθ θ  r sin θ
y  r cosθ
y  r sin θ θ  r cosθ

K 
1
2

m x 2  y 2 

1
2
 1
 
m r sin θ  r cosθ θ  m r cosθ  r sin θ θ
2

2

2

1 1
 mr 2  mr 2 θ 2
2 2
1

 m r 2  r 2 θ 2
2

The reference point for potential is taken as the point of support.

V   mgy
V   mgr cos
As force of constraint is acting on the particle there is corresponding P.E. of the
particle. This potential is minimum whenever the constraint i.e . r = l is
satisfied. Let the potential due to this constraint force be Vr constraint

Thus the modified Lagrange‟s function is

1
 
L  r 2  r 2 2  mgr cos  Vr
2
constraint

The Lagrange‟s equation for the radial co-ordinates is given by

d  L   L 
 0
dt  r  r
 2 Vrconstraint 
d
mr   mrθ  mg cosθ 
 0
dt  r 
V constraint
mr  mrθ 2  mg cosθ  r 0
r

Due to the deep and narrow potential Vrconstraint only one value of r i.e.
r=l(constant) is allowed.

Then the above equation reduces to

V constraint
 

 ml  2  mg cos   r  Qrconstraint (Similar to E  v )
r

(Diagram)

The direction of this force of constraint is along the string inward if the string is
tight. This gives us the tension in the string. In the equilibrium state
L 
r=l(constant) implies that 0
r

In other words, from the expression for L We get,

L Vr
constraint

 0
r r
 1 
Where L   mr 2 2  mgr cos 
2  r l
Vr
constraint


constraint
Qr
r
L

constraint
Qr
r r l
This way the force of constraint can be determined. So far we have considered
Holonomic constraints acting on the system of particle. Holonomic constraint is
the one that can be expressed as a relation between the co-ordinates of the
system.
 
e.g. Rigid body – The constraint is ri  r j  constant
i.e. the distance between any two particles is always constant. This is relation
between co-ordinates. Hence it is a Holonomic constraint.

Generalization:-
Consider a system of n particles with 3N as the total number of co-ordinates.
Let there be c number of constraints on the motion of the system. These
constraints are assumed to be Holonomic i.e. they can be expre ssed as a
relation between the co-ordinates. Out of 3N co-ordinate first c co-ordinates are
taken as dependant co-ordinate.

f j x k t , t   0 k  1,2...,3N j  1,2...c

In the case of simple pendulum the relation between the co-ordinates can be
written as (r – l = 0)

Out of 3N co-ordinates first c co-ordinate are selected as dependant co-

ordinates. They can be selected as

q j  f j x k t , t   0
i.e. qj = 0 j = 1,2,3…c

These first c co-ordinate gives rise to a potential, which is called potential due
to the force of constraints. This potential due to constraint depends on the first
c coordinates V constraint q1 , q 2 ,    q c , t 
This potential due to constraints has very deep and narrow minimum at
q1 , q 2 ,    qc , t   0
The Lagrange function is given by
L  K  V  V constraint
 L  V constraint

The solution of the possible motion of the system can be obtained by solving the
following equations
d  L  L
  0 j = c+1,c+2…3N
dt  q j  q j

In order to determine the force of constraint then we introduce the first c co-
ordinates and we get,
L

constraint
Qj j = 1,2…c
q j
Simple Pendulum with oscillating support: -
Consider a simple pendulum with both of mass „m‟ attached to one free
end the other end is attached to a support which is not fixed but can oscillate.

(Diagram)

(Degree of freedom for simple pendulum: the degree of freedom is one)

But since support is not fixed, but oscillating it can oscillate is sidewise or
vertically up and down. Thus it has two degrees of freedom. Hence, degree of
freedom in this case is three N = 3N – k
Here N = 2, the bob and 2nd one is oscillating support.

But they can oscillate in vertical plane only

Thus, 1) z bob = 0 2) z s = 0

Also the length of the string is constant

3) x2 + y2 = l2 = constant

Thus n=3(2) – 3 = 3

Three independent co-ordinate xs, ys, 

Kinetic Energy K  mx  x s   m y  y s 
1 2 1 2

2 2
 mx 2  y 2   2 xx s  2 yy s  x s2  y s2
1
2
x  l sin x  l cos
y  l cos y  l sin 
1

 K  m l 2 2  2l cosx s  2l sin y s  x s2  y s2
2

Potential Energy
V  mg  y  y s 
 mg l cosθ  y s 

L  K V


1
2
 
m l 2 2  2 x s l cos  2 y s l sin   x y2  y s2  mg l cos  y s 

Lagrange‟s equation
d  L  L
 0
dt    
L
 ml 2  mx s l cos  my s l sin 

d  L 
   ml 2  mxy l cos  mys l sin   mx s l sin   my s l cos

dt   
L
 mx s l sin   my s l cos  mgl sin 


Lagrange equation
ml 2  mxy l cos  mys l sin  mx s l sin  my s l cos  mx y l sin  my s l cos  mgl sin  0

Divide by ml 2
g y x
  sin   s sin   s cos  0
l l l
  g ys  x
     sin    s cos
l l  l
ys and xs are the accelerations of the support. Let us assume that the support
oscillates along the horizontal direction and there is no motion along vertical
direction.
 Let x s  x 0 cos t (Oscillatory part)
& ys  0 i.e. (No up and down oscillation)
 x s   x0 sin t
xs   xω 2 x0 cosωt
 ω 2 x0 cosωt
ys  0

Equation of motion now becomes

θ  g sin θ   ω x cos ωt cos θ
2

0
l l

Let the amplitude be small   0

 sin    , cos  1
g
& Let     02 Natural frequency of simple Pendulum
l
 x ω2 
θ  ω02 θ   0  cosωt
 l 
This is an equation of forced oscillations. The pendulum performs simple
harmonic oscillations under the influence of periodic force. If the frequency of
oscillation of support is exactly same as the natural frequency of the pendulum;
then the pendulum will oscillate with maximum amplitude.

