61
©
Copyright 2004
5
Motion under the Influence
of a Central Force
The fundamental forces of nature depend only on the distance from the
source. All the complex interactions that occur in the real world arise from these
forces, and while many of them are usually described in a more complex manner,
their origin can be found in the fundamental forces that depend only on distance.
Thus even the intricate forces of aerodynamic drag can ultimately be described as
resulting from the electrostatic potential of the air molecules scattering with the
electrostatic potential of the molecules of the aircraft, and the electrostatic force is
essentially one that depends on distance alone. It is the presence of many sources
of the distance-dependent forces that enables the complex world we know to
exist. Thus, in order to understand complex phenomena, it is appropriate that we
begin with the simplest. Therefore we will begin by applying the tools of
mechanics developed in the previous two chapters to describe the motion of an
object moving under the influence of a single source of a force that depends only
on the distance. We will call this object a "test particle" to make clear that its
motion in no way affects the source of the potential. Such a situation is known as
a central force problem since the source may be located at the origin of the
coordinate system making it central to the resulting description.
5.1
Symmetry, Conservation Laws, the Lagrangian, and Hamiltonian for
Central
Forces
Since there is a single source producing a force that depends only on
distance, the force law is spherically symmetric. If this is the case, then there can
be no torques present in the system as there would have to be a preferred axis
about which the torques act. That would violate the spherical symmetry so
0
dt
L
d
N
=
=
r
r
. (5.1.1)
Equation (5.1.1) clearly means that the total angular momentum of the test
particle does not change in time. Specifically, it means that the direction of the
angular momentum vector doesn't change. Since there is only one particle in this
system, this is little more than a statement of the conservation of angular
momentum, but it has a great simplifying implication. The radius vector
r
r
and the
particle's linear momentum
p
r
define a plane. Since
p
r
L
r
r
r
×
=
, (5.1.2)
the angular momentum is always perpendicular to that plane and being constant in
space requires that the motion of the particle is confined to that plane. Thus we
can immediately reduce the problem to a two dimensional description.
Since there is only one particle in the system and we require the total
energy of the system to be constant, the total energy of the particle must be
constant. Thus such a force is conservative and we may use the Lagrangian
formalism of Chapter 3 to obtain the equations of motion. We begin this
procedure by choosing a set of generalized coordinates. Remember that the only
requirement for the generalized coordinates is that they span the space of the
motion and be linearly independent. For motion that is confined to a plane defined
by the action of a central force, the logical choice of a coordinate frame is polar
coordinates with the center of the force field located at the origin of the coordinate
system. However, since the kinetic energy is more obviously written in Cartesian
coordinates, let us use the definition of the Lagrangian to write
(
)
)
r
(
m
y
x
m
V
T
2
2
2
1
Φ
−
+
=
−
=
&
&
L
, (5.1.3)
where
Φ
(r) is the potential giving rise to the conservative central force. The
transformation from Cartesian coordinates to polar coordinates is
âŽ
⎬
⎫
θ
=
θ
=
sin
r
y
cos
r
x
, (5.1.4)
62
so that
⎪âŽ
⎪
⎬
⎫
θ
θ
+
θ
θ
θ
−
θ
=
θ
θ
+
θ
θ
θ
−
θ
=
2
2
2
2
2
2
2
2
2
2
2
2
cos
r
cos
sin
r
r
2
sin
r
y
sin
r
cos
sin
r
r
2
cos
r
x
&
&
&
&
&
&
&
&
&
&
. (5.1.5)
