Solution Manual For Classical Mechanics Second Edition - Tai L. Chow

Download as pdf or txt
Download as pdf or txt
You are on page 1of 209

SOLUTIONS MANUAL FOR

CLASSICAL
MECHANICS
SECOND EDITION

by
Tai L. Chow
SOLUTIONS MANUAL FOR
CLASSICAL
MECHANICS
SECOND EDITION

by
Tai L. Chow

Boca Raton London New York

CRC Press is an imprint of the


Taylor & Francis Group, an informa business
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2013 by Taylor & Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Printed on acid-free paper


Version Date: 20130918

International Standard Book Number-13: 978-1-4822-2780-2 (Ancillary)

This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and
information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and
publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission
to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any
future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic,
mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or
retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact
the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides
licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment
has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation
without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Preface
This Solutions Manual has been prepared to accompany the text Classical Mechanics
Second Edition. by Tai L. Chow. It is intended for use only by instructors using Classical
Mechanics as a textbook, and it is not available for sale to students.
A generous amount of detail has been included in the solutions, my aim, of course, was to
spare the instructor some time and effort. I accept the responsibility for any errors that may have
eluded me and would appreciate notification of whatever errors you find as well as suggestions
for improving the solutions.
This Solutions Manual only cover for chapters from 1 to 13; solutions to chapters 14,15,
and 16 are forthcoming.

Tai L. Chow
Physics Department
California State University, Stanislaus
Chapter 9

9.1 Show that the superposition principle does not hold for a nonlinear differential
equation such as
d2x
2  cx  0
2
dt
where c is a constant.
Solution:
d2x
 cx 2  0 (1)
dt 2
Now if x1 (t )  and  x2 (t ) are two solutions of Eq.(1), then we have
d 2 x1 d 2 x2
2  cx1  0 2  cx 2  0
2 2
and
dt dt
We now add these equations (multiplied by arbitrary constants  1  and   2 ) and obtain
d2
( x   2 x2 )  c( 1 x12   2 x2 2 )  0 (2)
dt 2 1 1
We can identify the linear combination in the first term as x: x   1 x1   2 x2 , but there is
no such linear combination in the second term: c( 1 x12   2 x2 2 )  x 2 . Thus, we can not
put Eq.(2) into the original form, Eq.(1). In other words, the superposition principle does
not hold in this case.

9.2 Consider the system in Fig. 9.2. Assume that the tension T in the wire is linear with
its elongation  . If the initial tension is T0 , derive the equation of motion and lin-
earize the problem.
Solution:

Fig. 9.2
In Fig. 9.2 is a particle initially held at rest by two identical light elastic strings of
force constant k(along A0A’).The tension in each string is T.The particle is then pulled to
the side and released. When the displacement of the particle is x, The length of each
string is increased to a quantity L, the tension then being given by
T  T0  k ( L  B)
The equation of motion of the particle is
x
mx  2T sin   2[T0  k ( L  B)]
L

132
Now L  x 2  B 2  B 1  ( x / B) 2 , and the equation of motion becomes


mx  2 T0  kB( 1  ( x / B) 2  1)  Bx 1
1  ( x / B) 2
which is a difficult nonlinear differential equation. If x < B, it is possible to find an ap-
proximate equation of motion which is less formidable. To this end, let us expand the
radicals, on the right-hand side of the above equation, by means of binomial theorem to
yield
 1 x2 1 x4 
1  ( x / B)  1 
2
2  4 .......
 2B 8B 
If we neglected power of x higher than the third, the equation of motion reduces to
2T0 kB  T0 3
mx  x x  0.
B B3
The case represented by the above equation of motion is physically similar to that
of a spring system as shown in Fig. 9.2b, where a particle is initially held at rest by two
identical springs of natural length B. Assuming that there is no gravity, each spring is
stretched an amount b.

Fig. 9.2b
Let the particle be pulled to side and released. When the particle is displaced by x,
the actual amount that the spring is stretched is
x 2  B 2  ( B  b)
The potential energy of the system is
1
V ( x)  2  k[ x 2  B 2  ( B  b)]2
2
The equation of motion is found from mx  dV / dx with the result
2 kx[b  B  x 2  B 2 ]
mx  0
x2  B2
which is a difficult nonlinear differential equation. If x is less than B, we can expand the
radicals to get an approximate equation of motion. To this end, let us first rewrite the po-
tential energy as
V ( x)  k[ x 2  B 2  2( B  b) x 2  B 2  ( B  b) 2 ]
and then expand the radical on the right side by means of the binomial theorem to yield

133
 1 x2 1 x4 
x 2  B 2  B 1  ( x / B) 2  B1   .......
 2B 
2
8 B4
To this approximation, the potential energy is, after rearrangement of terms
kb 2 k ( B  b) 4
V ( x)  kb 2  x  x
B2 4B3
and the approximate equation of motion is
2 kb k ( B  b) 3
mx  x x  0.
B B3

9.3 Construct a phase diagram for the potential in Fig. 3.12 of Chapter 3.
Solution:
Using the general procedure explained in the text, the phase diagram is con-structed
as follows.

Fig. 9.3

9.4 If the amplitude of the motion of a pendulum is not small, the free horizontal mo-
tion may be represented approximately by
g g
x
 x  3 x3  0
b 2b
where the notation is that of Fig. 7-2. Verify this equation. If a force F0 cos t is
also acting, find an approximate steady-state expression for x(t), considering (g/b)x3
to be small.

Fig.9.4
Solution:
(a) The forces acting on m are the tension T in the string and the weight mg. If x
is

134
measured positively to the right, the equation of motion along the x direction is
mx  Tx
and Tx  (mg sin  ) cos

b 2  y 2   x 2 
1/ 2
x y
Now sin   , cos    1    
b b b  b  
and the equation of motion along the x direction becomes
g  1 x2 
x
 x1  
b  2 b2 
g g
or x
 x  3 x 3  0. (1)
b 2b
(b) If a force F0 cos t is also acting, the above equation becomes
g g F0
x  x  3 x3  cos t
b 2b m
o x   x   x 3  A cos t (2)
g g F0
where     3 A
b 2b m
We now seek an approximate steady-state expression for x(t), considering  to be
small. We assume, as a first approximation, that
x1  B cos t
Substituting this in the right side of Eq.(2), we obtain
x2   B cos t   B 3 cos3  t  A cos t (3)
Employing the trigonometric identity
3 1
cos3  t  cos t  cos 3 t
4 4
Eq.(3) becomes
 3  1
x2   B   B 3  A cos t   B 3 cos 3 t
 4  4
Integrating twice, we have, as a second approximation to Eq.(2),
1  3   B3
x2  3  B   B  A cos t 
3
cos 3 t
  4  36 2
where the integration constants are taken to be zero.

9.5 Lord Rayleigh used the equation


x  (a  bx 2 ) x  0 2 x  0
in his discussion of nonlinear effects in acoustic phenomena. Show that this equation
can be converted to the following equation (Van der Pol’s equation)

135
a
y
 ( y 2  y 2 ) y  0 2 y  0, with y  y0 3b / a x
y0 2 0
Solution:
Differentiation of Rayleigh’s equation yields
 x  0 2  0
x  (a  3bx 2 ) (1)
The substitution
3b
y  y0 x
a
implies that
a y a y a 
y
x  , x
 , x

3b y0 3b y0 3b y0
When these expressions are substituted in Eq.(1), we find
a 
y a 3ba y 2  y a y
 a  2   0 2 0
3b y0 3b  3b y0  y0 3b y0

Multiplying by y0 3b / a and rearranging, we arrive at Van der Pol’s equation


a
y  2 ( y0 2  y 2 ) y  0 2 y  0.

y0

9.6 Consider the equation of motion of a particle


x   x   x 3  0;   0,   0
Sketch the potential function. Find a Fourier series solution for the case where en-
ergy has a small positive value.
Solution:
V 1 1
The potential energy per unit mass is    x 2  x 4 . A variety of cases can
m 2 4
occur depending on the signs of   and   . If   0  and    0, is as sketched in Fig.
9.6a. The only type motion which can occur is periodic. This is sometimes referred to a
hard spring. If   0  but    0, the potential function  is like Fig. 9.6b, and both
periodic and non-periodic motion can occur. This case is often referred to as a soft spring.