Constraints
Holonomic Non – Holonomic
1) For rigid body, 1) An insect kept in a rectangular box.
 
ri  r j = constant, j  i = 1,2…N It cannot go out. x  a, y  b, z  c
2) For Simple pendulum 2) An insect kept in a sphere of radius
r = l, i.e. r – l = 0 R
3) A wheel rolling x2 + y2 + z2  R2

Rhenomic Scaleronomic Rhenomic Scaleronomic

Time independent Time dependent Time independent Time dependent
 
ri  r j = constant An insect crawling
on a balloon and
we are passing air
inside the balloon.

Problem:
1) Consider a particle of mass „m‟ moving in a plane and a central force acts
on it. Obtain the Lagrange function and hence write Lagrange‟s equation.

(Diagram)

Origin of the co-ordinate system is taken as the centre of force.

Degree of freedom is 2 since a particle is confined to a plane.

 x  r cos  x  r cos  r sin 

y  r sin   y  r sin   r cos

Kinetic energy
K  mx 2  y 2 
1
2
1
  
 m r cos  r sin   r sin   r cos
2
2

2

1

 m r 2  r 2 2
2

Central force, by definition is given by

F r   F r rˆ
k (Inverse square Law)
  2 rˆ
r
(-ve sign shows the force is inwards)
r
V r     F r dr
0
The potential energy function does not involve the co-ordinate . It depends
only on the distance of the particle from the centre i.e. the co-ordinate „r‟.
L  K V

 
 m r 2  r 2 2  V r 
1
2
The Lagrange‟s equation for two co-ordinates r and  are as follows
d  L  L
  0
dt  r  r
d
mr  mr 2  V r   0
dt r
V r 
mr  mr 2  0
r

d  L  L
  0
dt    
d  L 
 0  0
dt   
L
 constant

mr 2  constant

If Lagrange does not involve particular co-ordinate then the co-ordinate is said
to be cyclic or ignorable co-ordinate. The corresponding generalized momentum
is always conserved.
Thus, cycle co-ordinate always leads to some kind of conservation principle.
If the co-ordinate time is absent in Lagrange‟s function (i.e. cycle coordinate),
then the corresponding conjugate momentum i.e. energy is always conserved.

2) Find the Lagrange function and hence equation of motion of compound

pendulum.

(Diagram)

Degree of freedomn: 3N – k

Constraints (k)
1. z1 = 0
2. z2 = 0
3. l 1 = constant
4. l 2 = constant
n  3(2)  4  2 .

x1  l1 sin θ1
x1  l1 cosθ1θ1
y1  l1 cosθ1
y1  l1 sin θ1θ1

x 2  l1 sin  1  l 2 sin  2
x  l cos   l cos 
2 1 1 1 2 2 2

y 2  l1 cos 1  l 2 cos 2
y  l sin    l sin  
2 1 1 1 2 2 2

K 
1
2
 1

m1 x12  y12  m2 x 22  y 22
2
 
  
 m1 l1212  m2 l1212  l 2222  2l1l 212 cos  1 cos  2  sin  1 sin  2 
1
2
1
2

 m1l1212  m2 l1212  m2 l 2222  m2 l1l 212 cos  1   2 
1 1 1
2 2 2

V  m1 gy1  m2 gy2 

 m1 gl1 cosθ1  m2 gl1 cosθ1  m2 gl2 cosθ 2
L  K V

m1l12 θ12  m2 l12 θ12  m2 l 22 θ22  m2 l1l 2 θ1θ2 cosθ1  θ 2   m1 gl1 cosθ1  m2 gl1 cosθ1  m2 gl 2 cosθ 2
1 1 1

2 2 2

d  L  L
  0
dt  1  1
L
 m1l12 θ1  m2 l 22 θ1  m2 l1l 2 θ2 cosθ1  θ 2 

θ1
d  L 

dt  θ1 
 

  m1l12 θ1  m2 l12 θ1  m2 l1l 2 θ2 cosθ1  θ 2   θ2 sin θ1  θ 2  θ1  θ2  

L
 m1 gl1 sin θ1  m2 gl1 sin θ1  m2 l1l 2 θ1θ2 sin θ1  θ 2 
θ1
 
m1l12 θ1  m2 l12 θ1  m2 l1l 2 θ2 cosθ1  θ 2   m2 l1l 2 θ1  θ2 θ2 sin θ1  θ 2 
 m gl sin θ  m gl sin θ  m l l θ θ sin θ  θ   0
1 1 1 2 1 1 2 1 2 1 2 1 2

m1l121  m2 l121  m2 l1l 22 cos1   2   m2 l1l 212 sin 1   2 

 m2 l1l 212 sin 1   2   m1 gl1 sin 1  m2 gl1 sin 1  m2 l1l 212 sin 1   2   0

1l12 m1  m2   gl1 sin  1 m1  m2   m2 l1l 2 2 cos 1   1   22 sin  1   2   0

Special cases of compound Pendulum

Take m1 = m2= m and l 1 = l 2 = l
Consider small amplitudes i.e. 1 & 2 are very small. The equation are
2ml 21  ml 22 cos 1   2   ml 222 sin  1   2   2mgl sin  1  0

1 and 2 are very small, 1 – 2 is still small i.e.  1   2  0

 cosθ1  θ 2   1 sin θ1  θ 2   0
 2ml 21  ml 22  0  2mgl 1  0
g
21  2   1  0
l

Also,
L
 m2 l 222  m2 l1l 21 cos 1   2 
2
d  L 
   m2 l222  m2l1l21 cos1   2   m2 l1l21 sin 1   2 1   2 

dt   2 
 m 2 l 222  m 2 l1l 21 cos 1   2   m 2 l1l 21 sin  1   2 1  m 2 l1l 212 sin  1   2 

L
 m2l1l212 sin 1   2   m2 gl2 sin  2
2
L
 m2 l1l 212 sin  1   2   m2 gl2 sin  2
 2

Lagrange‟s equation
d  L  L
  0
dt  2   2

m2 l 222  m2 l1l 21 cos1   2   m2 l1l 211 sin 1   2 

 m2 l1l 212 sin 1   2   m2 l1l 212 sin 1   2   m2 gl2 sin  2  0

m2 l 222  m2 l1l 21 cos1   2   m2 l1l 211 sin 1   2 

 m2 gl2 sin  2  0

Special Case:
m1  m2  m, l1  l 2  l
1 and  2 verysmall
 sin  2   2 1   2   0
sin 1   2   0, cos1   2   1

ml 22  ml 21  mgl 2  0

g
2  1   2  0
l

Non –Linear Mechanics (Chaos)

In Newtonian Mechanics, the laws are predictable and the equations of
motions are linear. The solutions of the equations are not very sensitive to the
initial conditions. If there are certain errors in specifying the initial
uncertainties in the solution, there are corresponding predictable uncertainties
in the solution.
There are systems, which are governed by non-linear equations, and for
certain values of parameters of the equation, the solutions are quite
unpredictable. In other words, the solutions are very sensitive to the initial
conditions. Small change in initial conditions makes very large change in
solutions thereby making the output quite unpredictable.
The essential difference between a chaotic system and a non-chaotic one
is the degree of predictability of the motion given the initial conditions to some
level of accuracy.

