MAT 434 – CLASSICAL MECHANICS_

TOPIC: LAGRANGIAN FORMULATION OF MECHANICS

Basic Concepts
Constraints (Restrictions)
Constraint is defined as a restriction on motion.
Constraint is anything that restricts or prevents the motion of a particle or a system of
particles in any direction, and the restricted system is called constrained system.
In other words, constraints are those conditions that limit the motion of a system.
Alternatively, if the motion of a particle (system) is restricted and it cannot extend freely
in three dimensions, then the motion is said to have some constraints.
The forces that are responsible for restricting the motion of the object are called constraint
forces or forces of constraints. They are as such unknown forces.
The constraints (restrictions) can be expressed in terms of equations as appropriate

conditions, involving coordinates r i and (or) velocities r i .

To describe the motion of a single particle or a system of particles, we have to know the
coordinates. With respect to a system of 3-dimensional coordinates: If N particles are moving
freely in three dimensions, the motion is described by 3N space coordinates. But all the
coordinates are not independent when they are constraints in the system. This means there are
restrictions imposed on the coordinates: change in one coordinate affects other coordinates.
The ‘restrictions’ can be expressed in terms of equations or appropriate conditions.
Examples:
(i) Consider a simple pendulum bob oscillating in a two dimensional space ( x, y) . The
motion of the pendulum bob is such that we have x 2  y 2  l 2 , where l is the length of
the string. This equation is a relation between the two coordinates and thus, it is a
constraint. Therefore, the motion is described by a single independent coordinate.
(ii) Next, consider a particle moving on the surface of a sphere. The particle is restricted
by the constraint so that it can only move on the surface or in the region exterior to the
surface (not inside). In this case, the three Cartesian coordinates satisfy:
x2  y 2  z 2  a 2 . Thus, in this case we have two independent coordinates. If we know
x and y , we can know z .
(iii) But if we say the particle can be anywhere inside the sphere, we have the following
inequality condition: x2  y 2  z 2  a 2 . In this case, the constraint is not in the form of
an equation as before.

1
Degree of Freedom (DOF):
The number of independent coordinates that are required to describe the motion of a system
is called the number of degrees of freedom of the system.
If in a system of N particles, there are k number of constraints, we have n  3N  k
number of independent coordinates, n = number of degrees of freedom.
And these n independent coordinates are called generalized coordinates and they are
denoted by q1 , q2 , q3 , , qn .

Degree of Freedom (DOF): No. of independent coordinates required to represent the entire
motion =3 × (No. of particles) – No. of constraints
Simple pendulum
Since the system has one particle ( N  1), we need 3 space coordinates ( 3N ).
Constraints:
(i) The particle moves in a plane, hence z  0 .
(ii) The length of the string is constant; l  const.
Therefore, we need n  3N  k  3(1)  2  1 independent variable. That is, it has one degree of
freedom. (Generalized coordinate: q   )

Classification of Constraints
Constraints are classified according to their nature:
(i) Holonomic Constraints: If the differential equations of the constraints are integrable.
(ii) Non-Holonomic constraints: If the differential equations of the constraints are not
integrable.

Holonomic Constraints:
If the differential equations of the constraints are integrable, then the constraints are holonomic
and the holonomic constraints are expressible by equations of the following form:

fi (r1 , r 2 , r3 , , rN , t )  0 ,

where i  1, 2,3, ,k .

Note: A holonomic constraint can be mathematically described as an algebraic equation between

the coordinates of particles in the system and time.
A dynamical system is said to be holonomic if it has holonomic constraints.

2
Examples:
The following are examples of holonomic constraints:

(i) A particle is constraint to move in the x-y plane, the equation of constraint is z  0 , the
constraint is holonomic.
(ii) A particle is constraint to move on a circle of radius r in the x-y plane, the equations
of constraints are z  0 , x2  y 2  r 2  0 . The constraints are holonomic.
(iii) Constraint of Rigidity: The simplest example of holonomic constraint is a rigid body,
where the constraint on the motion of the particles keep the distances rij  ri  rj
unchanged. So, the constraints are expressed by the equation of the form

ri  rj  cij

(Length between any two points is constant).