Fig.9.6a Fig.9.6b

For the periodic motion, a Fourier series representation of the solution exists:

x(t )  a0   (a n sin nt  bn cos nt ) (1)
n 1

136
Many of the coefficients in Eq.(1) can be shown to be zero by using the symmetry
proper-ties of the motion. The symmetry of the potential function means that the average
position of the particle is zero. Therefore
a 0  0x(t )d (t )  0
1 2

The shape of x(t) for x < 0 must be exactly the same as for x > 0. For simplicity
and without loss of generality, let us measure the time at an instant when the velocity
vanishes: x(0)  0. Under this initial condition, the displacement x(t) must be a sym-
metric function of time t. To see why, we first sketch the quantitative form of x(t) as a
function of t , and we examine the equation of motion which is not affected by a re-

Fig. 9.6c
versal of the direction of t. This means that, for example, the solution for    t  2 is
a mirror reflection of the solution for 0   t   . And

 x(t ) sin td ( t )  0 ;
1 2
a1 
 0

it is evident that the integral over    t  2 cancels that over 0   t   as shown in


Fig. 8.6c. By a similar argument, all the coefficients of sine terms, a n , can be shown to be
zero. Fig. 9.6d shows the first cycle of x(t) and also cos2 t . The coefficient b2 is given b

Fig. 9.6d

 x(t ) cos 2 td ( t )
1 2
b2 
 0

a moments of inspection shows that the contribution to b2 from  / 2   t   cancels


that from 0   t   / 2 . Thus b2  0 . Similarly, b4  b6 ....  0. It can be concluded
that the periodic solution for the potential functions of the form given in Fig. 8.6a and b is
of the form
x(t )  b1 cos t  b3 cos 3 t  b5 cos5 t ........
provided that x  0 at t  0. To determine the coefficients b1 , b3 , b5 , and so on, let us
sub-stitute this solution into the differential equation and neglect the fifth and higher
harmonics, we find
b1 (   2 ) cos t  b3 (  9 2 ) cos 3 t
  (b1 cos t  b3 cos 3 t ) 3  0

137
which can be rewritten in a convenient form by using the following trigonometric identi-
ties
1 3
cos 3  t cos 3 t  cos  t
4 4
1 1 1
cos 2  t cos 3 t = cos 5 t  cos 3 t  cos  t
4 2 4
the final result is
 3 3 
b (  2
)   b 3
  b 2
b cos t
 1
4 1
4 1 3

 1 3 
 b3 (  9 2 )   b13   b12 b3  cos 3 t  0
 4 2 
where the fifth harmonic has been neglected. In order for the above equation to be an
identity for all values of t, it is necessary that the coefficients of harmonics cos  t and
cos 3 t separately vanish. Therefore
3 3
b1 (   2 )   b1 3   b12 b3  0
4 4
1 3
b3 (  9 2 )   b1 3   b12 b3  0
4 2
which gives two relations for three unknowns: b1 , b3 and  . Let us determine b3 and 
in terms of b1 . Neglecting last term in the first relation (which is 5th order) gives the
frequency
3
 2     b12
4
Substituting this into the second relation yields b3
 b13  b13  21 2
 b13
b3   1   b  .
 21  32  32 1  32
32 1   b12 
 32 

9.7 If a simple pendulum is oscillating in a medium and experiences a resistive force


proportional to the velocity, then its equation of motion becomes
g
     sin   0
b
where b is the length of the pendulum and  is the coefficient of friction. Discuss the
stability of such a pendulum near its equilibrium point.
Solution:
It is more convenient to rewrite the equation of motion of the pendulum as a pair
of first order equation. To this end, let us put    . Then we can write
g
   ,       sin 
b
as the equations of motion of the pendulum. The singular points are obtained by putting

138
  0 and   0. These are
  0,   0, n ; n  1,  2, ....... .
So, defining ( , ) as these points we get
 

(  ,   )  (0, n ); n  0,  1,  2,........
Let us now examine the stability near the singular point (0, 0). We put
     x,     y
where x and y are small quantities whose squares and higher powers may be neglected.
Thus, we linearize the equations of motion near the origin and obtain
g
y  x, x    x  y
b
g
or y   y  y  0

b
The general solution of this equation is of the form exp(  t ), and  satisfies the
characteristic equation:
g
2     0
b
The roots are
   2  4 g / b
 
2
We see that so long as  > 0, the roots have always a real negative part and hence
the solution for  goes to zero exponentially as t  . In other words, the pendulum
executes a stable motion near the origin, a well-known fact.

9.8 Show that a secular term arises in the first-order solution to the equation
x  0 2 x   x 2

where  is a small quantity. Remove the secular term from the solution by using
the Bogoliuboff-Kryloff procedure.
Solution:
Let us attempt a perturbation solution by writing, to first order
x(t )  x0    x1
The equation of motion then separates into
x0  0 2 x 0  0
x1  0 2 x1  x 0 3
It is obvious that the zero-order solution is of the form
x0 (t )  A cos0 t
then the first-order equation becomes
1 3
x1  0 2 x1  A3 cos 3 0t  A (3 cos 0t  cos 30t )
4

139
The solution to this equation is
A3
x1 (t )  (120 t sin 0 t  cos 30 t )
320 2
The first term in the parenthesis is proportional to the time and so the secular term arises
in the first-order solution.
To remove the secular term, let us use Bogoliuboff-Kryloff procedure. We expand
x(t) and the frequency:
x (t )  x0  x1  2 x2 .....
0 2   2   1    2 2 ....
Substituting these expansions into the equation of motion and equating to zero the coeffi-
cients of like powers of  , we have
x0   2 x 0  0
x1   2 x1  x 0 3   1 x 0
x2   2 x 2  3x 0 2 x1   1 x1   2 x 0
The initial conditions are
x(0)  A; x(0)  0
which, in view of the expansion, lead to
x0 (0)  A, xi (0)  0; x 0 (0)  0, xi (0)  0; i  1, 2, 3,....
The zero-order equation describes a simple harmonic motion, its solution is
x0 (t )  A cos t
which satisfies the initial conditions. Substituting this zero-order solution the first-order
equation we have
x1   2 x1  A 3 cos3  t   1 A cos t
3 1
 ( A 3   1 A) cos t  A 3 cos 3 t
4 4
It is the first term (  cos t ) on the right side that gives rise to the secular term in the
first-order solution. To eliminate this troublesome term, we require
3
 1  A2
4
then the first-order equation reduces to
1
x1   2 x1  A 3 cos 3 t
4
its solution satisfying the initial conditions is
A3
x1 (t )  (cos t  cos 3 t )
32 2
and the complete solution, correct to first order, is
  A2   A3
x (t )  1  2  A cos t  cos 3 t
 32  32 2
Also, correct to first order, the frequency is

140
3 2
0  ( 2   1 ) 1/ 2  ( 2  A )
4
or
 3 A2  
1/ 2
3 A2 
  0 1     (1  ).
 4 0  8 0 2
2 0

This procedure can be extended to second order approximation (to any desired
order, in fact).

9.9 Consider the motion of the an-harmonic oscillator whose potential energy is
1 2 1
kx  m x 3
V ( x) 
2 3
under the combined action of the two sinusoidal driving forces
F  F1 cos1t  F2 cos2 t
Show that the solution of the equation of motion contains sinusoidal terms whose fre-
quencies are equal to the sum or difference of the driving frequencies. This phenome-
na known as intermodulation. What are possible values of 1 and 2 for resonance?
Solution:
This problem can be solved by using either perturbation or successive approxima-
tion. But it is very tedious, so we just list answers below:
F1 cos1t F2 cos2 t
x0  2 
m(0  1 ) m(o 2  2 2 )
2

  F1 2  F2 2
x1  
2m 0 2 ( 0 2  1 2 ) 2 2m 0 2 ( 0 2   2 2 ) 2
 F12 cos 21t  F2 2 cos 22 t
 