Example: - Consider two simple harmonic oscillators be driven by some

sinusoidal periodic force. Let the two oscillators start with different initial
conditions but after some time when transients are over both the oscillators
oscillate with same frequency as that of the driver‟s frequency. This is the
example of non- chaotic system.
The well-known example of chaotic system is the weather system.

Note: Even though Brownian motion of gas molecule is random motion, it is not
a chaotic motion. This is because the motion of the gas molecule becomes
unpredictable as initial conditions of the gas molecules are not known. If
somebody provides the initial conditions, then the trajectory of a gas molecule
can be predicted very well.

Anharmonic Oscillator: The best example of chaotic system for a certain

parametric values is the Anharmonic oscillator. The simple pendulum is an
approximation of an Anharmonic oscillator. In this case, the amplitude of
oscillations was assumed to be very small.
Consider an Anharmonic oscillator of mass „m‟ and subjected to a force
F x   k x  x 3 

The P.E. of the oscillator is

x
V x     F x dx
0

  k x  αx 3 dx
x

1 2 x4
 kx  k
2 4
x 2
x 
4
 k   
 2 4 
Following plots give the variation of P.E. V x  as a function of x for different
values of k and  . For simplicity, k is taken as  1 and similarly  is taken as
1.

If  =0 and k is positive then the particle performs usual S.H.O. but if  is

positive then the particle tries to come back to its mean position quickly as if
the restoring force has increased. It is called a hard spring. If  is negative (
<0) then the spring is called soft spring.
In the above diagrams plot (a) corresponds to a single equilibrium position at
x=0. It is stable equilibrium.
Plot (b) also has only one equilibrium position at x=0. But it is possible only for
small values of energy.
Plot (c) has equilibrium at x=0 but it is unstable.
Plot (D) has three equilibrium positions. Two are stable and one at x=0 is
unstable.

At equilibrium position
f x   0
 
 k x  x 3  0
 kx1  x   0 2

x0
or
1
1  x 2  0  x 2  1  x 2  


But in (d)  = –ve

1
x2 

1
x

1 1
x ,
 

Most stable and sustained oscillatory motions are possible for the plots (a) and
(d) shown above.

Consider a particle, in addition to F x  experiences a damping force given by

dx
Fdamping  2m , where,  – coefficient of damping per unit mass of the particle
dt

Let external periodic force be acting on the particle given by,

Fdriving  fm cos t
Where f –amplitude of periodic force per unit mass

Applying the Newton‟s second law of motion, we get,

d 2x
m 2  Frestoring x   Fdamping  Fdriving
dt


md 2 x
2
  dx
 k x  x 3  2m  fm cos t
dt
dt
k k 
2
x  2x   x     x 3  f cost
 m m
 k 
But,     0 = natural frequency of the oscillation

 m 

k
Let 
m
 x  2x   02 x  x 3  f cos t
This is the equation of motion of an Anharmonic Oscillator called as Duffin g‟s
oscillator.
By selecting the scale of time properly, the natural frequency  0 can be
made one. In other words, the time scale is so adjusted that the periodic time of
the particle is exactly 2. Therefore  0 = 1. Similarly, adjusting scale of x
properly  can be made  1.

In this case, the equation becomes

x  2x  x  x 3  f cos t

This is called reduced equation of motion of Duffing‟s oscillator. There are only
three parameters involved, damping coefficient  the external periodic force
frequency and amplitude.

Quadratic Map (Logistic Map): -

Consider the population of certain species. It depends on the food supply
available. If the food supply is unlimited then population goes on increasing.
But if the food supply is finite then population will attain a maximum and
afterwards will decrease to zero. If we normalise the population by dividing it by
the maximum possible population then population is expressed as a fraction .
This fractional population in n  1th year depends on the fractional population
of the previous year non-linearly. It is given by,
x n 1   1  x n x n

This is known as quadratic equation or the logistic map.

xn and xn +1 can take the values between 0 and 1.
Therefore  lies between 0 and 4.

(Diagram)

At maximum
dxn 1
0
dxn max
 1  x n   x n  0
1  2 xn  0
x n  0.5
 xn 1   0.50.5
 0.25

x n 1 
4
0 4

For different values of parameter , the logistic map exhibits various aspects of
chaotic systems. There are three interesting ranges of .

1) 0<<1
2) 1<<3
3) 3<<4.

Consider
1) 0<<1
xn+1 = xn(1 – xn)
Let  = 0.7
n xn

0 0.5
1 0.175
2 1.1011
3 0.636
4 0.417
5 0.0279
6 0.0189
7 0.0129
8 0.0089
9 0.0062
10 0.0043

Thus for 0<<1, population attains a stable value as zero after se veral
iterations. The starting population may be anything but finally it will reach
zero. Analytically it can be shown as follows:-
0<<1.
Let xn 1  x n  x
x  x 1  x 
 x 1   1  x   0
 x0
In this case, x  0 is the fixed point. This fixed point is called as attractor
because the populations in the vicinity are finally attracted towards x  0 .
And,
1   1  x   0
 1  x   1
1
1 x 

1
x  1

1
This is not possible as 0    1   1 . Thus, x is –ve which is not allowed.

Hence, there is only one attractor.