So, it is a holonomic constraint.
(iv) Motion of a particle moving down the surface of a sphere:
A particle placed on the surface of a solid sphere is subject to the constraint that it can
move only on the surface or in the region exterior to the sphere. So, the constraint is
expressed by the equation: x2  y 2  z 2  a 2 . This is a holonomic constraint.
(v) An object sliding down an inclined plane is subject to the constraint that it can move
only on the inclined plane.

So, the constraint is expressed by the equation:

y
tan    tan   const.
x
y  x tan 
This is a holonomic constraint.

(vi) An object sliding down an inclined plane where the inclination angle of the plane varies
with time:

3
 y
Constraint: tan t  (  t ).
x
(vii) Sliding bead on a circular wire of radius ‘a’ in XY-plane.

Constraints: z  0, x 2  y 2  a 2 .

The particle restricted to move along a curve is also an example of holonomic

constraint.
Note: (1) An important characteristic of holonomic constraints is that they are integrable to get a
function of space coordinates and time.
(2) Sometime the constraints are written in terms of velocity and momentum, etc but they can be
integrated to find relations among coordinates ri , t .

Non-Holonomic Constraints:
The constraints, which cannot be expressed in the form

fi (r1 , r2 , r3 , , rN , t )  0

are called Non-holonomic constraints.

The differential equations of such constraints are not integrable,
The non-holonomic constraints may also appear in the form of inequalities.
For examples:

(i) If ri  rj  cij , the system is non-holonomic and rigidity also vanishes.

(ii) The equation of constraint in the case of a2 particle moving on or outside the
a
surface of a sphere of radius is x 2
 y 2
 z  a 2
if the origin of the coordinate
system coincides with the centre of the sphere. This inequality is a non-
holonomic constraint.

4
 A particle placed in the interior of a sphere of radius of a : x  y  z  a .
2 2 2 2
(iii)
(iv) Gas molecules confined in a spherical container of radius “a”. If ri is the position
vector of the ith molecule, then the constraint equations are: xi2  yi2  zi2  a 2 . Here,
the centre of the sphere is the origin of the coordinate system.

Note: Non-holonomic constraints are harder to solve.

Another classification of Constraints:

Constraints can also be classified into the following ways:
(i) Scleronomic: When the constraints are independent of time.
(ii) Rheonomic: When the constraints have explicit dependence on time.
Examples:
For example, a pendulum with an inextensible string is scleronomic, and the condition of the
constraint is x 2  y 2  l02 , where x and y are the coordinates of the pendulum bob, and l0 is the
length of the string. Note that the length of the string is constant.
On the other hand, a pendulum with an extensible string is rheonomic; the condition of constraint
in this case is x 2  y 2  l 2 (t ) where l (t ) is the length of the string at time t . Here, the length of
the string varies with time. The bob is constrained to move in a plane in both cases.

The constraints in the case of a rigid body are Scleronomic constraints while that of a
bead of a rotating wire loop is Rheonomic.

5
Degrees of Freedom - DOF (n) – (Revisited)
The number of independent coordinates needed to specify the position of a particle or
configuration of a system is called its degree of freedom.

Degree of freedom of a particle moving in R 3 is free from constraints, so its number of

degrees of freedom, n  3 .
A system of N particles, free from constraints, has 3N independent coordinates or 3N
degrees of freedom.
If there are k constraints, then we have n  3N  k independent coordinates or n  3N  k
degrees of freedom.

Configuration Space:
The first step that an observer takes in describing the motion of a system of particles is the
complete specification of each of their positions in 3D Euclidean space. This specification is
referred to as the configuration of the system. The set of all possible configurations of a system of
particles is known as the configuration space of the system.

This configuration space can be identified with R n .

Note: Configuration space is the space of the degrees of freedom.
Examples:
1. A particle constrained to move on a straight rigid wire has 1 degree of freedom, with
independent coordinate, x .
2. A particle constrained to move on a plane has 2 degrees of freedom, n  2 , with
independent coordinates ( x, y) or (r , ) .
3. Simple pendulum
Degree of freedom, n  1 ,
with independent coordinate  .

Generalized Coordinates: (Revisited)

If there are k holonomic constraints on a system of N particles, the number of degrees
of freedom is n  3N  k .
We will need n  3N  k independent coordinates to specify the configuration space of the
system.
These n  3N  k independent coordinates are called generalized coordinates (good
coordinates), and they are denoted by q1 , q2 , q3 , , qn .