2m( 0 2  41 2 )( 0 2  1 2 ) 2 2m( o 2  4 2 2 )( 0 2   2 2 ) 2
 F1 F2 cos(1  2 )t

m[ 0  (1   2 ) 2 ]( 0 2  1 2 )( 0 2   2 2 )
2

 F1 F2 cos(1  2 )t

m[ 0  (1   2 ) 2 ( 0 2  1 2 )( 0 2   2 2 )
2

9.10 Solve the following equation of motion for an-harmonic oscillator


mx  kx   x 2  0
by the method of successive approximations and the method of perturbations.
Solution:
(a) Method of successive approximations
Considering the  x 2 term small, we first omit it, and obtain
x0  kx0  0

which describes a simple harmonic motion and its solution may be written as

141
x0  A cos(0 t   ), 0  k / m
This is solution to a first approximation. We now put this value of x in the small  x 2
term of the equation of motion, obtaining
mx1  kx1   A2 cos2 (0 t   ) = 0
Using the trigonometric identity cos 2  2 cos2   1, we can rewrite the above equation
as
1
mx1  kx1   A 2 [1  cos 2(0 t   )] = 0
2
For its solution, we try
x1  A cos(0 t   )  B cos 2(0 t   )  C
Substituting this into the last equation, we have
1 1
3kB cos 2(0 t   )  kC   A 2   A 2 cos 2(0 t   )  0
2 2
This last equation must be an identity for all values of t, hence we must have
 A2  A2
C , B
2k 6k
and our second approximation to the solution then is
 A2  A2
x1  A cos(0 t   )  cos 2(0 t   ) 
6k 2k
The integration constants A and  are determined by the initial conditions for a particular
problem.
(b) Method of perturbations
We expand the solution in a series involving the small parameter:
x  x0   x1  2 x2 ......
In the following discussion, we retain only the first two terms. Substituting this into the
equation of motion, we get
x0   
m( x1 )  k ( x0   x1 )   ( x0   x1 ) 2
The above equation must be an identity for any value of and this, in turn, necessitates
equating the coefficients of each power of on the left side to the coefficients of the
same power on the right side of the last equation. Hence we have
mx0   kx0
mx1   kx1  x0 2
The first equation is that for SHM and its solution is, as part (a)
x0  A cos(0 t   ), 0  k / m
Substituting this into the second equation, we have
mx1   kx1  A 2 cos2 (0 t   )
1 2
  kx1 A [1  cos 2(0 t   )].
2
The same equation was found in part (a). The two methods thus give the same result. But

142
the perturbation method is more elegant, and using as a label for small quantities makes
it easier to distinguish different orders of small quantities, particularly in further
approximations.

9.12 Use the ritz method to analyze the forced oscillations of a harmonic oscillator in-
corporating a nonlinear spring whose deflection characteristic is a cubic
my  ky   y 3  F0 sin t .
Solution:
According to Hamilton’s principle, a conservative, holonomic system always be-
haves such that the time integral of its Lagrangian is minimal. But this statement is not
applicable directly to our system, due to the presence of the nonconservative forcing
func-tion. The generalized Hamilton’s principle for nonservative, non-holonomic systems
may be stated as
I  t (L  Qq)dt  0,
t2
q(t1 )  q(t 2 )  0
1

A simple proof of this statement is provided at the end of the solution. From this general-
ized statement we infer that the functional I of a nonconservative system takes the form

I  t ( L  Qq)dt
t2

Applying this to our system, we have


1 2 1 2  4 
L  T V  my   ky  y 
2 2 4 
and Q  F0 sin t
The functional (we call action integral in the text) describing our system is then given by

I  t ( L  Qy )dt
t2

t2  m  
  t  y 2  y 2  y 4  ( F0 sin t ) y dt
k
1  2 2 4 
We now come to the crucial step of selecting a suitable trial function. For our problem, it
is reasonable to assume a periodic response at the forcing frequency. The non-linearty in
the spring may lead to distortion, phase shift, etc. Let us adopt a simple trial which ignore
these secondary effects:
y  a sin t
Before we substitute this and its time derivatives into the functional we need to set the in-
terval so that the variation in y vanishes at the end points, we choose an interval equal to
one period. Thus, we have
m  
I  0
2  / k
( a cost ) 2
 ( a sin t ) 2
 ( a sin t ) 4
 F sin 2
t dt
 2 2 4 0

Integration gives
 3 4 F0
I (m 2  k )a 2  a  a
2 16 
We now evaluate a by requiring that

143
I 3
 (m 2  k )a  a 3  F0  0
a 4
This cubic equation will yield either one or three real roots for the response amplitude a,
depending on the system parameters.
______________________________________________
On Generalized Hamilton’s principle
For simplicity, we consider a system describable by a discrete set of generalized
coor-dinates and having a Lagrangian L  L(q, q, t ) , its variation can be written as
L L
L  q  q
 q q
Note that time variation doesn’t exist explicitly, since time is the independent variable.
Now
d  L   L d d  L 
  q  ( q )    q
dt   q   q dt dt   q 
L d  L 
  q    q
 q dt   q 
from which we get
L d  L  d  L 
 q    q     q
 q dt   q  dt   q 
Substituting this into the variation of L, we obtain
 d  L   L  d  L 
 L     q    q
 dt   q   q  dt   q 
Integration with respect to time gives
 d  L   L  d  L 
t12 ( L   dt   q    q  q)dt  t12 dt   q  q dt
t t
(A)
 
The right side vanishes:
t d  L  L t
t12 dt   q  q dt   q  q t2  0   q(t 2 )   q(t1 )  0
1
The expression inside the square brackets on the left side of Eq.(A) vanishes for
conserva-tive, holonomic systems but equals the generalized force Q in the case of
systems which are nonconservative. Hence we extend (or generalize) Hamilton’s
principle to read
 I  t ( L  Q q)dt  0
t2

_____________________________________________________

8.13 A particle of mass m is placed on the inside of a smooth paraboloid of revolution


having equation cz  x 2  y 2 at a point P which is at height h above the horizontal
plane (the xy-plane). Assuming that the particle starts from rest. Find
(a) the speed of the particle at the vertex 0;
(b) the time taken to reach 0;
(c) the period for small oscillations. If the oscillations are not small, show that

144
the period is given by
c  4h
  2 {1  (1 / 2) 2 k 2  (1  3 / 2  4) 2 k 4 / 3
2g
 (1  3  5 / 2  4  6) 2 k 6 / 6 ........}
Solution:

Fig. 9.13
(a) As shown in Fig. 9.13, we choose the point P in the yz-plane so that x  0
and cz  y 2 . By the principle of conservation of energy we have if Q is any point on the
path PQ0,
(P.E. + K.E.) at P = (P.E. + K.E.) at Q
1 1
mgH  m(0) 2  mgz  m(ds / dt ) 2
2 2
where s is the arc length along OPQ measured from 0. Thus
(ds / dt ) 2  2 g( H  z), or ds / dt   2 g( H  z)
using the negative sign since s is decreasing with t.
Putting z= 0, we see that the speed at the vertex is 2gH .
(b) We now rewrite ds/dt in terms of x, y, and z. Since x  0, and cz  y , we have
 ds   dx   dy   dz 
2 2 2 2

        
 dt   dt   dt   dt 
 dy  4 y 2  dy   4 y 2  dy 
2 2 2

    2    1  2  
 dt  c  dt   c  dt 
Combining this with the result of part (a), we obtain
 4 y 2  dy 2
1  2    2 g ( H  y / c)
2

 c  dt 
from which we obtain
c2  4 y 2
 2 gcdt  dy
cH  y 2

Integrating, using the fact that z  H and y  cH at t  0 while y  0 at t  


c2  4 y 2
  2 gcdt   cH
 0
dy
0
cH  y 2

145
Letting y  cH cos , the integral can be written as
c  4 H 2
 
2g
0
1  k 2 sin 2  d , with k  4 H / (c  4 H ) < 1

The integral in the last equation is an elliptic integral and can be evaluated in terms of
series but not in terms of elementary functions.
(c) The particle oscillates back and forth on the inside of the paraboloid with
period given by
P  4
For small oscillations the value of k can be very small, if k is so small so as to be zero for
practical purposes, then
c  4 H 2
P  4  4
2g 0
 1  0  sin 2  d  2 (c  4 H ) / 2 g .

9.14 A very important type of nonlinear equation was extensively studied by Van der Pol
in connection with relaxation oscillations in vacuum tube circuits. The Van der Pol
equation may be written as
x   (1  x 2 ) x  x  0
where  is small, positive parameter. Solve the Van der Pol equation by the pertur-
bation method.
Solution:
We write x  x0   x1   2 x2   3 x3 ......
Substituting in the Van der Pol equation and equating the coefficients of like powers of 
we obtain
x0  x0  0 (1a)
x1  x1  ( x0 2  1) x0  0 (1b)
x2  x2  2 x0 x1 x0  ( x0 2  1) x1  0 (1c)
A solution of Eq.(1a) is
x0  A cos t (2)
where we have chosen the phase to be zero. Substituting this into Eq.(1b) we get
 A2  A3
x1  x1    1 A sin t  sin 3t
 4  4
The first term on the right hand side produces a secular term in the solution, we eliminate
it by requiring A=2. Then, a solution of this differential equation becomes
1
x1  B sin t  sin 3t (3)
4
Substituting Eqs.(2) and (3) into Eq.(1c) and simplifying we get
9  5  5
x2  x2    2 B  cos t    B  cos 3t  cos 5t
4  4  4