Consider,
2) 1<<3
Let =2.5
xn+1 = xn(1 – xn)
Let x0=0.5

n xn
0 0.5
1 0.625
2 0.5859
3 0.6066
4 0.5966
5 0.6017
6 0.5991
7 0.6004
8 0.5998
9 0.6001
10 0.5999
11 0.6000

From the above analysis, it is obvious that from 1<<3 there is only one fixed
 1
point x  1    0.6 . It is an attractor after several iterations irrespective of the
 
1
starting value of xn finally, it will converge to x  1  . The other fixed point

x  0 is never attained, it is called „repeller‟.

3) Consider =3.2

xn+1 = xn(1 – xn)

Let x0 = 0.5

n xn
0 0.5
1 0.8
2 0.512
3 0.7995
4 0.5129
5 0.7995
6 0.5129
7 0.7995
8 0.5129
Thus for  = 3.2, there are two fixed points which are stable called as attractors.
The population fluctuates between these two fixed values. The third point in
between which is the intersection of straight line and parabola is repeller.

Till  = 3, there was only one stable fixed point. As  increases beyond 3
there are two fixed points. This transition as the value of  is raised passed a
critical value 3 from one fixed point to a pair of stable period–2 points is known
as a bifurcation or period doubling.
In this period doubling phenomenon xn+2 matches with xn i.e. xn+2 = xn

We have,
x n  2  x n 1 1  x n 1 
  x n 1  x n 1  x n 1  x n 
Let x n 2  x n  x say
 x   x1  x1  x  x 2 
2


 x 1  2 1  x 1  x1  x   0 
 x  0 or 1  2 1  x 1  x1  x   0
It is a cubic equation in x. There are three roots. The roots are
 1  1      3 
x  1  ; 1 
   2    1 

The graph for xn+2=F(xn) cuts the region line at three points. The extreme two
points are the stable attractors and the middle one is repeller.

As  increase further, the fixed points keep on multiplying. First, it is period-2

bifurcation then period-4 bifurcation and so on. For  3.57, it is observed that
there is not a single fixed point. The population keeps on fluctuating between
zero and one. It is non-repetitive and non-periodic. We say that the system
exhibits a chaotic behaviour.

Such bifurcations occur faster and faster until an

infinite number of bifurcations occur at  = 3.56994 as shown in the above
graph. Denoting by (k) the critical value of  at which the bifurcation from a
stable period k set of points to a stable period k+1 set occurs it is found that
Lim  k   k 1
 4.669201
k    k 1   k
This number is called Fiegenbaum number. This ratio turns out to be universal
for any map with a quadratic maximum and in seen in a wide range of physical
problems.

V.Imp.
For any mapping function x n 1  F x n  , show that a fixed point x will be stable if

F x   1 where F x  

dF
dx x x

Solution
[Point x is called fixed point if xn 1  x n  x (Constant) for large n]

  1  xn   xn
dF

dx
  1  2 x n 
dF

dx xn  x

1) But 0<  <1

We have x  0
dF
   1
dx x
2) Consider 1    3
1
Then x  1 

dF   1 
   1  21  
dx x    
 2
  1  2  
 
   2
dF
 1
dx x

3) Consider 3<<4
We have fixed points
1     3 
x 1 
2    1 
dF  1     3 
   1  1  
dx x      1 
   3 
   1   1 
   1 
dF
 1
dx x

Consider x n 1  x n 1  x n 
Consider a value of xn, which is very close to x
Let x n  x   n
x n 1  x   n 1 x is a fixed point
Substituting,
x   n 1   x   n 1  x   n 

  x   n  x 2   n x   n x   n2 
Since  n is small  is negligible.
2
n


x   n 1   x n   n  x 2   n x 
 x Is fixed point
Then by definition, x
xn 1  x n  x
 x  x 1  x 
x 1   1  x   0
 n 1   x  x 2   n  2 n x   x
 n1  x  x 2   n  2 n x   x

n n
  
x 1  x  2 n  n   x
  x 
n

 x   1  x   1  2x  n   n

n
 x   1  x   1
  2x  
n
 n 1
    x
n
  1 2 x 

 n 1   1  2 x  
dF
n dx

If n 1  1
n
  n 1   n

In other words as xn approaches the fixed point x within the difference  n , x n 1

also approaches x with difference less than  n . It is means it is a converging
process and after se veral steps x n 1 x n becomes same as x . Thus x must be the
fixed point called as attractor.
 n 1
1
If  n
  n 1   n
As xn approaches x , x n 1 goes away (diverges) further and further from x , it is not
a fixed point. It is called a repeller.

Geometric Interpretation:

Problem 2) For the iterated

quadratic map x n  2  F F x n   F2 x n 
Show that its fixed point satisfy

x 3  2x 2  
   1
x 
2  1  0
   3
Solution:
F2 x n   x n  2  x n 1 1  x n 1 
x n  2   x n 1  x n 1  x n 1  x n 

Let x n 2  x n  x
x   x 1  x 1  x1  x 
x  2 x  2 x 2 1  x  x 2 
x  2 x 1  x 1  x  x 2 

x  2 x 1  x  x 2  x  x 2  x 3 
x  2 x  3 x 2  3 x 3  2 x 2  3 x 3  3 x 4
x  2 x  3 x 2  3 x 3  2 x 2  3 x 3  3 x 4  0
x 1  2  3 x  3 x 2  2 x  3 x 2  3 x 3   0
x 0
1  2  3 x  3 x 2  2 x  3 x 2  3 x 3  0
3 x 3  23 x 2  3  2 x  1  2  0
 1
x 3  2 x 2  1   x 

2  1
0

  3
x3  2x 2  
   1
x 

2  1
0

   3

In addition to x  0 , there are three other cubic roots of above equation. Out of
these three roots one value is
1
x  1


Factorise this and find the other two values of x

Roots of above equation are
x   x   x     0
 1
 x  1  x   x     0
 

x3  2x 2 
  1 x   2  1 
  3 
 
 1   1    1
x 3  1   x 2  1   x 2  
      

Fractal Dimensions:
The general objects such as point, line, square, cube has integral dimensions.
But, there are some geometric constructions for which the dimensions are
fractional and not integral.