6
Illustrations:
1. Simple Pendulum:

Consider the oscillations of the bob of a simple pendulum of length l oscillating in the X-Y plane
to be described by the Cartesian coordinates x and y with the origin O coinciding with the point
of suspension.

The coordinates x and y are not independent of one another but are connected by the equation
of constraint as

x2  y 2  l 2

 x  l 2  y2

7
One cannot change x independently, any change in x will automatically change y .
Note that:
x  l sin  and y  l cos .

If you choose  as the only coordinate, it can represent the entire motion of the bob in XY
plane.
.
In this problem, only one coordinate  is sufficient which is the sole independent coordinate. We
can write
x y
q    sin 1 or q    cos 1
l l
In this case, number of particles = 1;

The constraints are: x 2  y 2  l 2 and z  0 .

So, the number of constraints = 2.

DOF: No. of independent coordinates required to completely specify the motion
 3 (no. of particles)  (No. of constraints)
=3 1  2  1
DOF = 1; The angle  serves as the generalized coordinate.

2. Double Pendulum:
To describe the motion double pendulum in XY plane, one needs four coordinates ( x1, y1 , x2, y2 ) in
Cartesian coordinate system.

8
The Cartesian coordinates are not independent of one another, they are related by constraint
equations.

The system is subject to four holonomic constraints. The holonomic equations, z1  0 and
z2  0
constrain the particles to be moving in a plane, and, if the strings are kept taut (subjected to great
tension), we have the additional holonomic constraints

x12  x22  l12

( x2  x1 )2  ( y2  y1 )2  l22

So, the number of constraints = 4

z1  0; z2  0;
x12  y12  l12 ;
( x2  x1 ) 2  ( y2  y1 ) 2  l22 .

DOF: No. of independent coordinates required to completely specify the motion

 3 (no. of particles)  (No. of constraints)
=3  2  4  2
Thus, only two coordinates are needed to describe the system, and they could conveniently be the
angles that the two strings make with the vertical.

 
If you choose 1 and 2 as the coordinates, then they can adequately describe the motion of the
double pendulum at any instant. (they are complete)

9
Generalized coordinates: The two angles
1 and  2 served as the generalized coordinates in
this case. So, we can write

q1  1 q2  2
and .

10
 The Principle of Virtual Work and D’Alembert’s Principle

Consider a system described by n generalized coordinates q1 , q2 , q3 , , qn undergoing a

displacement such that it does not take any time and that it is consistent with the constraints of the
system. Such displacements are called virtual displacements because they do not represent actual
displacements of the system. (A virtual displacement is an imaginary, instantaneous
displacement which is consistent with the constraint relation at a given instant, i.e. without
allowing real time to change, and denoted by  r ).
Since there is no actual motion of the system, the work done by the forces of constraints in such a
virtual displacement is zero.

If  ri is the virtual displacement of the ith particle on which a resultant force Fi acts and the
system is in equilibrium, the virtual work done,  Wi  Fi   ri , is zero. (No work is done in the
virtual displacement).
The total virtual work done is then given by
N
 W   Fi   ri  0 (1)
i 1

The resultant force acting on the ith particle is made up of two forces, namely

Fi a called the applied force

f i called the force of constraint

Hence, we can write Fi  Fi a  fi . (2)

Therefore, we get
N

 (F
i 1
i
a
 fi )   ri  0 (3)

Expanding this, we have

N N

 Fi a   ri   fi   ri  0 .
i 1 i 1
(4)

N
Under a virtual displacement, the work done by the forces of constraints is zero (  fi   ri  0 ),
i 1

therefore, the last equation reduces to

N

F
i 1
i
a
  ri  0 , (5)

11
That is, for the equilibrium of a system, the virtual work of applied forces is zero.

Principle of virtual work.

The principle of virtual work states that if a system that is in equilibrium under the action of a
set of forces is subjected to a virtual displacement, then the total virtual work done by the forces
will be equal to zero.
Note: This principle deals only with statics.

D’Alembert’s Principle
The principle of virtual work deals only with statics and the general motion of the system is not
relevant here. D’Alembert suggested a principle that involves the general motion of the system.
We know that according to Newton’s second law of motion, the force acting on the ith particle is
given by

Fi  mi ai
dvi
 mi
dt
d
 (mi vi )
dt
dp
 i  pi
dt

That is, Fi  pi

or

Fi  pi  0 . (6)

This means that ith particle will be in equilibrium under the action of a force equal to the actual
force Fi plus a “reversed effective force” (also called inertial forces)  pi .