Substitution of these expressions into the Lagrangian gives
(
)
)
r
(
m
r
r
m
2
2
2
2
1
Φ
−
θ
+
=
&
&
L
. (5.1.6)
Lagrange's equations of motion for polar coordinates will then be
⎪
⎪
âŽ
⎪
⎪
⎬
⎫
=
θ
∂
∂
−
⎟⎟
âŽ
⎞
⎜⎜
âŽ
⎛
θ
∂
∂
=
∂
∂
−
⎟⎟
âŽ
⎞
⎜⎜
âŽ
⎛
∂
∂
0
dt
d
0
r
r
dt
d
L
L
L
L
&
&
. (5.1.7)
In terms of the polar coordinates [r,
θ
] the quantities required for Lagrange's
equations of motion are:
⎪
⎪
⎪
⎪
⎪
âŽ
⎪⎪
⎪
⎪
⎪
⎬
⎫
=
θ
∂
∂
∂
Φ
∂
−
θ
=
∂
∂
θ
=
θ
∂
∂
=
∂
∂
0
r
)
r
(
m
mr
r
mr
r
m
r
2
2
L
L
L
L
&
&
&
&
&
. (5.1.8)
so that the explicit equations of motion become
( )
⎪
⎪
âŽ
⎪⎪
⎬
⎫
=
θ
+
θ
=
θ
=
∂
Φ
∂
+
θ
−
0
mr
r
mr
2
mr
dt
d
0
r
)
r
(
m
mr
r
m
2
2
2
&&
&
&
&
&
&&
. (5.1.9)
63
Now in Chapter 3 we developed the Hamiltonian from the Lagrangian and
the generalized momenta [see equations (3.3.1- 3.3.3)] and it is illustrative to do
this explicitly for the case of a central force. The generalized momenta can be
obtained from the Lagrangian by means of equation (3.3.1) so for polar
coordinates we have
⎪âŽ
⎪
⎬
⎫
θ
=
θ
=
]
mr
,
r
m
[
p
]
,
r
[
q
2
i
i
&
&
. (5.1.10)
From equation (3.3.3) we can then write the Hamiltonian as
(
) (
)
E
U
T
)
r
(
m
mr
r
m
mr
r
m
)
t
,
,
r
,
p
,
p
(
H
2
2
2
1
2
2
1
2
2
2
r
=
+
=
Φ
+
θ
+
−
θ
+
=
θ
θ
&
&
&
&
. (5.1.11)
As long as the Lagrangian and generalized coordinates were not explicit functions
of time, the Hamiltonian is an integral of the motion. Since from equation (5.1.11)
it is clear that the Hamiltonian is also the total energy, we have effectively
recovered the law of conservation of energy. The Hamilton canonical equations of
motion [see equations (3.3.6)] effectively add nothing new to the problem since
we already have two constants of the motion (i.e., the angular momentum and the
total energy). Thus we can turn to the implications of these two constants.
5.2
The Areal Velocity and Kepler's Second Law
The
θ
-equation of motion that gives us the constancy of angular
momentum enables us to write
L
mr
mr
2
2
=
ω
=
θ
&
. (5.2.1)
The differential area included between two radius vectors separated by a
differential angle d
θ
is just
( )
θ
=
rd
r
dA
2
1
. (5.2.2)
Let us call the time derivative of this area the areal velocity so that
.
const
m
L
dt
d
r
dt
dA
2
1
2
2
1
=
=
θ
=
(5.2.3)
Thus we could say that the areal velocity of the test particle is constant and
merely be re-stating the conservation of angular momentum. This is indeed the
64
way Johannes Kepler gave his second law of planetary motion. However, Kepler
had no conception of angular momentum and his laws dealt only with the planets.
Here we see that Kepler's second law not only applies to objects moving under the
influence of the gravitational force, but will hold for an object moving under the
influence of any central force regardless of its distance dependence. Thus Kepler's
second law of planetary motion is far more general than Kepler ever knew.
We may use this result to eliminate from the first of the two Lagrangian
equations of motion and thereby reduce the problem to that of one dimension.
Solving equation (5.2.1) for and substituting into the r-equation (5.1.9) we get
θ
&
θ
&
0
r
)
r
(
mr
L
r
m
3
2
=
∂
Φ
∂
+
−
&&
. (5.2.4)
We can obtain the result of the Hamiltonian directly by multiplying equation
(5.2.4) by r
&
and re-writing as
[ ]
[
]
m
r
)
r
(
dt
d
m
mr
2
L
)
r
(
dt
d
r
m
dt
d
2
2
2
1
2
2
2
2
1
θ
+
Φ
−
=
⎥
⎥
⎦
⎤
⎢
⎢
⎣
⎡
+
Φ
−
=
&
&
, (5.2.5)
which becomes
(
)
[
]
(
)
0
dt
dE
V
T
dt
d
)
r
(
m
r
r
m
dt
d
2
2
2
2
1
=
=
+
=
Φ
+
θ
+
&
&
. (5.2.6)
5.3
The Solution of the Equations of Motion
Finding the two integrals of the motion goes along way to completing the
solution of the problem. These two constants essentially amount to integrating
each of equations (5.1.9) once so that the equations of motion can be written as
⎪
⎪
âŽ
⎪⎪
⎬
⎫
=
=
θ
=
=
⎥
⎥
⎦
⎤
⎢
⎢
⎣
⎡
Φ
+
+
.
const
L
mr
.