In general, dimension d of a set of points in „p‟ dimensional space is defined as.

ln N  
d  lim
 0
 
ln 1
When N( ) is the number of p dimensional cubes of side  needed to cover the
entire set. This calculation of dimension is explained as follows
1) Consider a point only point cube is of side  requiring filling the given point.
 N    1
 ln N    0
d 0

2) A line of length ‘l’. This line of length „l’ can be filled with small line segments
of length 

[Diagram]
 N    l

 d  lim
 
ln l
 0
ln 1 

 ln l  ln  
 lim  
 0 

ln 1   
 
 ln l   1 ln  
 lim  
 0 

ln 1
  


 
 ln l  ln  1 
 lim  
 0  1 
 ln    
 

 lim 
 ln l  ln 1   
 
 0 

ln 1
   

 ln l 

 lim 1 
 0 

ln 1 
  
1 1
1    0,  , ln    
  

3) Consider a square of side l

[Diagram]

 N    l
2

2

d  lim

ln l
2
2

 0
 
ln 1
  1 
ln l 2  2 
  
 lim 
 0
ln 1

ln l  ln  2
2
 lim
 0
ln 1

 1 
2

 ln   

   
 ln l 2
 lim
 0 
ln 1 ln 1 
   
 

 lim
 ln l 2
 
2 ln 1  
 
 0 

 
ln 1
  
ln 1 
 
 
lim  ln l 2 
   2
  0  1 
 ln    
 
2

4) Consider a cube of side l

[Diagram]
 N    l
3

3

d  lim
ln l  3
3

 0
 
ln 1
1
ln l 3  ln
 lim 3
 0
ln 1


 lim 

ln l 3

 
ln 1 
 
3

 0 

ln 1   
 ln 1
 


 lim 
 ln l 3

 
3 ln 1 
 
 0 

ln 1   
 ln 1
 

3

So far the dimension of the objects are integral consider the following
construction in a set called as “Cantor Set”.

Cantor Set is defined as follows:-

Consider a line of unit length. Divide it into three equal parts. Then the middle
third is removed leaving the two line segments. The process is continued by
removing at each step the middle third of each segment.

To cover this set at step greater than „k‟

N k   2 k
In each case the side of the cube i.e.
k
1
  
 3
ln 2 k
d  lim
k   ln 3 k

ln 2
 lim
k   ln 3

d  0.6309

Henon Map:
In many of the examples such as Cantor set, scale invariance was observed.
Invariably in scale invariant constructions fractional dimensions was observed.
Another scale invariant example is the two dimensional Henon map:
xn1  1  cxn2  y n
y n1  y n

If the parameter  = 0, then the Henon map reduces to quadratic equation or

logistic map. The Henon map consists of two-dimensional curves covering only
part of two-dimensional space for c = 1.4 and  = 0.3 the variation of y with
respect to x is shown in the following graph.
In figure (a) a small rectangle is shown. In figure (b) the same small rectangle is
expanded. Figure (b) also has a small rectangle which is enlarged to figure (c)
and so on. All expansions show identical pattern. This is what is called as scale
invariance. Wherever scale invariance is observed the construction or set has
fractional dimension. The dimension in Henon‟s map case is d =1.264

Lyapunov Exponent
It is known that the non-linear systems are very sensitive to initial conditions.
Whether the given system exhibits non-linearly or not can be measured with a
coefficient called as Lyapunov exponent. Consider two particles with d0 as
initial separation between them. As time changes the dist may also change. The
dist of separation between the two particles at any time „t‟ is given by:
d t   d 0 e Lt
Where, L is called Lyapunov exponent.

Case (i) If L  0
The distance remains same, or keeps on decreasing, it converges to zero after
sufficiently long interval of time. We say that the system exhibits linear
characteristic. It means the system is not very sensitive to the initial conditions.

ii) If L  0
As time advances the distance of separation keeps on increasing exponentially.
The system is said to be non-linear. It is very sensitive to the initial conditions.
Thus, the Lyponove exponent is a measure of whether the system is linear or
non-linear.

Approximate analytic steady state solution of Duffing‟s oscillator:

The reduced equation of motion of a Duffing‟s oscillator is
x  2x  x  x 3  f cos t

The anharmonic oscillator is subject to periodic force. The system takes some
time to respond fully to the applied periodic force. We consider the solutions
when the system is stabilized or after sufficiently long interval of time when the
initial transients are died down.

The trial solution of a non-linear equation is assumed to be a linear

combination of the applied periodic force and its derivative.
Let xt   A  cost    
A is the amplitude of Duffing‟s oscillator and  is the phase. These two arbitrary
constants depend on ' ' .
Substituting this in the above equation we get,
x   A sin t   
x   A 2 cost   
  A 2 cost     2A sin t     A cost     A3 cos 3 t     f cos t
A1   2 cost     2A sin t     A3 cos 3 t     f cos t

This equation must be true for all values of time „t‟.

We use the following identities:

cos3 t     cos3t     cost   
1 3
4 4
cos t  cost -     
 cost -   cos  sin t -  sin 

Substituting this in above equation we get,

   1
A 1   2 cost     2A sin t      A3 cos3t     A3 cost   
3 
 4 4 
 F cost    cos  F sin t   sin 

 A1     F cos  4 A  cost     F sin   2A sin t     4 A cos 3t     0

 2 3 3 1 3

As the above relation is true for all values of time the coefficients of
cost   sin t    and cos 3t    must be independently equal to zero.
A1   2   f cos  A3  0
3
4
  3 2 
 A1    4 A   F cos   0
2

   
 2A  f sin    0
1
 A3  0
4
Third condition is difficult to obey as it makes A = 0 and the entire solution
does not exist. Therefore we make approximations here we assume A to be very
small so that A3 vanishes but A is still non-zero. The first two equations give us
f cos  A1   2   A 2 
 3 
 4 
f sin   2A
 can be eliminated from the above equation by squaring and adding the above
equation.

2  3 2
2

f  A 1    A   2  
2 2 2

 4  
f
A
2
 3 2
1    A   2 
2 2

 4 
Substitute the value of A in any of the above equation it can be solved for .

It is obvious that if xt   1 then the x3 term in the equation of motion of

Duffing‟s oscillator vanishes and we get, damped forced oscillator.