Thus, the principle of virtual work takes the form:

N

 ( F  p ). r  0 .
i 1
i i i (7)

Now, since Fi  Fi a  fi , we have

N

 (F
i 1
i
a
 fi  pi ). ri  0 (8)

12
which can be arranged to give
N N

 ( Fi a  pi ). ri   fi . ri  0 .
i 1 i 1
(9)

Using the assumption that the virtual work done by forces of constraints is zero, the last equation
reduces to
N

 (F
i 1
i
a
 pi ). ri  0 . (10)

This is D’Alembert’s principle.

The principle states that the work done by the applied forces Fi a , plus the work done by the inertial
forces,  pi in virtual displacement  ri is zero.

Derivation of Lagrange’s Equation from D’Alembert’s Principle

Let us consider a system of N particles. The coordinate transformation equations are
ri  ri (q1 , q2 , q3 , , qn , t ) (11)

where qk ( k  1, 2, , n ) are the generalized coordinates and t is the time.

Differentiating the last equation with respect to t (using chain rule), we have
dri ri dq1 ri dq2 ri dqn ri dt
    
dt q1 dt q2 dt dqn dt t dt
ri dq1 ri dq2 ri dqn ri
    
q1 dt q2 dt dqn dt t

Or

dri n
r dq j ri
r  i  (12)
dt j 1 q j dt t

Or

n
ri r
r  qj  i (13)
j 1 q j t

The virtual displacements of the particles can be expressed in the form

ri r ri
 ri   q1  i  q2    qn
q1 q2 qn

13
Or
n
ri
 ri    qj . (14)
j 1 q j

(Since by definition, virtual displacements do not depend on time,  t  0 )

According to D’ Alembert’ s principle:
N

 (F
i 1
i
a
 pi ). ri  0 (15)

Breaking this into two terms, we have

N N

 Fi a . ri   ( pi ). ri  0
i 1 i 1

Or
N N

F
i 1
i
a
. ri   pi . ri  0
i 1
(16)

N
Consider the first term on the LHS: Fi 1
i
a
. ri

n
ri
Using the result  ri    q j in this term gives
j 1 q j

N N n
ri
F
i 1
i
a
. ri   Fi a . 
i 1 j 1 q j
 qj

n  N 
r
    Fi a . i   q j

j 1  i 1 q j 
The expression in the brackets,
N
ri
Q j   Fi a . , (17)
i 1 q j

is called a generalized force. Note that Q j does not need to have the dimensions of force but
Q j  q j must have the dimensions of work.

With this notation, we can write the work done by the applied forces in the form
N n

Fi 1
i
a
. ri   Q j  q j . (18)
j 1

14
 N
Let us now consider the work done by the inertial forces: (2nd Term):  pi . ri  0
i 1

N N

 pi . ri   mi ri . ri
i 1 i 1

N  n r 
  mi ri .   i  q j 
 j 1 q 
i 1  j 
n N 
r 
   mi ri . i   q j

j 1 i 1  q j 

n N  
N
r
  pi .  ri    mi ri . i
j 1 i 1  q j
  q j (19)
i 1 
Let’s go backward a bit.
Consider the following time derivative: (Using product rule of differentiation)

d r  d  r  r
 mi ri . i   mi ri .  i   mi ri . i (20)
dt  q j  dt  q j  q j

Rearranging the above equation, we obtain:

ri d r  d  r 
mi ri .   mi ri . i   mi ri .  i  (21)

q j dt  q j dt  q j
 
which can be expressed as:

ri d r  d  r 
mi ri .   mi vi . i   mi vi .  i  (22)
q j dt  q j  dt  q j 
d ri
(since ri   vi )
dt
n
ri r
Recall that r  qj  i (see Equation 12)
j 1 q j t

Taking the partial derivative of r with respect to q j , we have

r r
 i (23)
q j q j

This result may be remembered by looking it at simply as the cancellation of dots.

15
(Note that ri does not depend on q j ).

