Chapter VII

Hamilton’s Principle-
Lagrangian & Hamiltonian Mechanics

Recommended problems: 7-1, 7-2, 7-3, 7-4, 7-6, 7-7, 7-10, 7-12, 7-13, 7-14,
7-15, 7-25, 7-26, 7-27, 7-29, 7-33, 7-34, 7-37, 7-39,
.
Hamilton’s Principle
Of all the possible paths along which a dynamical system may move from one
point to another within a specific time interval (consistent with any constraint),
the actual path followed is that which minimizes the time interval of the
difference between the kinetic and potential energies, i.e.,
t2
  T  U dt  0 (7.1)
t1
This means that the integral of T-U must be an extremum.
Defining the difference of T-U to be the Lagrangian as

L xi , xi ; t   T  xi   U  xi  (7.2)

Eq.(7.1) now reads
t2
  L xi , xi ; t dt  0 (7.3)
t1

Comparing Eq.(7.3) and Eq.(.3) with f  L, x  t y  x y  x

We get the Euler’s equations as
 L d  L 
  0 (7.4) Lagrange Equation
x dt  x 

Example Find the equation of motion of the 1-Dimensional harmonic oscillator.

Solution The Lagrangian of the system is

L  T  U  12 mx 2  12 kx 2 
L L
 kx &  mx 
x x
d k
 kx  mx   0   kx  mx  0  x  x0
dt m
Example Find the equation of motion of the simple pendulum shown.

Solution With respect to the top, the potential

energy of the ball is

U  mgL cos 
Lcos
And for the kinetic energy we have
L
T 1 mv 2  1 mL2 2 
2 2

L  12 mL2 2  mgL cos 

m
L L
 mgL sin  &  mL2
 
Using Lagrange equation we get

 mgL sin  
d
dt

mL2  0    mgL sin   mL2  0 
g
  sin   0
L
Generalized Coordinates

Generalized coordinates are any set of independent coordinates qi (not

connected by any equations of constraint) that completely specifies the state of
a system. The required number of generalized coordinates is equal to the
system’s number of degrees of freedom.
A single particle free to move in 3-dimensional space requires 3-coordinates to
specify configuration and hence has a 3-degrees of freedom. n-free particles
would require 3n coordinates and so on. .
For each constraint equation the number of generalized coordinated is
decreased by one coordinate. This means that if there are m equations of
constraints, then 3n-m coordinates are independent, and the system is said to
posses s=3n-m degrees of freedom. The equations of constraints must be
expressible of the form

f k  xi , t   0 i  1,2,n k  1,2,m (6.5)

Constraints that can be expressed in the form of Eq.(6.5) is called holonomic
constraint, otherwise it is nonholonomic.
Since the Lagrangian is a scalar function it is invariant to coordinates
transformation. In terms of the generalized coordinates, the Lagrange’s
equations can be written as

L d  L 
 0 j  1,2, s (6.6) Lagrange Equations
 
q j dt  q j 

There are s of these equations, and together with the m equations of

constraints and the initial conditions, they completely describe the motion of the
system.
It is important to realize that the validity of Lagrange’s equations requires the
following 2-conditions:

1- The forces acting on the system (apart from any constraint forces) must
be conservative.

1- The Equations of constraints must be in the form of Eq.(6.5), i.e., the

constraint must be holonomic..
Example Find a suitable set of generalized coordinates for a point particle
moving on the surface of a hemisphere of radius R whose center is at the
origin.
Solution The particle is constrained to move on the surface, so we have

x2  y2  z 2  R2  x2  y2  z 2  R2  0
If we choose the Cartesian coordinates our generalized coordinates will be

q1  x, q2  y z  R2  x2  y2
Example Use the (x,y) coordinates system as shown in
the figure to find the K.E. T, P.E. U, and the Lagrangian L
for a simple pendulum moving in the x-y plane. Determine
the transformation equations from the (x,y) coordinates to
the coordinates . Find the equations of motion.
Solution Using the rectangular coordinates we have

 
T  12 m x 2  y 2 U  mgy

 
L  T  U  12 m x 2  y 2  mgy
To transformation x&y into the coordinates  we have

x  l sin  & y  l cos  x  l cos & y  l sin  

 
L  12 m x 2  y 2  mgy  12 ml 2 2  mgl cos
To find the equation of motion we apply Lagrange’s equation (Eq.7.4). We have