For a given frequency , f 2 is cubic in A2 [i.e. (A2)3] and gives one or three
real values for A2. Alternatively, by regarding the equation as determining 2 for
given A, a quadric equation results which is easier to solve.
For the following graph we plot A versus  for the case of hard spring with the
damping coefficients  = 0.1 and force constant f = 0.5. The dashed curve is the
amplitude frequency relation for undamped free oscillator (i.e.  = 0, f = 0)

3 2
We have   1  A
4
1


 4
  2
Or A     2  1 
 3 
This curve is called the spine of the resonance. It is the locus of
resonance peak as f is varied in the   0 limit. The anharmonic cubic term
causes the resonance amplitude term to lean over and if f is sufficiently large
then A and  become triple valued over an interval in .
The middle value in fact corresponds to an unstable steady motion. The
phase angle  of  calculated from the previous equations is shown in the
following graph.

Dynamics of rigid body

Angular momentum and angular velocity of a rigid body:
Consider, a simple system of two particles attached at the ends of a massless
rod i.e. distance between the two particles always remains constant. Let the
system be dumbbell shaped rotating about an axis passing through the center
of rod.

(Diagram)

Angular momentum of the system is given by

  
L  L1  L2
   
 r1  p1  r2  p 2
    
L  m1 r1  v1  m 2 r2  v 2

We have,
m1  m2  m
 
r1  r2  l
2
  
r1  r2  r
  
v1  v 2  v
  
 L  mr  v 1   1 1
  
L  2mr  v

Obvious from the above relation, that the direction of angular momentum L is
perpendicular to the rod. Therefore, it is not parallel to the direction of angular
velocity or axis of rotation. Thus, in general the direction of angular momentum

is not along the axis of rotation of the rigid body. Only when  = 90, L will be

parallel to  .

Consider an inertial frame. According to Newton‟s law of motion we have,


 dL
N
dt
We introduce a rotating frame such that the rotating rigid body is at rest in this
non-inertial frame. In other words, the starred frame rotates with angular
velocity . We have
 
 dL   dL   
       L
 dt   
  rest  dt  rot

But in the rotating frame, rigid body is at rest. Hence, d  L / dt   0 .

Therefore, we get the relation:
  
N L
   
If  and L are parallel than   L  0
This implies N  0 . Thus, whenever direction of angular momentum is parallel
to the axis of rotation, then the body rotates freely i.e. without any torque
acting on it.

Moment of inertia
Inertial mass is a measure of opposition to the change in linear motion of the
particle. Analogously, moment of inertia is a measure of opposition to the
change in angular motion of the rigid body. It not only depends on the mass of
the rigid body but also on the distribution of mass about the axis of rotation.
Consider a rigid body, (Rigid body consist of large number of particles with the
constraints that distance between any two particles is always constant) rotating
about an axis passing through a point in the rigid body. Therefore, there is no
translational motion of the body but only the rotational motion. Each particle
has the same angular velocity ' ' but different linear velocities and the position
vectors of all the particles are different.

For ith particle

  
vi    ri
The linear momentum of the i th particle
 
p i  mi v i
Angular momentum of the i th particle
  
Li  ri  pi
  
Li  mi ri  vi
Hence Angular momentum of the rigid body is
 
L   Li
 
  mi ri  vi 
i
  
  mi ri    ri 
i
     
  mi  ri  ri   ri ri   
i
   
  mi ri 2   mi ri   ri
i i
In a Cartesian co-ordinate system we have,

ri  xi iˆ  yi ˆj  zi kˆ
 ri 2  xi2  y i2  z i2

and    x iˆ   y ˆj   z kˆ
   


 L   mi xi2  yi2  zi2  x iˆ   y ˆj   z kˆ   mi xi x  yi y  zi z xi iˆ  yi ˆj  zi kˆ
i i


Let L  Lx iˆ  Ly ˆj  Lz kˆ
 
 L x   mi y i2  z i2  x   mi y i xi y   mi z i xi z
i i i

L y    mi x i y i  x   mi x  z  y   mi z i y i  z  2
i
2
i 
i i

L z   mi xi z i x   mi y i z i y   mi xi2  y i2  z  
i i

let
I xx   mi  y i2  z i2 
i


I yy   mi xi2  z i2 
i

I zz   m x i
2
i  y i2 
i

I xy   mi xi y i  I yx
i

I xz   mi xi z i  I zx
i

I yz   mi y i z i  I zy
i

 L x  I xx x  I xy y  I xz z

L y  I yx x  I yy y  I yz z
L z  I zx x  I zy  y  I zz  z
 L x   I xx I xy I xz   x 
L   I  
 y   yx I yy I yz   y 
 L z   I zx I zy I zz   z 
 I xx I xy I xz 
  
I
 yx I yy I yz   I
 I zx I zy I zz 
 
= Moment of inertia tensor

The component Ixy, Iyz, Izx is called products of moment of inertia of rigid body.
The components Ixx, Iyy, Izz are called components of moments of inertia of rigid
body.

The angular momentum of rigid body can now be expressed as

  
L  I 

Where I is the moment of inertia tenser having 9 components in general.

Principle axis and principle momentum:

For a regular symmetric bodies it is always possible to select the special co-
ordinate system i.e. principal axis such that products of moments of inertia
vanish, only the components Ixx, Iyy and Izz are non-zero. These non-zero
components are called principle moments of inertia. Therefore, for principle axis
the angular momentum of a rigid body can be expressed as

L  I  iˆ  I  ˆj  I  kˆ
xx x yy y zz z

Euler‟s equation
Consider a rigid body rotating about an axis passing through some point in the
rigid body. Let „L‟ be the angular momentum of the rigid body. According to
Newton‟s 2nd law of motion, the torque acting on the rigid body should be same
as rate of change of angular momentum of the rigid body. In general, the

 dL
angular momentum of the rigid body is given by, N 
dt
This is the set of Euler‟s equation for rigid body motion.