146
Again, the first term on the right hand side produces a secular term in the solution, we
can eliminate it by requiring B = 9/8. Then the solution is
19 5
x2  C cos t  cos 3t  cos 5t (4)
64 96
Thus, an approximate solution of Van der Pol equation is
9 1 
x (t )  2 cos t    sin t  sin 3t 
8 4 
 19 5 
  2  C cos t  cos 3t  cos 5t   .......
 64 96 

147
Chapter 10

10.1 This is a long problem, see text p.333 for the detail.
Solution:
(a) The equations of motion are
dp1 dp2
m1  F, m2  F
dt dt
from which we have
dp1 dp2
m1  m2 0
dt dt
Integration yields the law of conservation of momentum
p1  p2  P  cons tan t .
(b) When the mutual force F is a function of x only, it is conservative, and we can
in-troduce a potential energy function V(x) as before. The law of conservation of energy
takes the form
T + V = constant,
With
1 1
T  m1 x12  m2 x 2 2 .
2 2
(c) If we denote the initial velocities by u1 , u2 , and the final velocities by v1 , v2 , then
we have,
1 1 1 1
m1v12  m2 v2 2  m1 u12  m2 u2 2
2 2 2 2
from which we have
1 1 1 1
m1v12  m1 u12  m2 u2 2  m2 v2 2 . (1)
2 2 2 2
And from law of conservation we have
m1v1  m1u1  m2 u2  m2 v2 (2)
Dividing Eq.(1) by Eq.(2) we obtain
v1  u1  u2  v2 , or v2  v1  u1  u2 (3)
Eq.(3) shows that the relative velocity is just reversed in the collision. We may now solve
Eqs.(3) and (2) for the final velocities in terms of the initial velocities. For the case u2  0
we find
m1  m2 2m1
v1  u1 , and v2  .
m1  m2 m1  m2
10.2 A mass m strikes a smooth flat surface as shown in Fig. 10.2. If the coefficient of
restitution is e, determine the angle of rebound  , the final velocity v f , and the en-
ergy loss per impact in terms of initial velocity u and the angle of incidence  .
Solution:

148
Fig. 10.2
We assume that the velocity of the flat surface is zero (with which is associated a
very large mass). Furthermore, since the surface is smooth (frictionless), there will be no
component of force acting on the mass in the horizontal direction during the impact.
Therefore, ux  v x or
u cos  v cos  (1)
In the vertical direction, the coefficient of restitution becomes, with the velocity of the
flat surface to be zero before and after impact,
( velocity of separation) v sin 
e 
velocity of approach u sin 
or e u sin  v sin  (2)
dividing Eq.(2) by Eq.(1) we find that the angle of rebound is
e tan  tan 
and we see that only if the collision is perfect elastic (e = 1) will the angle of rebound
equal the angle of incidence.
The magnitude of the final velocity is

v  v x 2  v y 2  (u cos ) 2  (eu sin  ) 2


and we see that v  u only if the collision is perfectly elastic (e = 1).
The loss of kinetic energy per collision is
1
T  m(u 2  v 2 )
2
1 1
 m[u 2  (u cos ) 2  (eu sin  ) 2 ]  mu 2 sin 2  (1  e 2 )
2 2
When e = 1 (perfect elastic), the loss of kinetic energy is zero; when e = 0 (perfect inelas-
tic), a maximum loss of kinetic energy occurs.

10.3 Two spheres of mass m and M respectively, collide obliquely. Find their velocities
after impact in terms of their velocities before impact.
Solution:

Fig. 10-3

149
   
Let u , U and v , V be the velocities of the spheres before and after impact respect-
tively, as depicted
 in Fig. 10-3. Choose a coordinate system so that the xy plane is the

plane of u & U ,and so that at the instant of impact the x axis passes through the centers
C1 and C2 of the spheres.
By the conservation of momentum, we have
   
mu  MU  mv  MV (1)
From Fig. 10.3 we see that
 
u  u(cos 1 i  sin  1 j ), U  U (cos 2 i  sin  2 j )
  (2)
v  v(cos  1i  sin  1 j ), V  V (cos  2 i  sin  2 j )
Substituting these in Eq.(1) and equating coefficients of i and j , we have
mu cos 1  MU cos 2  mv cos  1  MV cos  2
(3)
mu sin  1  MU sin  2  mv sin  1  MV sin  2
Applying the definition of restitution in the x direction, we have
velocity of separation = -e (velocity of approach)
   
or v  i  V  i  e(u  i  U  i) (4)
which on using Eq.(2) becomes
v cos 1  V cos  2  e(u cos 1  U cos 2 ) (5)
Furthermore, the tangential velocities before and after impact are equal:
   
u  j  v  j , v  j  V  j
or
u sin1  v sin 1 , U sin 2  V sin  2 (7)
From Eq.(5) and the first equation of Eq.(3) we find
(m  Me)u cos 1  M (1  e)U cos 2
v cos  1 
m M
m(1  e)u cos 1  ( M  me)U cos 2
V cos  2 
m M
Combining these with Eq.(7) we find

v  v (cos  1i  sin  1 j )
(m  Me)u cos 1  M (1  e)U cos 2
 i  u sin  1 j
m M

V  V (cos  2 i  sin  2 j )
m(1_ e)u cos 1  ( M  me)U cos 2
 i  U sin  2 j
m M

10.4 A gun is fired horizontally point-blank at a block of wood which is initially at rest
on a horizontal floor. The bullet becomes imbedded in the block, and impact causes

150
the system to slide a certain distance s before coming to rest. Given m (mass of the
bullet), M (mass of the block, and  (coefficient of sliding friction between the
block and the floor), find the muzzle velocity of the bullet.
Solution:
We first note that the coefficient of restitution e is zero in our case. Now let u and v
be the initial velocity of the bullet and the velocity of the system (block + bullet) immedi-
ately after impact, respectively. By conservation of momentum, we have
mu  (m  M )v
Next, the magnitude of the retarding frictional force is equal to (m  M ) g , and, by
New-ton’s second law, we have
(m  M ) g  (m  M )a, or a g
where a is the deceleration of the system after impact.
By using the kinematics relation v f 2  vi 2  2as , we have

v2
s
2 g
since in our case, v f  0, vi  v and a    g (deceleration). But v  mu / (m  M ) ,
substituting this in the last equation for s we obtain
 mu   1 
2
v2
s     
2 g m M  2 g 
Solving for u we obtain
u  [(m  M ) / m] 2  gs
for the initial velocity of the bullet in terms of the given quantities.

10.5 A wedge of mass M rests on a horizontal plane and with the inclined face at an
angle  to the horizontal. It is hit on its inclined face by a particle of mass m mov-
ing along the plane with speed u in a direction perpendicular to the edge of the
wedge. Show that the speed of the wedge immediately after impact is
mu sin 2  / ( M  m sin 2  )
The particle then slides up the face of the wedge and back to the horizontal plane.
Find the final velocities of the particle and the wedge. It is to be assumed that there
are no frictional forces and that the impact occurring are inelastic, in other words
relative motion in the direction of the impact is destroyed.
Solution:

Fig. 10.5a Fig. 10.5b


Since the impact between the particle and the wedge is inelastic, there will be no re-

151
bound and the particle will proceed to slide up the face of the wedge with initial speed u1.
Let the initial speed of the wedge be u2 . The momentum of the particle parallel to the
wedge face is conserved:
u cos  u1  u2 cos
Furthermore, the horizontal momentum of the system (particle + wedge) is unchanged:
mu  mu1 cos  (m  M )u2
Solving the above two equations, we find
Mu cos mu sin 2 
u1  , u2  (1)
m sin 2   M m sin 2   M
The next stage of motion of the particle is that it slides smoothly up the face of the
wedge and returns to its original position. The final speeds v1 of the particle relative to
the wedge and v 2 of the wedge can be determined by two conservation equations. Since
all forces acting on the system are vertical, the horizontal momentum of the system is
conserved.
Thus
mu  mu1 cos  (m  M )u2  mv1 cos  (m  M )v2 (2)
The energy is also conserved. Since there is no net change in potential energy for the
particle during the up and down motion ion, the energy conservation equation amounts to
conservation of kinetic energy:
1 1
m[(u2  u1 cos ) 2  u12 sin 2  ]  Mu2 2
2 2
(3)
1 1
 m[(v 2  v1 cos ) 2  v12 sin 2  ]  Mv 2 2
2 2
Eqs.(2) and (3) can be solved for v1 and v2 in terms of u1 and u2 and hence, by Eq.(1), in
terms of u. The results are
mu{(m  M ) sin 2   2 M }
v2 
(m  M )(m sin 2   M )
u(m 2 sin 2   mM  M 2 cos2  )
v1 cos  v 2 
(m  M )(m sin 2   M )