L
 mgl sin 
L
 ml 2   mgl sin  

d ml 2
0
 
  dt
g
  sin   0

l
y
Example Consider the problem of a
projectile motion under gravity in 2-
dimensions. Find the equations of
motion in both Cartesian and polar vo
coordinates.
 x
Solution Using the Cartesian coordinates
we have, taking U=0 at y=0,
 
T  12 m x 2  y 2  U  mgy 


L  T  U  12 m x 2  y 2  mgy 
Here we have 2-generalized coordinates, x & y. So Lagrange’e equations give

L d  L  L d  L 
  0 &   0
x dt  x  y dt  y 
d
For the x-coordinate we have 0  mx   0  x  0
dt
d
 mg  my   0  y   g
dt
In polar coordinates, the generalized coordinates are r & . Now we have


T  12 m r 2  r 2 2  U  mgr sin  

 
L  12 m r 2  r 2 2  mgr sin 
Lagrange’e equations give for the 2-generalized coordinates, r &  gives
L d  L 
For the r-coordinate we have   0 
r dt  r 
d
mr 2  mg sin   mr   0  r 2  g sin   r  0
dt
L d  L 
For the -coordinate we have   0 
 dt   
d
 
 mgr cos  mr 2  0   gr cos  2rr  r 2  0
dt
Example A particle of mass m is constrained to move
on the inside surface of a smooth cone of half-angle
. Determine the generalized coordinates and the
constraints. Find the equation of motions.
Solution here we use the cylindrical coordinates r, ,
& z. The equation of constraint is
z  r cot 
So we 2-degrees of freedom and the generalized coordinates are r & . Now

    
T  12 m r 2  r 2 2  z 2  12 m r 2  r 2 2  r 2 cot 2   12 m r 2 csc2   r 2 2 
U  mgz  mgr cot   
 L  1 m r 2 csc2   r 2 2  mgr cot 
2
L d  L 
For the r-coordinate we have   0 
r dt  r 
d
 
mr  mg cot   mr csc2   0  r 2  g cot   rcsc2   0
 2
dt
L d  L 
For the -coordinate we have   0 
 dt   
d

0  mr 2  0
dt
  mr 2  constant

But L  I  mr 2  constant

So we recover the conservation of angular momentum about the axis of

symmetry of the system.
Example The point of support of a simple
pendulum of length b moves on a massless rim
of radius a rotating with constant angular
velocity . Obtain the equation of motion of m.
Solution Taking the origin of the coordinates
be at the center of the rotating rim we get

x  a cost  b sin  y  a sin t  b cos

x  a sin t  b cos y  a cos t  b sin 

The K.E & P.E are now

   
T  12 m x 2  y 2  12 m a 2 2  b 2 2  2ba sin  t 
U  mgy  mg a sin t  b cos  
 
L  12 m a 2 2  b 2 2  2ba sin  t   mg a sin t  b cos 
The only generalized coordinate is . SO

L
 mba cos  t   mgb sin 

 L

 
 m b 2  ba sin  t 

d  L 
    mb   mba cos  t    
2
dt   
Applying the Lagrange’e equation we get

mb 2  mba cos  t      mba cos  t   mgb sin  

mb 2  mba 2 cos  t   mgb sin  

a 2
g
  cos  t   sin 
b b
Note that for =0 we get
g
  sin   0
b
Which is the equation of motion of the simple pendulum.
Example Find the frequency of a simple a
pendulum placed in a rail-road car that has
a constant acceleration a in the x-direction.
l
Solution We choose a fixed coordinate l 
system with x=0, and v=vo at t=0. The
position of the mass m is m m
x  l sin   vot  12 at 2 y  l cos

x  l cos  vo  at y  l sin 

The K.E & P.E are now

 
T  12 m x 2  y 2  12 mvo  at  l cos   12 ml sin  
2 2

U  mgy  mgl cos 

L  12 mvo  at  l cos   12 ml sin    mgl cos

2 2

The only generalized coordinate is again . SO

L
 ml sin  vo  at  l cos   ml 2 2 sin  cos  mgl sin  

 L
 ml sin  vo  at   mgl sin 

g a
   sin l cos  cosl sin   cosl cos  sin l sin 
l l
L d  L 
Applying the Lagrange’s equation   0 
 dt   
g a
   sin   cos
 ( prove as HW )
l l
If the mass doesn’t oscillate with respect to the train, then   0 
g a
0   sin  l  cos l  tan  l   a
l l g
Because the oscillations are small we can assume   l    is very small
g a
     sinl     cosl   
l l
Expand the sin and cosine functions we get
L
 ml cos vo  at  l cos   ml 2 sin 2  