const
E
)
r
(
m
)
mr
2
(
L
r
m
2
2
2
2
2
1
&
&
. (5.3.1)
65
The first of these is obtained by integrating equation (5.2.6) and replacing from
equation (5.2.1). The other is equation (5.2.1) itself. These are two first order
differential equations and require two more constants to completely specify their
θ
&
solution. These two additional constants are known as initial conditions and it is
worth distinguishing them from integrals of the motion. An
integral of the motion
will have the same value for the entire temporal history of the system while an
initial value or boundary condition
is just the value of one of the dependent
variables of the problem at some
specific
instant in time. Surely the latter is a
constant since it is specified at a given time, but an integral of the motion is some
combination of the dependent variables that is constant for all time. Integrals of
the motions are exceedingly important to any dynamics problem and knowledge
of them, as we shall see later, places very useful constraints on the history and
nature of the system.
The solution of the second of equations (5.3.1) yields the temporal history
of the variable
θ
and can be obtained by direct integration of that equation so that
∫
θ
+
=
θ
t
0
0
2
)
t
(
mr
Ldt
)
t
(
. (5.3.2)
Here the constant
is the initial value of
θ
at t = 0. We have chosen the initial
value of the time to be zero, but that is arbitrary and the initial value could have
been anything. The first of equations (5.3.1) is somewhat more difficult to solve.
Direct integration gives
0
θ
( )
∫
−
Φ
−
=
r
r
2
/
1
2
2
/
1
2
m
0
]
mr
2
/
L
)
r
(
E
[
dr
)
r
(
t
. (5.3.3)
Thus the r-coordinate is given as an implicit function of time t(r). This implicit
function must be inverted to have the same form as equation (5.3.2). The
parameter is the second initial value, being the value of r at t = 0. Thus the two
initial values
, and the integrals of the motion [L,E] completely specify the
motion of the particle. However, to solve a specific problem we must specify
Φ
(r)
because the integral, and subsequent inversion of equation (5.3.3), cannot be done
without knowledge of
Φ
(r).
0
r
]
r
,
[
0
0
θ
In any event, we may put rather general limits on the range of solutions
that we can expect for any given
Φ
(r). The second of equations (5.3.3) is
essentially a one dimensional equation in r, so we will define anew potential
2
2
2
r
m
2
L
)
r
(
)
r
(
+
Φ
=
∅
. (5.3.4)
66
We can require that the potential energy vanish as
∞
→
r
as a reasonable
boundary condition. Thus the kinetic energy is
)
r
(
m
E
T
∅
−
=
. (5.3.5)
Now if
then the force law is repulsive, and there is some minimum
distance r
0
)
r
(
≥
∅
min
to which the particle can approach the source before the right hand
side of equation (5.3.5) would become negative, implying a nonphysical negative
kinetic energy. This is a plausible result that simply says that a repulsive force
which increases in strength as its source is approached will eventually stop the
approach. The total energy for such a system must also always be greater than
zero.
For the more interesting case of power law potentials where
, we
can write
0
)
r
(
≤
∅
2
)
2
n
(
2
2
1
r
r
k
m
L
)
r
(
⎥
⎥
⎦
⎤
⎢
⎢
⎣
⎡
−
=
∅
−
, (5.3.6)
where k is some positive constant. Interesting things will happen at
(
)
[
]
)
2
n
/(
1
2
0
L
mk
2
r
−
=
. (5.3.7)
If power law dependence is such that n > 2, then serves as an upper bound for
particles with a total energy E < 0.
0
r
For the more pertinent case of gravity where n = 1, then serves as a
lower bound inside of which particles may not approach. The physical
interpretation of this lower bound is simple. If L is not zero, then there is some
angular motion of the particle as it orbits the central source. However, as it
approaches the central source, the conservation of angular momentum will require
an increase in its angular velocity keeping the particle from approaching closer.
Colloquially one could say that the particle is repelled by centrifugal force.
Indeed, this part of the pseudo-potential
∅
is often known as the "rotational
potential". Thus, if the total angular momentum L > 0, then the motion of the
particle will be kept beyond . If the total energy E < 0, then there will also be
an upper bound r
0
r
0
r
l
since the kinetic energy must remain positive. Thus for cases in
which the total energy E < 0, the particle's motion will be confined between
. This is the case for virtually all motion in the solar system.