10.6 A particle of mass m lies at the middle, A, of a hollow tube of length 2b and mass
M. The tube, which is closed at both ends, lies on a smooth table. The coefficient of
restitution between m and M is e. If m be given an initial velocity u along the tube.
(a) Find the velocities of m and M after the first impact.
(b) Find the loss in energy during the first impact.
(c) Find the time required for m to arrive back at A traveling in the original direc-
tion.
Solution:
(a) Let v’ and V’, respectively, denote the velocities of m and M after the first colli-
ision. By conservation of momentum, we have

152
mv  M  0  mv' MV ' (1)
and from the definition of coefficient of restitution, we have
velocity of separation v'V '
e 
velocity of approach v
or
ev  v'V ' (2)
Solving Eqs.(1) and (2) we find
m  eM m(1  e)
v'  v V ' v. (3)
m M m M
(b) Initially the total kinetic energy is
1 2
T mv
2
After the first collision the total kinetic energy T’ is
1 1
T'  MV ' 2  mv' 2
2 2
where v’ and V’ are given by Eq.(3). Therefore, the change in kinetic energy is given by
1 1 1 2 1 mM (1  e 2 ) 2
T  MV '  mv'  mv  
2 2
v .
2 2 2 2 (m  M )
(c) We can use the definition of the coefficient of restitution to solve Part (c). The
ve-locities before and after the first collision are v and  ev . Similarly, after the second
colli-sion the relative velocity is e 2 v . Therefore, the total time elapsed until m is again
at A and traveling in the original direction is
b  1
2
b 2b b
t   2  1   .
v ev e v v  b 

10.7 (a) In the system of Prob.10.6, determine the velocities relative to the center of mass
before and after the first collision.
(b) Find the loss in energy during the first impact in terms of a reference system
with origin at the center of mass.
(c) How far has the center of mass traveled during the time mass m has arrived at A
(i.e., during the time given by Part (c) of Prob. 10.6).
Solution:
(a) Let v  and  v' be the velocities of m relative to the mass center before and after
the first impact and V  and  V ' the corresponding quantities for M. If v c denotes the velo-
city of the center of mass (which remains constant), we then have
m
vc  u
m M
M m
and v  u  vc  u V  u
m M m M

153
To find v’ and V’, let us go back to Newton’s rule and the law of conservation of
momentum. Since the center of mass is not being accelerated, the equations of motion in
a center of mass system are the same as that in an inertial laboratory system, i.e.
dp1 / dt   F dp2 / dt  F
Consequently the same subsequent procedure may be carried out to give us Newton’s
rule:
v'V '  e(v  V )
Of course, here the velocities are with respect to the center of mass. Relative to the center
of mass the conservation of momentum is expressed by
mv  MV  0 mv' MV '  0
These may be combined with Newton’s rule given above to yield
v'  ev V '  eV
In terms of u they become
eM eM
v'   u V ' u
m M m M
(b) The initial kinetic energy relative to the center of mass is
1 2 1 1 mM 2
Tr  mv  MV 2  u
2 2 2 m M
The final value of the kinetic energy is
1 1 1 e 2 mM 2
Tr '  mv' 2  MV ' 2  u
2 2 2 m M
And so the change in the kinetic energy is given by
1 mM
T ' r  T ' r Tr   (1  e 2 )u 2  Tr (1  e 2 )
2 m M
it is identical to the result of Prob.10.6.
(c) When the number of collisions becomes very large, the final result is that the
kinetic energy relative to the center of mass approaches zero, leaving only kinetic energy
of the translation of the mass center. This is
1
Tc  (m  M )v c 2
2
which is just (mu 2 / 2)  Tr . The distance through which the center of mass moves is
mb(1  e) 2 / (m  M )e 2 .

10.8 In the elastic scattering of neutrons by protons (mn  mp ) at relatively low energies,
it is found that the energy distribution of the recoiling protons in the laboratory sys-
tem is constant up to a maximum energy that is the energy of the incident neutrons.
Find the angular distribution of the scattering in the C.M. (center of mass) system.
Solution:

154
Fig.10.8
In the C.M. system, whenever the neutron is scattered through the angle  , the pro-
ton recoils at the angle      . Thus, the neutron scattering cross section is equal to
the recoil cross section at the corresponding angles:
dN n dN p

d( ) d( )
Thus,
dN n dN p dTp

d( ) dTp d( )
where dN p / dTp is the energy distribution of the recoil protons. According to experiment,
dN p / dTp = const. Since mp  mn , Tp is expressed in terms of the angle  Then
Tp  T0 sin 2 ( / 2)
Thus,
dTp 1 d
 {T0 sin 2 ( / 2)}
d( ) 2 sin  d
or
dTp T0 d     T0   T0
 sin 2   sin cos 
d( ) 2 sin  d  2  2 sin  2 2 4
Therefore, we find for the angular distribution of the scattered neutron,
dN n dN p T0

d( ) dTp 4
Since dN p / dTp  const ., dN n / d is also constant. That is, the scattering of neutrons by
protons is isotropic in the C.M. system.

10.10 A beam of alpha particles, each of which has an energy of 4  106 ev, is projected
toward thin target gold foils.
(a) Compute the cross section for scattering through an angle of  / 2 or more.
(b) The target gold foils in Rutherford experiment had a thickness of ~5  105 cm.
An area of 1 cm2 , times that thickness, yields a volume density of about 1016 a-
toms. If this volume is bombarded with an incident flux of 10 4 alpha particles
per cm2  sec, how many backward scattering would we expect to observe in a
minute?
Solution:
(a) As Fig. 10.10 shows, for impact parameter b, the scattering angle is . If the par-
ticle enters the atom within the disc of area  b 2 , its scattering angle will be larger than

155
 . Now the relationship between b and  is
zZe 2  zZ e 2 
b cot  cot
8  0 K 2 2 K 4  0 2

Fig. 10.10
For scattering at  / 2 ( 900 ) or more, the impact parameter b can be found easily from
the above equation:
2  79
b . eV  nm) cot 450  7.0  1015 m
(144
2  (4.0  10 eV )
6

so  b 2  154
.  1028 m2 /nucleons;
where we have used the fact that K  4  106 eV , e 2 / 4  0  144
. eV  nm .
(b) A scattering target foil may contain thousands or tens of thousands of atoms. Let
t be the thickness of the foil and A its area, and let d and M be the density and molecular
weight of the material of which the foil is made. The volume of the foil is just At, its mass
is dAt, the number of moles is dAt/M and the number of atoms or nuclei per unit volume
is
dAt 1 N Ad
n  NA  N A  Avogadro' s number
M At M
Now the target is assumed to be so thin that all nuclei are exposed to the bombarding par-
ticles. Thus, as seen by an incident particle, the number of nuclei per unit area is nt. If the
total number of incident particles is N, the fraction that scattered at angles greater than 
is
f  Nnt ( b 2 ) .

from which we find, with n  1016 atoms / cm3 , t  5  105 cm, N  104 / cm2  sec
f  7.7  109 per cm2  sec .

10.11 The interaction between an atom and an ion at distances greater than contact is
given by the potential energy
C
V (r )   4
r
[ C  (e 2 / 2) P 2 , e is the ion charge and P is the polarizability of the atom.] Make a
sketch of the effective energy vs. the radial coordinate. If the total energy of the ion
exceeds the maximum value of the effective potential energy, the ion spirals inward
to the atom. Find the cross section for an ion of velocity v 0 to strike an atom. You
can assume that the ion is much lighter than the atom.
Solution
C 2
Veff   
r 4 2mr 2

156
dVeff 4C 2
0 
dr r0
r0 5 mr0 3
from which we find
4mC
r0 2 
2
From Veff  V0 at r  r0

4
we obtain V0  .
16m2 C
1
To reach r = 0, the initial energy E 0  mv 0 2 at great distance must be greater than V0 :
2
4
E 0  V0 or E0 
16m2 C
Now   mv0b  2mE 0 b
(2mE 0 ) 2 b 4
and so E0 
16m2 C
Solving for b we find
b 2  2 C / E0 .
Then the cross section for particles penetrating to r = 0 is
2
   b 2  2 C / E 0  2C / m .
v0

10.12 A uniform beam of particles each of mass m and energy E is scattered by a fixed
center of force. For a repulsive force F  K / r 3 , find the differential scattering
cross section. What will be the total cross section? Give physical discussion on its
behavior.
Solution:
K
The potential for the given force law is V (r )  .
2r 2
We first make a change of variable, z  1/ r. Then, from Eq.(10-48), we have
z max  b
  0
zmax bdz b z
 sin 1 
K K z max 0..... K
1  (b 2  )z 2 b2  2 b2 
mv 0 2 mv0 2 mv 0 2
Solving for b,
K 2
b( ) 
mv0 2   4 2
2