Using Taylor expansion for sin and cos to first order of  we get
g a
   sin l   cosl   cosl   sin l  
l l
1
    g sin l  a cosl     g cosl  a sin l 
l

The first bracket is already zero leaving     g cosl  a sin l 
l
a a g
As tan l    sin   & cos  
2 2 2 2
g a g a g

  

2 2
g 2
a 2
  l
g 2  a2 
l g a

g 2  a2
   0 
l
g 2  a2
We have SHM with frequency   2
l

Note that when the car is at rest (a=0) we get 2  g l

Example A bead slides along a smooth wire bent in the
shape of a parabola z=cr2. The bead rotates in a circle of
radius R when the wire is rotating about its vertical
symmetry axis with angular velocity . Find the value of c.
Solution Choosing the cylindrical coordinates we get


T  12 m r 2  z 2  r 2 2  U  mgz

The equation of constraint is z  cr 2  z  2crr

   
L  12 m r 2  4c 2 r 2 r 2  r 2 2  mgcr 2  L  m 4c 2 rr 2  r 2  2mgcr
r
L
r
  d  L  1
 
 m r  4c 2 r 2 r     2r  8c 2 r 2 r  16c 2 rr 2
dt  r  2
r1  4c 2 r 2   r 2 4c 2 r   r 2 gc   2   0
This equation can be simplified if r=R=constant. In such a case we have

2gc     0
2
 c
2
2g
Example Consider the double pulley
system shown. Use the coordinates
indicated to determine the equations of
motion.
Solution If l1 & l2 be the lengths of the
rope hanging freely from each pulley. The
distances x & y are measured from the
center of the pulleys. So we have

x1  x & x2  l1  x  y & x3  l1  x  l2  y 

x1  x & x2   x  y & x3   x  y 

T  12 m1x12  12 m2 x22  12 m3 x32
 12 m1x 2  12 m2  y  x 2  12 m3  y  x 2

U  U1  U 2  U 3
 m1gx  m2 g l1  y  x   m3 g l1  x  l2  y 
L  12 m1x 2  12 m2  y  x 2  12 m3  y  x 2
 m1gx  m2 g l1  y  x   m3 g l1  x  l2  y 
Here we have 2-generalized coordinates, x & y. For the x-coordinate we have

L d  L 
  0 
x dt  x 
L L
 m1g  m2 gx  m3 gx  m1x  m2  y  x   m3  y  x 
x x

 m1x  m2 x  y  m3 x  y  m1  m2  m3 g

L d  L 
For the y-coordinate we have   0 
y dt  y 
L L
 m2 gy  m3 gy  m2  y  x   m3  y  x 
y y

 m2 x  y  m3 x  y  m2  m3 g

Lagrange’s Equations with Undetermined Multipliers
As we mentioned before, Holonomic constraint is the constraint that can be
expressed in the form of Eq.(7.5). Such a constraint enables us to find an
algebraic relations between coordinates. Any constraints that can be expressed
in terms of the velocities, i.e., in the form

f k  xi , xi , t   0 i  1,2,n k  1,2,m (6.7)

Such constraints is nonholonomic unless the equations can be integrated to

yield relations among the coordinates. It is called semiholonomic.
Consider a nonholonomic constraint in the form of

 Ai xi  B  0 i  1,2,n (6.8)

i
But if Ai & B have the forms
f f
Ai  & B , f  f  x, t  (6.9)
xi t

Then Eq.(6.8) becomes

 f dxi f df
   0 (6.10) 
i xi dt t dt

f  x, t   constant  0 So the constraint is actually holonomic.

From Eq.(10) we conclude that any constraint expressible in the form

f f
 dqi  dt  0 (6.11)
i qi t
Are equivalent to the constraint given by Eq.(6.5). Referring to the previous
chapter (Eq. 6.23) we can write
L d L f
   k t  0 (6.12)
qi dt qi k qi

With k(t) are called the Lagrange undetermined multipliers. It its related to the
force of constraint. The generalized forces of constraint are given by
f
Qi   k t  (6.13)
k qi
Example Reconsider the case of the disk
rolling down an inclined plane. Find the
equation of motion, the force of constrained,
and the angular acceleration.