0
1
r
r
r
≥
≥
67
5.4
The Orbit Equation and Its Solution for the Gravitational Force
To see more clearly the solution to the equations of motion, let us
eliminate time as the independent variable from equations (5.1.9). Thus, we
search for a single equation involving r as a function of
θ
. We can do this by
noting that the total time derivative operator can be written as
θ
=
θ
θ
=
d
d
mr
L
d
d
dt
d
dt
d
2
, (5.4.1)
and
( )
θ
−
=
θ
d
r
1
d
d
dr
r
1
2
. (5.4.2)
Thus, replacing the time derivatives of equation (5.1.9) by equation (5.4.1) and
using equation (5.4.2) we can write
( )
)
r
(
f
r
m
mr
L
d
r
1
d
mr
L
3
2
2
2
2
2
≡
∂
Φ
∂
−
=
−
θ
−
. (5.4.3)
This is the so called orbit equation since its solution is r(
θ
), the orbit of the
particle. This equation is more amenable to solution if we re-write it by
substituting
r
1
u
≡
, (5.4.4)
so that
(
u
1
f
u
d
u
d
m
u
L
2
2
2
2
−
=
⎟
⎟
âŽ
⎞
⎜
⎜
âŽ
⎛
+
θ
)
. (5.4.5)
The quantity f(r) or f(l/u) is the force law, which for gravity is
2
2
GMmu
r
GMm
)
r
(
f
−
=
−
=
. (5.4.6)
Thus the orbit equation for the gravitational force takes the following relatively
simple form:
.
const
L
GMm
u
d
u
d
2
2
2
2
=
β
≡
=
+
θ
(5.4.7)
This is a second order equation so we can expect two constants of integration in
the most general solution. As is customary with differential equations of this
form, we can guess a solution to be
(
)
β
+
θ
−
θ
=
0
cos
A
u
, (5.4.8)
which we can re-write as
68
69
)
]
(
[
0
cos
e
1
P
r
θ
−
θ
+
=
, (5.4.9)
so long as
⎪
⎪
âŽ
⎪
⎪
⎬
⎫
=
=
β
=
2
2
2
2
GMm
AL
e
GMm
L
1
P
. (5.4.10)
The reason for this last transformation is that equation (5.4.9) is the general
equation for a conic section with a focus at the coordinate origin and an eccentricity e.
Thus we recover the essence of Kepler's first law, namely that the planets move in
ellipses with the sun at one focus. Certainly ellipses are conic sections, and should the
mass of the sun greatly exceed the mass of the earth, then the sun may be regarded as the
source of the central force of gravity. We have now only to decide which type of conic
section the orbit will be and on what parameter the kind will depend.
We can make this determination by generating the total energy from the
solution to the orbit equation. First differentiate the solution with respect to time
and after some algebra find the velocity to be given by
(
)( )
(
)
[
]
.
r
P
2
1
e
P
m
L
v
2
2
2
2
2
+
−
=
(5.4.11)
Now form the kinetic energy and add the potential energy to get
(
)
( )
1
e
P
GMm
E
2
2
1
=
. (5.4.12)
It is clear from equation (5.4.12) that the sign of the total energy will determine
the sign of the eccentricity e and hence the type of orbit namely
⎪
⎪
âŽ
⎪⎪
⎬
⎫
<
⇒
<
=
⇒
=
>
⇒
>
orbit
elliptic
:
1
e
0
E
orbit
parabolic
:
1
e
0
E
orbit
hyperbolic
:
1
e
0
E
2
2
2
. (5.4.13)
Thus we see that Kepler's first law of planetary motion implies that the total
orbital energy of the planets is negative so that the planets are bound. In addition,
any bound test particle in orbit about the sun will have an elliptic orbit.
Chapter 5: Exercises
1.
Given that a particle is moving under the influence of a central force of the
form
3
2
r
c
r
k
f
+
−
=
,
where k and c are positive constants. Show that the solution to the orbit
equation can be put in the form
( )
( )
αθ
+
−
=
cos
e
1
e
1
a
r
2
,
which is an ellipse for e < 1 and
α
= 1. Discuss the character of the orbit for
and e < 1. Derive an approximate expression for
α
in terms of the
dimensionless parameter
γ
= [c/(ka)].
1
â‰
α
2.
Discuss the motion of a test particle moving in a potential field of the form
Φ
(r) = (
α
/r) + (
β
/r
3
) ,
in terms of the rotational potential and conservation of energy.
3.
A particle moves in a circular orbit of radius under the influence of a
central force located at some point inside the orbit. The minimum and
maximum speeds of the particle are
and
respectively. Find the orbital
period in terms of these speeds and the radius of the orbit.
0
r
1
v
2
v
4.
Suppose that all the planets move about the sun in circular orbits under the
influence of an inverse cube force law. Assuming conservation of
momentum and energy, find a relation between the orbital period and the
radius of the orbit for the planets. (i.e. a new "Kepler's third law").
70