According to Eq.(10-47)
  (   ) / 2

157
so that b( ) can be rewritten as b( ):
K  
b( ) 
mv0 2  (2   )
The differential cross section can now be computed from
b db
 ( ) 
sin  d
with the result
K 2 (   )
 ( ) 
mv0 2 2 (2   ) 2 sin 
The total cross section is
2 K 3 (   )d
t  2 0 
 
 ( ) sin  d 
mv0 2 0  2 (2   ) 2
The integrand in above equation can be written as
  A B (2   ) A  B 2
  
 2 (2   ) 2  2 (2   ) 2  2 (2   ) 2
By comparing coefficients of powers of  in the leftmost and rightmost members of the
last equation, we can determine the values of A and B:
A  B  1/ 4
Therefore, the total cross section becomes
K 2  d d
2 [ 0 2  0

t  ]
2mv 0  (2   ) 2
K 2 1 1 
 2 [  ]
2mv 0  2   0
The total cross section then is infinite, because it diverges for  = 0. This result is ex-
pected, since the force has an infinite range. For a finite force, the situation will be dif-
ferent: only incident particles with impact parameter b less than the range of the force
will “feel” the force, while incident particles with impact parameter b greater than the
range of the force will continue their motions unaffected.

10.13 If, in Example 4 (scattering by a rigid sphere), the energy lost by the scattered par-
ticle to the sphere is  , show that
 R2
dcm ( )   cm ( )d  d
 max
Solution:
 
The initial and final momenta of the particle in the C.M. system are pi and p f , and
the change in momentum is
( p) 2  pi 2  p f 2  2 pi p f cos

158
 
Because the collision is elastic, the magnitude of pi and p f are equal:

pi  p f  2mT0 '
where T0 ' is the initial kinetic energy of the particle in the C frame. Thus
( p) 2  2  2mT0 '2  2mT0 'cos  4mT0 '(1  cos )
or

( p) 2  8mT0 'sin 2
2
Therefore, the energy loss of the scattered particle (or the energy transfer to the sphere) is
( p) 2 
E     4T0 'sin 2
2m 2
This energy loss (or the energy transfer) is maximum for    :
 max  4T0 '
so we can write

   max sin 2
2
  1 1
and d   max 2 sin cos d   sin  d
2 22 2 max
Now from Example 4, we have
R2 R2
d  d  2 sin  d (d  2 sin  d )
4 4
Thus
1 2
d 2 R  sin  d  R 2
  .
d 1  max
 max sin  d
2

10.14 A fixed force center scatters a particle of mass m according to the force law
F  K / rn , n  0.
If the initial velocity of the particle is u , determine the differential cross section for
small-angle scattering.
Solution:
We see that sine n is positive; we do have a weak field for large impact parameter b.
Eq.(58) gives
2b  nK
2 b 
dr 2bnK r dr
  n 1  2 (1)
mv0 r r 2  b2 mv0 b r n1 r 2  b 2
We now make a change of variable:
b2 2 2
z  2 , dz   3 b 2 dr   z 3/ 2 dr
r r b

159
This change of variable converts the integral in Eq.(1) to a beta function, which can be
ex-pressed in terms of gamma functions:
( x)( y)
 ( x, y)  0 t x1 (1  t ) y 1 dt 
1

( x  y)
Thus
n 1
( n 1)/ 2
2K   ( )
 dz 1  z  mv 2b n (n / 2)
nK 1 z 2
 
mv0 2 b n 0 0
Solving for b we find
n 1
2K   ( )
2 1
b 
n
mv 0 2 (n / 2) 
from which we obtain
 n 1 
2/n

d b db 1  n2 2  2 K  ( 2 ) 
    mv 2 (n / 2) 
d  d n
 0

 

160
Chapter 11

11.1 A uniform sphere is placed on the upper surface of a wedge and  the wedge is accel-
erated along a horizontal surface with a constant acceleration A. What must A be
in order that the sphere will remain at rest with respect to the wedge?
Solution:

Fig.11-1a Fig.11-1b

The force and pseudo-forces acting on the sphere are depicted in Fig.11-1b.
 N is
always perpendicular to the inclined plane of the wedge, the pseudo-force mA is in the
opposite direction of A . To prevent the sphere from rolling either up or down the wedge’s
inclined surface, the forces acting on the sphere must be zero:
F x  0, or N sin   mA
F y  0, or N cos  mg
From which we find
A  g tan  .

11.2 A bicycle travels with constant speed around a track of radius  . Find the accel-
eration of the highest point on one of its wheels.
Solution:

Fig.11.2
Let v 0 be the speed of the bicycle and b the radius of the wheel. We choose a
coordinate system with origin at the center of the wheel and with the x’ axis horizontal
pointing toward the center of curvature C of the track. Furthermore, we choose the z’ axis
remains vertical as shown in Fig. 11.2. Thus the 0’x’y’z’ system rotates with angular ve-
locity  which can be expressed as
 v
  k' 0


and the acceleration of the moving origin A0 is given by

161
 v0 2

A0  i '

Since each point on the wheel is moving in a circle of radius b with respect to the moving
origin, the acceleration in the 0’x’y’z’ system of any point on the wheel is directed
toward 0’ and has magnitude v0 2 / b . Thus, in the moving system we have

  v0 2
r '  k '
b
for the point at the top of the wheel. Also, the velocity of this point in the moving system
is given by

v '   j ' v0
so the Coriolis acceleration is
   v0   v0 2
2  v '  2 k '  ( j ' v0 )  2
 i'
  

Since the angular velocity  is constant, the transverse acceleration is zero. The centrip-
etal acceleration is also zero, because
   v2
  (  r ')  0 2 k'( k'bk')  0

Thus the net acceleration, relative to the ground, of the highest point on the wheel is
2 2
 v0 v0 
a3 i' k'
 b

11.3 A small mass hangs from a string in a car. What is the static angle of the string
from the vertical, and what is its tension when the car (a) moves with a constant
velocity, (b) moves with a constant acceleration?
Solution:

Fig.11.3
Let us analyze the problem both in an inertial frame and in a frame accelerating
with the car.
T cos  W  0 T cos  W  0
T sin   MA  0 T sin   F fict  0
F fict   MA
From which we find
MA A A
tan    tan  
W g g
T  M g 2  A2 T  M g 2  A2
From the point of view of a passenger in the accelerating car, the fictious force acts like

162
a horizontal gravitational force. The effective gravitational force is the vector sum of the
real and fictious forces.

11.4 A uniform cylinder of mass M and radius R rolls without slipping on a plank
which is accelerated at the rate A with respect to a horizontal surface (Fig. 11.4a).
Find the acceleration of the cylinder with respect to the plank and to the inertial
system.
Solution:

Fig.11.4a Fig.11.4b
The force diagram for the horizontal force on the cylinder as viewed in a system
which accelerates with the plank is shown in Fig.11-4b. a’ is the acceleration of the cylin-
der as observed in a system fixed to the plank. f is the friction force, and Ffict  MA with
the direction shown.
The equations of motion in the system fixed to the accelerating plank are
f  F fict  Ma '
Rf   I '
The cylinder rolls on the plank without slipping, so
 ' R  a'
These yield
a'
Ma '   I  F fict
R2
F fict
a'  
M  I / R2
Since I  MR 2 / 2, and Ffict  MA, we have
2
a '  A.
3
The acceleration of the cylinder in an inertial system is
1
a  A  a'  A .
3
Problem 11.3 and 11.4 can be worked with about the same ease in either an
inertial or an accelerating frame. But this is not always true. Many problems which are
rather complicated to solve in inertial systems, can be solved easily in accelerating
frames.

11.5 A bead slides on a smooth rod, the rod being constrained to rotate uniformly at an
angular velocity in a plane about an axis passing through one end perpendicular to
the rod. Find the position of the bead on the rod as function of time, and the reac-
tion of the rod on the bead. (Assume the axis is attached to an inertial system.)

163
Solution:
We choose a coordinate system rotating with the rod, with the origin at the inter-
section of the rod with the axis of rotation. The vector equation of motion is
      
mr  m  (  r )  2m  v  R
 
in which r is the radius vector, in the moving system,  from the origin to the particle, v
the velocity of the particle relative to the rod, and R the reaction of the rod on the par-
ticle. If we choose x to be positive along the rod and the y and z axes perpendicular to it,
with the z axis as that of rotation, the component equations of motion are
mx  m 2 x
my  2m x  R y
mz  Rz
 
Note that R has no x component since the rod is smooth and that  is along the z axis.
Now both y and z are zero for all time, and therefore both 
y and 
z must be put
equal to zero. This gives
Ry  2m x and Rz  0
Thus the reaction of the rod is in the positive y direction if  and x are both positive. The
quantity x may be determined from the solution of the x component equation of motion.

