Chapter 3

Lagrangian Mechanics

Module 1

Lagrange’s equation of motion of second kind

1
 It is well known that Newton’s equation of motion correctly describes the motion of a
particle or a system of particles in an inertial frame. Sometimes, the particle may be restricted in
its motion so that it is compelled to follow a specified path or some forces may act on the
particles to keep them on the surface. Such cases would render the problem impossible to treat it
using the Newtonian formalism. Moreover, Newton’s equation of motion remains undefined if
the forces of constraints operating on a system are not known priori. Lagrange’s equation of
motion of the first kind provides a technique to overcome such situation. In this case, Newton’s
equation of motion is solved by evaluating the forces of constraints from the constraint relations.
But it is a tedious procedure. Moreover, Newton’s equation of motion is valid in an inertial frame
only. The forces of constraints operating on a dynamical system restrict some of these
coordinates to vary independently. In other wards, all the coordinates which describe the
configuration of a dynamical system moving under the forces of constraints may not necessarily
be independent. Naturally, resulting equations of motion are not independent. As a result, a set of
independent coordinates is required to describe the configuration of the dynamical system and
also for obtaining a general equation of motion which is valid in any system of coordinates.

This leads us to derive a general equation of motion, known as Lagrange’s equation of motion,
which is valid in any coordinate system. Its derivation does not require the knowledge of
constraint forces, rather the knowledge of work, energy and principle of virtual work is needed.

Lagrangian methods can provide a much fresher way of solving some physical systems
compared to Newtonian mechanics, in particular for the system moving under some constraints.
Lagrangian mechanics, introduced by Joseph Louis Lagrange in 1788, is a reformulation of
classical mechanics in terms of arbitrary coordinates.

3.1 An Important result

2
 
coordinates which specify the position of the system at time t. Then each ri is a function of
 
q1, q2, ..., qn and time t i.e. ri = ri ( q1, q2, ..., qn ).

Time derivative of the generalized coordinate qi is called the generalized velocity of the i-th
particle. The velocity of the i-th particle is given by

 
 n ri ri
r  qj 
 .
j 1 q j t

Differentiating this equation with respect to generalized coordinate ‘ q j ’ again we have,

 
ri ri
 for j=1,2,…,n
q j q j

  ri
d  ri
Again,    .
dt  q j  q j

This proves that the order of differentiation with respect to ‘t’ and ‘ q j ’ are immaterial.

3.2 Lagrange’s equation of motion of second kind

Let the system contains N particles of masses mi (i  1, 2,..., N) . Let q1, q2, ..., qn be

n generalized coordinates which specify the position of the system at time t. If ri be the
position vector of the i-th mass then

 
ri = ri ( q1, q2, ..., qn ) (i  1, 2,..., N) .

  
  F  m r . r  0
N
From the generalized D’Alembert’s principle we have i i i i (3.1)
i 1

3
  
where Fi ’s are the external forces acting on the system and  ri ’s are the small instantaneous
virtual displacements consistent with the constraints.

N
  N  
From equation (3.1) we have,  i i . ri   F.i  ri
m 
r
i 1 i 1
(3.2)


   n ri
Since ri = ri ( q1, q2, ..., qn ),  ri    qe .
e 1 qe

  N   n ri 
N n
 N  ri 
 n
Then  w   F.i  ri   F.i    e   i
 q  F.  e  Qe qe
 q 
i 1 i 1  e1 qe  e 1  i 1 qe  e 1

w
where Qe  is the generalized force associated with the generalized coordinates
qe

qe , (e  1, 2,..., n) .

Now,

 
N
  N   n ri  n  N  ri 
 m 
r
i i . ri   mi i 
r .

 q e    i i m r .

 qe
i 1 i 1  e 1 q e  e 1  i 1 q e 

n    
d N
 r  N
 d  r 
     mi ri . i    mi ri .  i   qe
e 1  dt  i 1 qe  i 1 dt  qe  
n   
d N  ri  N    dri  
     mi ri .    mi ri .    qe
e 1  dt  i 1 qe  i 1 qe  dt  
 
n 
d N  ri  N  ri 
     mi ri .    mi ri .  qe
e 1  dt  i 1 qe  i 1 qe 

1 N 
T  Kinetic Energy of the system= 
2 i 1
mi ri 2

   
T n
 ri n
 ri ri ri
Now,   mi ri .   mi ri . (since,  ).
qe i 1 qe i 1 qe qe qe

4
 
T N
 ri
  mi ri .
qe i 1 qe

N
  n  d  T  T 
Therefore, 
i 1
mi ri . ri    

e 1  dt  q
  q .
e  qe  e

Substituting these in equation (3.2) and transferring all the terms in one side we have,
n   d  T  T 
 