Classification of rigid bodies:

Depending on the symmetry involved in the distribution of mass in a rigid body,
there are different types of rigid bodies:
1) Ix  Iy  Iz (asymmetric top)
2) Ix=Iy=Iz (spherical top)
3) Ix = Iy  Iz (symmetric top)
4) Ix =Iy; Iz=0 (Rotor)

Symmetric Top
It is difficult to solve Euler‟s equation for the symmetric top under the
application of non-zero torque, as it is difficult to express torque in terms of
body axis. But, it is worth discussing „torque free‟ motion of the symmetric top.

Since there is no torque acting on the symmetric top i.e. N  0 we have
N x  I x x  I z  I y  z  y  0
N y  I y y  I z  I x  x z  0
N z  I z  z  I y  I x  x y  0
Where Ix, Iy, and Iz are the principle moments of inertia of the symmetric top
about the body principle axis.

For symmetric top,

Ix =Iy  Iz
Therefore, above three equations simplify as
Let Ix = Iy
I x x  I z  I x  z  y  0 A]
I x y  I z  I x  z  x  0 B]
I z z  I x  I x  x y  0
C]
 I z z  0
The third equation implies that  z  0   z  const

Simplifying A] and B]
 I  I x  
 x   z  z  y  0
 I x  
 I  I x  
&  y   z  z  x  0
 I x  
 I  Ix 
Let  z  z  
 I x 
 x   y  0

 y   x  0

The above two differential equations are interdependent coupled ordinary

differential equations in  x and  y . They can be solved easily as follows:
Differentiating the first equation with respect to time we get,
 x   y  0
Substituting from 2nd equation we get,
x   x   0
x   2 x  0
Similarly we get,
 y  2 y  0
This is 2nd order, linear, ordinary different equation with constant coefficient.
The solutions are:
 x  A cost   0 
 x   A sint   0 
Substituting in 1st equation we get,
 x  A sin t   0    y  0
 y  A sin t   0 
 x2   y2  A2 (constant)
 x2   y2   z2  A2   z2 (constant)   z is constant
 || = constant

Thus, the magnitude of an angular velocity of the symmetric top is constant.

This angular velocity vector keeps on rotating about the z axis so that its z -
component remains constant as shown in the diagram.

(Diagram)

  I  Ix  
   z  z
 Ix 
 
If Iz > Ix, then direction of  is same as that of  z .

If Iz < Ix, then the direction of  is opposite to that of  z .
It implies that the symmetric top makes slow precessional motion with angular
velocity  about the body Z-axis making an angle  b with the Z-axis. The body
axis sweeps a cone about the Z-axis. This cone formed is called a body cone.
The solutions thus obtained refer to the body axis i.e. rotating frame. But, for
the observer fixed with respect to the space can be understood as the
symmetric top rotates about its own axis and subtends angle  s is space with
the body axis. The symmetric top also sweeps a cone in space, it is called space
cone. The body cone and a space cone of the symmetric top are shown in the
following diagram:

(Diagram)

Euler‟s angle  , , 

Rigid body has six degrees of freedom. Three co-ordinates are required to
specify the point in the rigid body say centre of mass. Remaining three co-
ordinates are required to specify the orientation of the rigid body. Three angles
 , ,  are used to specify the orientations of the rigid body. They are called
Euler‟s angles.
Starting from any orientation of a rigid body, any new general orientation can
be obtained through three independent rotations of the rigid body.

 
x, y, z  about Z x, y , z  about x x, y , z  about z 1,2,3
  

(Diagram)

Corresponding to the three Euler‟s angles there will be three types of angular
velocities , and  . The direction of  is along the old Z-axis whereas the
direction of  is along the new X  axis whereas  is along the new Z-axis as
shown in the following diagram. The angular velocities are always along the
axis of rotation.

(Diagram)

The angular velocity  of the rigid body can now be expressed in terms of the
components of angular velocity along the body principal axes (1,2,3). Thus, we
have

  1eˆ1   2 eˆ2   3 eˆ3

Where,
1   cos   sin  sin
   sin  cos   sin 
2

 3     cos

Symmetric Top with torque acting on it

(Diagram)

The torque acting on the symmetric top N = mgl sin 

For symmetric top, I 1  I 2  I 3
The Kinetic Energy of the symmetric top is given by,
1 1 1
K  I 112  I 2 22  I 3 32
2 2 2

But I1  I 2
1 1 1
 I 112  I 1 22  I 3 32
2 2 2
 I 1 12   22   I 3 32
1 1
2 2
1 

 I 1   cos   sin sin
2 
   sin cos   sin 
2 2

1
  2
I 3    cos 
2 

2

1 2 2 1
 
I 1    sin 2   I 3    cos
2
 2

Potential Energy V  mgl cos

 L  K V
1
2
 1
 
L  I 1  2   2 sin 2   I 3    cos
2

2
 mgl cos

 Lagrange equation
d  L  L
1)   0
dt    
L
 I 1

d  L 
   I 1
dt   
L
 I 1 2 sin  cos - I 3    cos son  mgl sin 

 
 I 1  I 1 sin  cos  I 3sin     cos  mgl sin   0
d  L  L
2)   0
dt    ??
L
   cons tan t

 
  I 1 sin 2   I 3    cos cos

d  L  L
3)   0
dt    
L
   constant


  I    cos
 3 

Central Force
Central force is force acting on the particle such that it depends only on the
distance of the particle from a fixed point and the direction of a force is always
along the line joining the particle and the fixed point. The fixed point is called
the center of force.

F  F r rˆ

e.g.
1) Gravitational force of attraction
 mm
F  G 1 2 2 rˆ
r
k
 2 rˆ
r

(Diagram)

2) Electrostatic force between two charge particles

 1 q1 q 2
F rˆ
4 r 2
Consider a particle acted upon by a central force. The angular momentum of a
particle is given by
  
Lrp

Take center of force as origin of co-ordinate system. p is the linear momentum
of particle.
  
Lrp

   dr
 r  mv  mr 
dt
Differentiating with respect to time
   
dL dr dr  d 2r
m   mr  2
dt dt dt dt
2
 d r
 0r m 2
 dt
  
dL  md 2 r
 rF  F
dt dt

 r  F r rˆ
 rrˆ  F r rˆ
 rF r rˆ  rˆ

dL
0
dt

L  constant

Thus, the angular momentum L of the particle is constant; constant in
magnitude and constant in direction.