11.6 Find the velocity and acceleration of a point P on the rim of a wheel of radius R
which is rolling down an inclined plane. The inclined plane makes an angle  with
the horizontal.
Solution:

Fig.11.6
Let 0’x’y’z’ be the coordinate system that is translating with the wheel, with the
origin 0’ at the center of the wheel. The stationary reference frame on the ground is
represented by 0xyz, as shown in Fig. 11-6. Furthermore, we orient the 0’x’y’z’ system
so that i'  i, j '  j ,k'  k ' at t = 0. Consider the motion of point P fixed to the rim of the
wheel and located on the Y’ axis. The position vector of P in the stationary system is
 
rP  r  bj' (1)

in which r is the position vector of 0’ with respect to 0. The velocity and acceleration of
P are then given by
   
rp  r  bj' rp  r  bj' (2)
where j '    and  j '   are the angular velocity and angular acceleration, respectively.

164
The wheel is subject to three forces: its weight Mg, a reaction N normal to the
plane, and a frictional force F tangent to the plane. Resolving these forces along and per-
pendicular to the plane, we obtain
F  Mg sin   mx
N  Mg cos  0 (3)
bF  I 
the third relation describes the rotation about the center of mass (the center of the wheel,
in the present case). For rolling without slipping, we also have
b   x
which connects the angle  through which the wheel rotates while it rolls a distance x.
The negative sign is due to the measurement of the angle  , which is customary measured
anti-clockwise. It follows that
b  b   
x
and that
1
F   Mx
2
Substituting this value for the friction in the first equation of Eq.(3) we get
2
x   g sin  (4)
3
The velocity and acceleration of point P can be determined from Eqs.(2) and (4).

11.7 An xyz coordinate system is rotating with respect to an XYZ coordinate system
having the same origin and assumed to be fixed in space (i.e., it is an inertial sys-
tem). The angular velocity of the xyz system relative to the XYZ system is given
by

  2t i  t 2 j  (2t  4) k
where t is the time. The position vector of a particle at time t as observed in the
xyz system is given by

r  (t 2  1)i  6tj  4t 3 k
Find (a) the apparent velocity and true velocity of the particle, and (b) the apparent
and the true acceleration at time t = 1.
Solution:
(a) The apparent velocity at any time t is

dr / dt  2t i  6 j  12t 2 k
 2i  6 j  12 k at t  1.
The true velocity at any time t is

165
i j k
  
dr / dt    r  (2ti  j  12t 2 k)  2t t 2 2t  4
t 2  1 6t 4t 3
 34i  2 j  2 k at t  1.
(b) The apparent acceleration is
 
d 2 r d  dr 
    2i  24tk
dt 2 dt  dt 
 2i  24 k at t  1.
The true acceleration is
 
d 2r  dr d    
 2    r    (  r )
dt 2 dt dt
At t = 1, this equals
~
40i  184 j  36k .

11.8 Find (a) the Coriolis acceleration, and (b) the centripetal acceleration of the par-
ticle in Problem 11.7 at time t = 1.
Solution:
 
(a) Coriolis acceleration = 2  dr / dt  48i  24 j  20k.
  
(b) Centripetal acceleration =   (  r )  14i  212 j  40k.

11.9 A particle slides over a smooth plane which is tangent to the earth’s surface at a
Northern latitude  . Find the reaction, due to Earth’s rotation, of the plane on the
particle.
Solution:
We shall neglect the acceleration of the earth’s center in space and also shall
restrict the motionto the neighborhood of the point of contact of the plane with the
earth’s surface If  is theconstant angular velocity of the earth, v the velocity of the
particle at any time, and R the reaction of the plane on the earth, then the equation of
motion is
    
mr  mg  2m  v  R (1)
Eq.(1) is written with respect to a system of axes rigidly attached to the earth. We take
the origin to be at the point of the plane which is tangent to the surface of the earth,
 and x
po-sitive to the south, y to the east, and z vertically upward. The component of  in this
sys-tem are
x   cos  y  0 z   sin 
Accordingly, the component equations of motion are
mx  2m y sin  (2)
my = -2m x sin  (3)
 = -mg + 2m y cos   R
mz (4)

166
Now the left side of Eq.(4) is zero, since z is zero for all values of the time. We therefore
have left
R  mg  2m y cos 
in which the quantity y may be obtained from solutions of Eqs.(2) and (3).

11.10 A bug crawls outward with constant speed


 v’ along the spoke of a wheel which is
rotating with constant angular velocity  about a vertical axis. Find all the forces
acting on the bug. How far can the bug crawl before it starts to slip, given the co-
efficient of friction  between the bug and the spoke?
Solution:

Fig.11.10
(a) We first choose a coordinate system fixed on the wheel, with the origin at the
center of the wheel and x’ axis point along the spoke. Then we have
 
r '  i' x'  i' v' r'  0
If we choose the z’ axis to be vertical, then

  k'
The various forces are then given by the following:
Coriolis force:
 
2m  r '  2m v'( k'i')  2m v' j '.
Centrifugal force
  
m  (  r ')  m 2 [ k'( k'i' x')]  m 2 x' i'
Thus, the equation of motion in the moving coordinate system reads

F  2m v' j 'm 2 x' i'  0

in which F is the real force exerted on the bug by the spoke.
(b) Since the force of friction has a maximum value of  mg , slipping will start
when

F   mg
or
[(2m v') 2  (m 2 x') 2 ]1/ 2   mg
Solving for x’, we find
[  2 g 2  4 2 (v') 2 ]1/ 2
x' 
 2

for the distance the bug can crawl before slipping.

167
11.11 If a projectile is fired due east from a point on the earth’s surface at a nothern lati-
tude  , with a velocity of magnitude v 0 and at an angle of inclination to the hor-
izontal of  , show that the lateral deflection when the projectile strikes the earth
is
4v 0 3
d  2  sin  sin 2  cos
g
where  is the rotation speed of the earth.
If the range of the projectile is R for the case  = 0, show that the change of range
due to the rotation of the earth is
1
R  2 R 3 / g cos  (cot 1/ 2   tan 3/ 2  ) .
3
Solution:
(a) We choose a right-handed coordinate system with a z axis vertically upward, x
positive to the south and y to the east. Then the lateral deflection of the projectile in this
system is in the x direction and the acceleration is
x  2z v y  2( sin  )(v0 cos )
a x   (1)
Integrating Eq.(1) twice and using the initial conditions x(0)  0 and x(0)  0, we get
x(t )   v0 t 2 cos sin  (2)
Now we treat the z motion of the projectile as if it were undisturbed by the Cori-
olis force. In this approximation, we have
1 2
z(t )  v0 t sin   gt
2
from which the time T of impact is obtained by setting z = 0
2v0 sin 
T (3)
g
Substituting this value for T into Eq.(2), we find the lateral deflection at impact to be
4 v0 2
d  x (T )  sin  cos sin 2  .
g
(b) In part (a) we assumed the z motion to be unaffected by the Coriolis force.
Actually, there is an upward acceleration given by 2 x v y so that
z  2 v0 cos cos   g

from which the time of flight is obtained by integrating twice, using the initial conditions,
and then setting z = 0:
2v0 sin 
T'  (4)
g  2 v0 cos cos 
Now, the acceleration in the y direction is

168
y  2x vz  ( cos  )(v0 sin  gt )
a y  
Integrating twice and using the initial conditions, y(0)  v0 cos and y(0)  0, we have
1
y(t )   gt 3 cos    v0 t 2 cos  sin   v0 t cos (5)
3
Substituting Eq.(4) into Eq.(5), the range R’ is
8  v 0 3 g sin 3  cos  4 v 0 3 sin 2  cos 
R'  
30 ( g  2 v 0 cos cos ) 3 g  2 v 0 cos cos 
2v 0 2 cos sin 

g  2 v 0 cos cos 
We now expand each of these three terms, retaining quantities up to order  but ne-
glecting all quantities proportional to  2 or higher. In the first two terms, this amounts to
neglecting 2 v0 cos cos  compared to g in the denominator. But for the third term we
have
2v 0 2 cos cos 2v 0 2 2 v 0
 cos sin  (1  cos cos )
2 v 0 g g
g (1  cos cos  )
g
4 v 0 3
 R sin  cos2  cos 
g2
where R is the range when the Coriolis effects are neglected:
2v 0 2
R cos sin  (6)
g
The range difference, R  R' R, now becomes
4 v0 2 1
R  2 cos  (sin  cos2   sin 2  )
g 3
Substituting for v 0 in terms of R from Eq.(6), we finally have the desired result