Solution The K.E. & P.E of the disk are

T  12 My 2  12 I 2  12 My 2  14 MR 2 2

U  Mg l  y sin 
Where l is the length of the inclined plane. The Lagrangian is therefore

L  12 My 2  14 MR 2 2  Mg l  y sin 

The equation of constraint is f  y,   y  R  0
Although the system has only one degree of freedom, we may continue to
consider both y &  as generalized coordinates and use the method of
Lagrange undetermined multipliers. Eq.(6.12) now reads
L d L f L d L f
  0 &   0 
y dt y y  dt  
L L f
 Mg sin  &  My & 1  Mg sin   My    0
y y y
L L 1 f 1
0 &  MR 2 &  R   MR 2  R  0
  2  2

These 2-equations together with the equation of constraint is sufficient to solve

for the 3-unkowns. Now from the equation of constraint we have
y  R  y  R 
2 g sin  2 g sin  Mg sin 
y  , 
 &  
3 3R 3
To find the generalized forces we use Eq.(6.13) to get
f Mg sin  f MgR sin 
Qy     Q     R 
y 3  3
These 2-genralized forces are the force of friction and the torque required to
keep the disk rolling without slipping.
Note that the problem can be solved using Eq.(6.6) by eliminating one of the
coordinates and keeping only one-generalized coordinates.
Example A particle of mass m starts at rest
on top of a smooth fixed hemisphere of
radius a. Find the force of constraints, and
determine the angle at which the particle
leaves the surface.
Solution The Lagrangian of the particle is

 
L  T  U  12 m r 2  r 2 2  mgr cos
And the equation of constraint is f  y ,   r  a  0
The generalized coordinates are r &  and Eq.(6.12) now reads

L d L f L d L f
  0 &   0 
r dt r r  dt  
L L f
 mr 2  mg cos &  mr & 1 
r r r
mr 2  mg cos  mr    0
L L f
 mgr sin  &  mr 2 & 0 
  
 mgr sin   2mrr  mr 2  0 
But r  a  r  0, & r  0 
g
ma 2  mg cos    0 mga sin   ma 2  0    sin 

a
To find  we have d d d d g
      sin  
dt d dt d a
 
g 2  g
 d  a  sin d   cos 
g

0 0 2 a a
  mg cos  ma 2  mg cos  2mg cos  2mg  mg 3 cos  2
The particle falls off the surface when =0  mg 3 cos o  2  0 

1  2 
 o  cos    48.2o
 3
Note that at the top of the hemisphere we have   mg
The Hamiltonian & Hamilton’s Equations:
For simple dynamical systems the potential energy is a function of coordinates
alone and the kinetic energy is a quadratic function of velocities, i.e.,

L  T qi , qi   V qi  (6.14) 

L T
  mi qi  pi (6.15)
qi qi
With pi is called the generalized momenta. Now from Lagrange’s equation we
have

L d  L  L dpi L
 0   0  p i  (6.16)
 
q j dt  q j  q j dt q j

Now defining the Hamiltonian function as

H   qi pi  L (6.17)
i
With L is the Lagrangian given by Eq.(6.14)
Now from Eq.(6.145) we have

L T
 qi pi   qi  qi  2T (6.18) 
i i qi i qi
H   qi pi  L  2T  T  V   T  V (6.19)
i
That is, the Hamiltonian function represents the total energy of the system. Now
L L
H   piqi  qipi  qi  qi (6.20)
i qi qi

But from Eq.(6.15) the 1ast and the 3rd term in the brackets cancel. Using Eq.
(6.16) for the 4th term, Eq.(6.20) now reads

H   qipi  p iqi (6.21)

i

Since H  H qi , pi   H H
H   pi  qi (6.22)
i pi qi

Now comparing Eq.(6.21) with Eq.(6.22) we get

 H
 qi
pi
(6.23) Hamilton' s equations of motion
H
 p i
qi
Example Use the Hamiltonian method to find the
equations of motion for a spherical pendulum of
mass m and length b.
Solution The K.E. & P.E of the particle are

 
T  12 m b 2 2  b 2 sin 2  2 & U  mgb cos

To find the Hamiltonian we have to find the generalized momenta. From

Eq.(6.15) we have
L L
p   mb 2 & p   mb 2 sin 2  
 
p2 p2
H  , , p , p   T  V    mgb cos
2mb 2
2mb sin 
2 2
Applying Hamilton’s equations, Eq.(6.23) we get

H p H p

  & 
 
p mb 2 p mb 2 sin 2 

H p2 H
p     mgb sin  & p    0
 mb sin 
2 2 