Also,

d  ri    d ri  ri v
      (24)
dt  q j  q j  dt  q j q j

Using Equation (23) and (24) in Equation (22), we have

ri d r  d  r 
mi ri .   mi vi . i   mi vi .  i 
q j dt  q j  dt  q j 
d r  v
  mi vi . i   mi vi . i

dt  
q j  q j

d v  v
  mi vi . i   mi vi . i
dt  q j  q j

Using this result in Equation (19), we have

n N 
N
d  ri  vi 
 pi .  ri     m v .
 i i q 
j 1 i 1  dt 
  m v
i i .

 qj
i 1
 j  q j 
n N 
d v  v 
    mi vi . i   mi vi . i   q j (25)

j 1 i 1  dt  q j  q j 

n 
d N v N
v 
    mi vi . i   mi vi . i   q j
j 1 
 dt i 1 q j i 1 q j 

We are almost there but not done yet.

Recall that the total kinetic energy of the system is given by
1 N
T 
2 i 1
mi vi2 (26)

Taking the derivative of the Kinetic Energy, T with respect to q j , we obtain

16
 T 1   N 
  
q j 2 q j  i 1
mi vi2 

1   N 
  
2 q j  i 1
mi vi .vi 

1 N 
  mi
2 i 1 q j
 vi .vi 
(26)
1 N  v v 
  mi  vi . i  i .vi 
2 i 1  q j q j 

1 N  v 
  mi  2vi . i
2 i 1   q j


N
vi
  mi vi .
i 1 q j

Similarly, differentiating the total kinetic energy, T partially with respect to q j , we have

T 1   N 2
N
v
  
q j 2 q j  i 1
mi i 
v  
 i 1
mi vi . i
q j
(27)

Substituting the last two results into Equation (25), we get

n 
N
d N vi N
v 
 pi .  ri     i i
j 1 
m v . 
q j i 1
 mi vi . i   q j
q j 
i 1  dt i 1
(28)
n 
d  T  T 
      qj

j 1  dt  q j  q 
 j

With Equations (18) and (28), Equation (16) becomes

N N

 Fi a . ri   pi . ri  0
i 1 i 1

n n  d  T  T 
  Q  q    dt  q    qj  0
  j  q j 
j j
j 1 j 1 

n  
 d  T  T  
  Q j   
     q j  0 .
j 1 
  dt  q j
  q j 


Since all the  q j are assumed to be independent variations, the individual bracketed terms in the
sum must vanish independently. So, it follows that
17
 
 d  T  T  
Qj      0

 dt  q j
  q j 


d  T  T
     Qj
dt  q j  q j

d  T  T
     Q j  0, j  1, 2, ,n (29)
dt  q j  q j
This set of n second-order differential equations for the n generalized coordinates q j is known
as Lagrange’s equations.
If the applied forces are all conservative, they can be expressed as the negative gradients of the
total potential energy q j V (r1 , r2 , , rN , t ) of the system. That is,

V
Fi a  iV   (30)
ri

The generalized force is then given by

N
ri
Q j   Fi a .
i 1 q j
N
ri
 Q j   Fi a .
i 1 q j
N
V ri
  .
i 1 ri q j
V

q j

V
Qj   (31)
q j

The Lagrange’s equations become in this case:

d  T  T
    Qj  0
dt  q j  q j

d  T  T V
      0,
dt  q j  q j q j

18
 d  T  T V
      0,
dt  q j  q j q j

d  T  (T  V )
     0,
dt  q j  q j

This can be written as

d  (T  V )  (T  V )
    0, (32)
dt  q j  q j

V
since V is not a function of q j , then  0.
q j

Equation (32) can be written as

d  L  L
    0,
dt  q j  q j
where
L  T V ,
the difference between the kinetic and potential energies, is called the Lagrangian function (or
simply Lagrangian) of the system. We have reached to Lagrange’s equation from D’Alembert’s
principle.
Lagarange’s equations provide one of the most convenient ways of writing down the
equations of motion for a wide range of mechanical systems. We can proceed as follows:
1. Select a set of generalized coordinates qi for each degree of freedom.
2. Express the kinetic energy T and potential energy V of the system in terms of the
generalized coordinates, their first time derivatives, and the time.
3. Form the Lagrangian, L  T  V .
4. Substitute L into Lagrange’s equations and perform the indicated differentiations.
Some examples will make this approach clear.

19
Plane Pendulum
A bob of mass m is suspended from the ceiling by a string of length l and can swing back
and forth in a vertical plane under the influence of gravity g . The system has one degree of
freedom, and we can take as generalized coordinate the angular displacement  from vertical
(Fig. …)

A mass m is suspended from the ceiling by a string of length l and swings back and forth in a
vertical plane (Fig. …..). Gravity g acts vertically down. Use Lagrange’s method to find the
equation of motion.