Consider
  
L  mr  v

Taking dot product with r̂

  
rˆ  L  mrˆ  r  v

 mrˆ  rˆ  v 

0  rˆ  rˆ  v   0

This implies that the angular momentum of a particle experiencing a central

force must be perpendicular to the position vector of the particle

Consider
    
v  L  mv  r  v 
0
This implies that the angular momentum of a particle is always perpendicular
to the velocity of the particle. In other words the particle mo ves in such a way
that its velocity always remain perpendicular to direction of angular
momentum.

It is obvious that the angular momentum L is always perpendicular to the
plane containing the radius vector and the velocity of the particle. Since
angular momentum is constant in direction and magnitude the plane
 
containing r  v must also be fixed.
Thus, a particle moving under the influence of a central force always perform a
motion confined to a single plane.
Plane polar co-ordinate

x  r cos
y  r sin 

Since position vector of r is,

r  xiˆ  yˆj

r  r cos iˆ  r sin  ˆj
We have,
 
r r
rˆ   
r r
rˆ  cos î  sin ĵ

ˆ 1 r
θ
r θ
1

 r  sin θ iˆ  cosθ ˆj
r

ˆ   sin  î  cos ĵ

r  rrˆ

 dr d
v  rrˆ 
dt dt
dr  drˆ 
 rˆ  r  
dt  dt 
dr  drˆ  dθ
 rˆ  r  
dt  dθ  dt

 rrˆ  rθ - sin θ iˆ  cos θθˆj 
 rrˆ  rθθˆ

v  v r rˆ  vθ θˆ
 vr  r and vθ  rθ

 dv
a
dt

d
dt
irˆ  rˆ 
drˆ dˆ
 rrˆ  r  rˆ  rˆ  r
dt dt
drˆ d dˆ d
 rrˆ  r  rˆ  rˆ  r
d dt d dt
 rrˆ  rrˆ  rˆ  rˆ  rˆ
 r  r 2 rˆ  2r  rˆ
Let a  ar rˆ  aθ θˆ
a r  r  r 2

a  2r  r


   
F  ma  mr  mr  2 rˆ  mr   2mr ˆ
 Fr rˆ  F ˆ
 Fr  mr  mr  2
F  mr   2mr 
Using above mathematics, it can be shown that a particle undergoing motion
under the influence of central force, its total angular momentum is always
conserved.

For a particle of mass „m‟ the angular momentum is given by

L  I
 mr 2
Consider
dL d

dt dt

mr 2 
m
d 2
dt
r  
 m2rr  mr 2

 mr 2r  r 

 r m r  2r 
 r ma  
rF

Since particle experiences central force


F  F r rˆ
 F  0
dL
Hence, 0
dt

Thus, particle moving under the influence of central force, its total angular
momentum is always conserved.
The direction of angular momentum can be obtained as follows:
  
Lrp

 rrˆ  mv
 
 mr rˆ  rrˆ  rˆ
 mr rrˆ  rˆ   mr 2rˆ  ˆ
 m 0  mr 2kˆ

L  mr 2kˆ

L is constant in magnitude and direction. Therefore, the motion of the particle
is confined to only XY plane.

Kepler‟s Laws of planetary motion:

Three laws based on observation of the planetary motion are given below and
called as Kepler‟s laws of planetary motion.
1) The orbit of a planet moving under the influence of central force is a close
orbit mostly elliptical. The motion of the planet is periodic.
2) As a planet moves around the Sun, it sweeps equal areas in equal interval of
time in other words, if the planet is closed to Sun it moves fast and if it is
away from the center, it moves slowly.
3) The cube of a radius of a orbit of a planet is directly proportional to the
square of the periodic time i.e. R 3  T 2

Proof of the 2nd law: i.e. each planet sweeps an equal area in equal intervals of
time.
[Diagram]

The area swept by the radius vector of the particle as it moves from P to Q in
time t is
1 
S  r r
2
1
 r 2 
2
 
 r  r 
S dr 1 2 
lim   r
 0  dt 2
mr 2
 (Dividing and multiplying by m)
2m
L

2m
= Constant

As angular momentum L is constant, dr/dt is always constant. It means radius
vector sweeps equal area in equal intervals of time.

Equation of an orbit:
Consider a particle like planet moving under the influence of central force. The
trajectory traced by the particle as a function of time is called the equation of
the orbit. The equation of orbit doesn‟t involve time explicitly.

The force is given by,


F  F r rˆ

 Fr  F r   m r  r 2 
 d 
2
d 2r
 m 2  mr 
dt  dt 
But L  mr 2
L2  m 2 r 4 2
 L2
  2 4
2

m r
2
L2
 m 2  mr 2 4  F r 
d r
dt m r
2 2
 F r 
d r L
m 2 
dt mr 3

Consider a special case when L = 0

   0    cons tan t
Thus, the particle moves along the straight line e.g.  particles heading towards
a heavy gold nucleus.

Consider L  0
1 1
Let r  or u  (Change of variable)
u r
L  mr 2
1 d
m 2
u dt
d Lu 2

dt m
dr d 1
Consider   
du du  u 
1
 2
u
dr dr du
Consider 
dt du dt
dr du d

du d dt
 1  du  Lu 
2
   2   
 u  d  m 
dr L du

dt m d
d 2 r d  dr 
Consider   
dt 2 dt  dt 
d  L  du 
  
dt  m  d 
 L  d  du  d 
      
 m  d  d  dt 
 L  d u  Lu 
2 2
    2  
 m  d  m 
 L2 u 2  d 2 u
  2  2
 m  d
From equation
 L2 u 2  d 2 u L2 u 3 1
   2   F 
 m  d m u
d 2u  m  1
  u   2 2  F  
d 2
 L u  u

Consider a special case of central force, where in force is inversely proportional

to the square of the distance of the particle from the center i.e. inverse square
law.
 F r   2
k
r
 ku 2
d 2u mk
u   2
d 2
L
The solution above equation is the complementary solution and the particular
integral. In this case, the particular integral itself is constant right hand side.
Therefore the solution is
 u   2  A cos   0 
1 mk
r L

*~*~*~*~*~*~*~*~*~*