 Q e     qe  0 .
 dt  qe  qe 

(3.3)
e 1  

Case (i): System with n degrees of freedom

In this case the coordinates are free coordinates and can be varied arbitrarily. So, the
coefficients of each  qe must vanish separately, giving

 d  T  T 
 
Qe      =0 for e  1, 2,..., n
 dt  qe  qe 
 

d  T  T
Or,    Qe for e  1, 2,..., n (3.4)
dt  qe  qe

These equations are called the Lagrange’s equations of motion of a dynamical system with n
degrees of freedom. These are second order differential equations.

V
If the system is conservative then Qe   where V  V (qe , t ) is the potential function.
qe

Substituting the value for Qe in equation (3.4) we have,

d  T  T V
  
dt  qe  qe qe

5
 d   T  V    T  V 
Or,   0,
dt  qe  qe

assuming that V is independent of the generalized velocity qe .

If we set L  T  V then Lagrange’s equations of motion can be written as

d  L  L
   0, e  1, 2,..., n
dt  qe  qe

Here L is called the Lagrangian of the system.

Note:

If the system contains some forces derivable from a potential function and some other forces
not derivable from a potential function then the Lagrange’s equation of motion can be written
as

d  L  L
   Qe/ , e  1, 2,..., n (3.5)
dt  qe  qe

where all the potential forces have been included in the Lagrangian L and the non-potential
forces are given by Qe/ .

Examples of such systems are the systems containing frictional forces.

Case (ii): Holonomic dynamical system with k bilateral constraints

In this case the generalized coordinates are connected by k independent relations of the
following form:

f j  q1 , q2 ,..., qn , t   0, j  1, 2,..., k (k  n) (3.6)

6
We now consider a virtual change of the system at time t consistent with the constraints in
which the coordinates q1, q2, ..., qn are changed to q1   q1 , q2   q2 ,..., qn   qn . Therefore,

from equation (3.6) we have, f j  q1   q1 , q2   q2 ,..., qn   qn , t   0 which can be expanded

in a series like

n f j
f j  q1 , q2 ,..., qn , t     qe  O( qe )2  0.
e 1 qe

n f j
Since changes  qe are small we have  q  q
e 1
e  0 for j  1, 2,..., k (k  n) . (3.7)
e

It is evident from (3.7) that the changes  q1 ,  q2 ,...,  qk are not independent (as they can be

solved in terms of the rest). We now introduce k arbitrary parameters 1 , 2 ,..., k . We now
k n f j
multiply the equation (3.7) by these k parameters and sum up to obtain j 
j 1 e 1 qe
 qe  0

n k
 f j 
or,    j
qe 
 qe  0 . (3.8)
e 1 j 1 

We now add equation (3.8) to the equation (3.3) and obtain,

n   d  T  T 
  k f j 
 
e 1 
Qe   
  
 

  j
 
 qe  0 . (3.9)
  dt  q e  qe  j 1 q e 

We now choose 1 , 2 ,..., k in such a way that the coefficients of  q1 ,  q2 ,...,  qk vanish
separately. This gives


 d  T  T   k f j
Qe        j  0 for e  1, 2,..., k ( n)

  e
dt q q 
e j 1 q e

d  T  T k f
Or,   
dt  qe  qe
 Qe  
j 1
 j j for e  1, 2,..., k ( n) .
qe
(3.10)

Now equation (3.9) takes the following form

Since the variations  qk 1 ,  qk 2 ,...,  qn are arbitrary and independent we must have


 d  T  T   k f j
Qe       j  0 for e  k  1, k  2,..., n
 dt  qe  qe 
  j 1 qe

d  T  T k f
Or,    Qe    j j for e  k  1, k  2,..., n . (3.12)
dt  qe  qe j 1 qe

Now, equations (3.10) and (3.12) together give

d  T  T k f
  
dt  qe  qe
 Qe  
j 1
 j j for e  1, 2,..., k , k  1, k  2,..., n .
qe
(3.13)

These are the Lagrange’s equations of motion for a holonomic dynamical system with k
bilateral constraints.

We have (n+k) unknown quantities q1, q2, ..., qn ; 1 , 2 ,..., k . To solve the equations given by

(3.13) (n number of equations) we have to supply k equations of constraints.