Solution:
The Cartesian coordinates of m are
x  l sin  , y  l cos  ,
so the Cartesian components of the velocity of m are

x  l cos , y  l sin  .

The kinetic energy, T is

20
 1
T m( x 2  y 2 )
2
1
 m (l cos  ) 2  (l sin  ) 2 
2
1
 m(l 2 2 cos 2   l 2 2 sin 2  )
2
1 2 2
 ml  (cos 2   sin 2  )
2
1 2 2
T  ml 
2
and the potential energy (relative to the ceiling) is
V  mgh  mgy  mgl cos .
Note that potential energy is mass  acceleration dueto gravity  height , and acceleration due to
gravity is negative since it is acting downwards.
The Lagrangian is thus,
1 2 2
L  T V  ml   mgl cos  .
2
The Lagrangian has the property that:

d  L  L
  0,
dt  qi  qi

where qi are the generalized coordinates of the system (in this case, we choose  as our
generalized coordinate). That is, q1  q   .

Thus, we have
d  L  L
   0.
dt    
So, we need to calculate the following:
L  1 2 2 
  ml   mgl cos  
   2 .
 ml 2

d  L  d
  (ml  )  ml 
2 2

dt    dt

21
 L  1 2 2 
  ml   mgl cos  
   2 
 mgl sin 

d  L  L
Substituting these back into the Lagrange’s equation,    0, we obtain the equation
dt    
of motion as

ml 2  mgl sin   0 .
This is equivalent to
g
 sin   0 .
l

Double Pendulum

A double pendulum consists of two simple pendulums, with one pendulum suspended from the
bob of the other. The “upper” pendulum has mass m1 and length l1 , the “lower” pendulum has
mass m2 and length l2 , and both pendulums move in the same vertical plane.

(a) Find the Lagrangian, using as generalized coordinates the angle 1 and  2 the pendulums
make with the vertical.
(b) Write down Lagrange’s equations of motion.
Solution:

22
23
The Cartesian coordinates of m1 are

x1  l1 sin 1 , y1  l1 cos  1 ,

The Cartesian coordinates of m2 are

x2  l1 sin 1  l2 sin 2 , y2  l1 cos 1  l2 cos  2 ,

Differentiating the quantities above with respect to time, we obtain the velocities of the bobs:
x1  l11 cos1 , y1  l11 sin  1 ,

x2  l11 cos 1  l22 sin 2 , y2  l11 sin 1  l22 sin  2

The kinetic energy of m1 is

1
T1  m1 ( x12  y12 )
2
1
 m1 (l11 cos 1 ) 2  (l11 sin 1 ) 2 
2
1
 m1l1212 (cos 2 1  sin 2 1 )
2
1
 m1l1212
2
1 2 2
 T1  m1l1 1 .
2
The kinetic energy of m2 is

1
T2  m2 ( x22  y22 )
2
1
 m2 (l11 cos 1  l2 2 cos  2 ) 2  (l11 sin 1  l2 2 sin  2 ) 2 
2
1
 m2 (l1212 cos 2 1  2l1l21 2 cos 1 cos  2  l22 22 cos 2  2 )
2
(l1212 sin 2 1  2l1l21 2 sin 1 sin  2  l22 22 sin 2  2 ) 
1
 m2 l1212 (cos 2 1  sin 2 1 )  2l1l21 2 (cos 1 cos  2  sin 1 sin  2 )  l22 22 (cos 2  2  sin 2  2 ) 
2 
1
 m2 l1212  2l1l21 2 cos(1   2 )  l22 22 
2
1
 m2 l1212  l22 22  2l1l21 2 cos(1   2 ) 
2

24
where above we used the fact that cos1 cos2  sin 1 sin 2  cos(1  2 ) .

So, the total kinetic energy, T , is

1 2 2 1
T  T1  T2  m1l1 1  m2 l1212  l22 22  2l1l21 2 cos(1   2 ) 
2 2
1 1 1
 m1l1212  m2l1212  m2l22 22  m2l1l21 2 cos(1   2 )
2 2 2
1 1
 (m1  m2 )l1212  m2l22 22  m2l1l21 2 cos(1   2 )
2 2
The total potential energy of the system is
V  m1 gy1  mgy2
 m1 gl1 cos 1  m2 g (l1 cos 1  l2 cos  2 )
.
 m1 gl1 cos 1  m2 gl1 cos 1  m2 gl2 cos  2
 (m1  m2 ) gl1 cos 1  m2 gl2 cos  2

The Lagrangian is L  T  V ,
1 1
L  (m1  m2 )l1212  m2l2222  m2l1l212 cos(1  2 )   (m1  m2 ) gl1 cos 1  m2 gl2 cos  2 
2 2
1 1
L  (m1  m2 )l1212  m2l2222  m2l1l212 cos(1  2 )  (m1  m2 ) gl1 cos 1  m2 gl2 cos 2
2 2
(b) Since the generalized coordinates are now 1 and  2 , the Lagrange’s equations are

d  L  L d  L  L
   0,    0.
dt  1  1 dt   2   2

We calculate this in stages as follows:

L
 (m1  m2 )l121  m2l1l22 cos(1  2 )
1

d  L 
  (m1  m2 )l1 1  (1   2 )m2l1l2 2 sin(1   2 )  m2l1l2 2 cos(1   2 )
2

dt  1 
 (m1  m2 )l121  m2l1l2 2 cos(1   2 )  (1   2 )m2l1l2 2 sin(1   2 )

L
 m2l1l21 2 sin(1   2 )  (m1  m2 ) gl1 sin 1
1

and

25
 L
 m2l1l21 cos(1  2 )  m2l222 ,
 2

d  L 
  m2l1l21 (1   2 )sin(1   2 )  m2l1l21 cos(1   2 )  m2l2  2
2

dt   2 
 m2l22 2  m2l1l21 cos(1   2 )  m2l1l21 (1   2 )sin(1   2 ),

L
 m2l1l21 2 sin(1   2 )  m2 gl2 sin  2 .
 2

Thus, the Lagrange’s equations are:

d  L  L
  0
dt  1  1


(m1  m2 )l121  m2l1l2 2 cos(1   2 )  (1   2 )m2l1l2 2 sin(1   2 )
  m2l1l21 2 sin(1   2 )  (m1  m2 ) gl1 sin 1   0

i.e.,
(m1  m2 )l121  m2l1l22 cos(1   2 )  (1   2 )m2l1l2 2 sin(1   2 )
 m2l1l21 2 sin(1   2 )  (m1  m2 ) gl1 sin 1  0

(m1  m2 )l121  m2l1l2 2 cos(1   2 )  m2l1l212 sin(1  2 )  m2l1l222 sin(1  2 )

 m2l1l212 sin(1  2 )  (m1  m2 ) gl1 sin 1  0
i.e.,
(m1  m2 )l121  m2l1l22 cos(1  2 )  m2l1l222 sin(1  2 )  (m1  m2 ) gl1 sin 1  0

Dividing through by (m1  m2 )l12 gives

m2l1l2 m2l1l2 (m  m2 ) gl1

1   cos(1  2 )   2 sin(1  2 )  1 sin 1  0
(m1  m2 )l12 2
(m1  m2 )l12 2
(m1  m2 )l12

Or
m2l2 m2l2 g
1   2 cos(1   2 )   22 sin(1   2 )  sin 1  0
(m1  m2 )l1 (m1  m2 )l1 l1

Or

26
 m2l2 g
1   2 cos(1   2 )   22 sin(1   2 )   sin 1  0
(m1  m2 )l1 l1

d  L  L
Similarly, for    0 we have
dt   2   2

 
m2l222  m2l1l21 cos(1  2 )  m2l1l21 (1  2 )sin(1  2 )  m2l1l212 sin(1  2 )  m2 gl2 sin 2  0
,
i.e.,
m2l222  m2l1l21 cos(1  2 )  m2l1l21 (1  2 )sin(1  2 )  m2l1l212 sin(1  2 )  m2 gl2 sin 2  0

This can be simplified as

m2l222  m2l1l21 cos(1  2 )  m2l1l212 sin(1  2 )  m2l1l212 sin(1  2 )  m2l1l212 sin(1  2 )

 m2 gl2 sin  2  0

i.e.,
m2l222  m2l1l21 cos(1  2 )  m2l1l212 sin(1  2 )  m2 gl2 sin 2  0 .

Dividing through by m2l22 yields

m2l1l2 m ll m gl
2  2
1 cos(1   2 )  2 1 22 12 sin(1   2 )  2 22 sin  2  0
m2l2 m2l2 m2l2

which reduces to
l l g
 2  1 1 cos(1   2 )  1 12 sin(1   2 )  sin  2  0
l2 l2 l2

Or

l1 g
2  1 cos(1   2 )  12 sin(1   2 )   sin  2  0 .
l2 l2

In summary, the equations of motion are

m2l2 g
1   2 cos(1   2 )   22 sin(1   2 )   sin 1  0
(m1  m2 )l1 l1
.
l g
 2  1 1 cos(1   2 )  12 sin(1   2 )   sin  2  0
l2 l2

It is a set of coupled differential equations.

27
In the small angle approximation, the equations become
m2l2 g
1   2 cos(1   2 )   22 sin(1   2 )   sin 1  0
(m1  m2 )l1 l1
l1 g
2  1 cos(1   2 )  12 sin(1   2 )   sin  2  0
l2 l2

Assignment
A double pendulum consists of two simple pendula, with one pendulum suspended from the bob of the
other. If the two pendula have equal lengths and have bobs of equal mass and if both pendula are confined
to move in the same plane, find Lagrange's equations of motion for the system. Do no assume small
angles.

A double pendulum consists of two simple pendula, with one pendulum suspended from the bob
of the other. If the two pendula have equal lengths, l , and have bobs of equal mass, m , and if both
pendula are confined to move in the same vertical plane under the influence of gravity g , find
Lagrange's equations of motion for the system.. Use  and  --the angles the upper and lower
pendulums make with the downward vertical (respectively)--as the generalized coordinates. (You are
required to state any assumptions used)

Write down the Lagrangian for a simple pendulum constrained to move in a single vertical plane
(See Figure 2). Find from it the equation of motion and show that for small displacements from
equilibrium the pendulum performs simple harmonic motion.
Solution:
Considering very small amplitude of oscillation, sin    , after simplification we obtain:
g
   0.
l

Obtain the Lagrangian and equations of motion for the double pendulum illustrated in Figure 2,
where the lengths of the pendula are l1 and l2 with corresponding masses m1 and m2 .

Is the Lagrangian formulation more advantageous than the Newtonian formulation? Why?

28
 MAT434
Review Questions
Short Answer Questions:
1. Define constraint motion.
2. What is degree of freedom?
3. What is virtual displacement?
4. Define Holonomic and non-holonomic constraints.
5. Define Scleronomous and Rheonomous constraints.
6. State the D'Alembert's principle in words.
7. Write the Lagrange's equation of motion for conservative system.
8. Write the Lagrange's equation of motion for non-conservative system.
9. Define cyclic coordinates.
10. Construct the Lagrangian for Atwood's machine.
11. Construct the Lagrangian for Spherical pendulum.

Long Answer Questions:

1. What are constraints? Explain, giving examples, the meaning of holonomic and
nonholonomic constraints.
2. Explain the meaning of Scleronomous and Rheonomous constraints. Give illustrations of
each.
3. Is the Lagrangian formulation more advantageous than the Newtonian formulation? Why?
4. What do you understand by cyclic coordinates? Show that the generalized momentum
corresponding to a cyclic coordinate is a constant of motion.
5. Explain the term 'virtual displacement' and state the principle of virtual work.
6. Describe the use of Rayleigh's dissipation function.
7. Define the Hamiltonian. When is it equal to the total energy of the system? When is it
conserved?
8. What is meant by a configuration space? How is this concept used to describe the motion of
a system of particles?
9. What are constraints? Discuss holonomic and Non-holonomic constraints with illustration.
10. Discuss various types of constraints with illustration.
11. Discuss the concept of generalized coordinates with illustrations.
12. Discuss the virtual work done for motion of a system and derive the mathematical statement
of D'Alembert’s statement.
13. Derive the Lagrange's equation of motion for a conservative system from D'Alembert's
principle.

29
14. Derive the general expression of kinetic energy and find the kinetic energy of double
pendulum from it.
15. What are cyclic coordinates? Show that total energy is conserved.
16. Construct the Lagrangian of Atwood machine and derive its the equation of motion.
17. Construct the Lagrangian of spherical pendulum and derive its the equation of motion. Also
show the conservation of total energy and constant of motion.

30