2R3 1
R   cos  (cot 1/ 2   tan 3/ 2  ) .
g 3

11.12 A spherical planet of radius R rotates with a constant angular velocity  . The ef-

fec tive gravitational acceleration g eff is some constant, g, at the poles and 0.8g at

the equator. Find g eff as a function of the polar angle  and g.
Solution:

g eff is given by Eq.(11-45a):
    
geff  g    (  R)
from which we have

169
        
geff 2  geff  geff  g 2  2 g    (  R)  [  (  R)]2 (1)
Using the vector identity
           
( A  B)  (C  D)  ( A  C)( B  D)  ( A  D)( B  C)
we can rewrite the third term on the right-hand side of Eq.(1) as
       
[  (  R)]2  [   ]2 (    R)
   
 (   )  (   )
     
  2 (   ) 2  (   )(   )
        
  2 [  (  R)]2  [  (  R)][(  R)   ]
  
  2 [  (  R)]2
Similarly, using the vector identity
        
A  ( B  C)  C  ( A  B)  B  (C  A)
we can rewrite the second term on the right-hand side of Eq.(1) as
         
g    (  R)  g     (     R)
   
   (g  ) 
(g = gR)
 
   ( gR   )
    
   g ( R   ) / R  g (  R) 2 / R
Substituting these into Eq.(1) we get
   
geff 2  g 2  2 g(  R) 2 / R   2 (  R) 2
 
Now, with    k and g  gR , we have
 
  R   R sin   R sin( / 2   )   R cos 
and
geff 2  g 2  2 g 2 R sin 2    4 R 2 sin 2  (2)
From which we find
g eff 2 (  0)  g 2
g eff 2 (   / 2)  g 2  2 g 2 R   4 R 2
 ( g   2 R) 2
and so
geff (   / 2) 2R
 1  0.8
g eff (  0) g
or
 2 R  0.2 g
From Eq.(2) we have
geff  g(1  0.36 sin 2  ) 1/ 2

170
Now
vesc 2  2 gR  v 2   2 R 2
Solving for v
v  (2 gR   2 R 2 ) 1/ 2  (18
. gR) 1/ 2
or
v
 (18
. / 2) 1/ 2  0.949 v  0.949vesc . .
vesc

11.13 In Fig.11.13, the 0-system, with its z-axis out from the page, is a fixed coordinate
system. The 0’-system
 is a rotating system, with x’ and y’ rotating with constant
angular velocity  about the z=z’ axis with respect to the fixed system. An ob-
server sitting at 0’ sees a rock on the ground fixed in the fixed system at a distance
r along the x-axis. To him the rock moves in a circular orbit with speed v'  r .
What force holds the rock in this circular orbit for the rotating observer?
Solution:

Fig. 11.13
From Eq.(11-26) we have
         
a  A  a '2  v '  (  r ')    r '

Since 0’ and 0 coincide, so A  0 .Furthermore, the rock is fixed in the 0-system, so
   
a  0, and   0 because  is constant. Solving the above equation for a' , we obtain
     
a '    (  r ')  2 'v '
From Fig. 11.13, it is easy to determine that the direction, including the minus sign, of the
first term on the right side is in the positive x direction. The magnitude of this centrifugal
acceleration is  2 r as expected. The direction, including the minus sign, of the second
term, is in the negative x direction or towards the left. Its magnitude is 2 2 r . The sum of
these two terms gives a net acceleration  2 r towards the left which is exactly what is
need-ed for this orbital motion. Hence, it is the Coriolis force which holds the rock in the
“orbit” observed by the observer in the rotating coordinate system.

11.14 A particle of mass m is constrained to move along a smooth rod A0B (Fig. 11.14).
The rod A0Brotates in a vertical plane about a horizontal axis with constant angu-
lar velocity  . The horizontal axis passes through the midpoint 0 of the rod and is
perpendicular to the vertical plane. Determine the motion of the particle P along the
rod.
Solution:

171
Fig. 11.14

The rod A0B rotates in yz plane about the x axis. At time t let r be the position vector of
the particle and  the angle made by the rod with the y axis. There are three forces acting
on P:
(i) The weight,

mg  mgk  mg sin  r  mg cos 
(ii) The centrifugal force,
  
m  (  r )  m i  ( i  rr)
 m[ i( i  rr)  rr( i   i)]
  m rr
2

(iii) The reaction force N  N of the rod which is perpendicular to the rod since
the rod is smooth.
The equation of motion of the particle P is then
d 2r
m 2 r  mg sin  r  mg cos   m 2 rr  N
dt
 (m 2 r  mg sin  )r  ( N  mg cos )
It follows that
N  mg cos
d r / dt 2   2 r  g sin 
2

Since    , a constant, we have  = t if we assume   0 at t  0 . Then the above


equation becomes
d 2r r / dt 2   2 r   g sin( t )
If we assume that at t  0, r  r0 , dr / dt  v0 , we find
 r0 v0 g   t  r0 v0 g   t g
r    
2 e    
2 e  sin( t )
 2 2 4   2 2 4  2 2
It is easy to check that if r0  0 and v0  g / 2 , r reduces to
g
r sin( t )
2 2
which shows that the particle oscillates along the rod with simple harmonic motion, and
the amplitude and period are given by g / 2 2 and 2 /  respectively.

172
Chapter 12

12.1 Find the center of mass of each of the following:


(a) A solid homogeneous hemisphere of radius a.
(b) A semicircular, homogeneous flat plate of radius a.
Solution:
(a)

Fig.12.1 Fig.12.1a
By symmetry the center of mass lies on the z axis (Fig.12.1). Subdivide the hemi-
sphere into solid circular plates of radius r, such as ABCDEA. If the center G of such a
ring is at distance z from the center 0 of the hemisphere, r2 + z2 = a2. Then if dz is the
thickness of the plate, the volume of each ring is
r2dz = (a2 - z2)dz
if  is the mass density (i.e., mass per unit volume), then we have

   z (a
a
2
 z 2 )dz 3
  a
0
zcm
   (a
a

0
2
 z 2 )dz 8

(b). We take the axes as shown in Fig.12.1a. From symmetry the center of mass
must lie on the y axis, i.e., as we have chosen our axes
x cm  0

 ydm  
a a2  x2

x  a y  0
ydydx 4a
y cm   
 dm   a a2  x2
dydx
3
a 0

12.2 Find the center of mass of the quadrant of an elliptical plate of constant thickness
enclosed by the two semiaxes. The density varies in such a way that at any point it
is directly proportional to the distance from the point to the major axis.
Solution:

Fig.12.2

173
We take the axes as shown in Fig.12.2. The density  can be written as  = ky.
Hence
  xdA  yxdA  y 2 dA
xcm   , ycm 
  dA  ydA  ydA
Choose
a 2
dA  xdy  b  y 2 dy
b
(the equation of the ellipse being x2/a2 + y2/b2 = 1). Then

 y b  y 2 dy 3 b
a b 2 2
b 0
y cm  

a b 16
y b 2
 y 2
dy
b 0
To obtain xcm, we note that the center of mass of each of the horizontal strips is x/2 which
equals (a / 2b) b 2  y 2 . Thus the whole mass of each strip ( ky  dy  (a / b)  b 2  y 2 )
may be considered to be located at the distance (a / 2b)  b 2  y 2 from the y axis. There-
fore

1 ba 2 a 2
b  y2 b  y 2 ydy 3
2 b
0 b
x cm   a.
a1 3 8
b
b3

12.3 Find (1) the moment of inertia and (2) the radius of gyration of each of the follow-
ing
(a) A thin circular hoop or cylindrical shell about its axis.
(b) A right circular cone of height h and radius a about its axis.
Solution:
(a)

Fig.12.3
For a thin circular hoop or cylindrical shell, all particles lie at the same distance from the
central or symmetrical axis. Thus
Iaxis = ma2
where a is the radius and m is the mass.
The radius of gyration K is defined by the equation I = mK2. Thus, in the case of
a thin circular hoop or cylindrical shell, the radius of gyration is simply equal to the
radius a.
(b) Subdivide the cone, one quarter of which is shown in Fig. 1, into elements of

174
mass dm as indicated in the figure.In cylindrical coordinates (r, , ) the element of mass
dm of the cylinder is dm = rdrddz, where  is the mass density. The moment of inertia
of dm about the z axis is
r2dm = r3drddz
Now (h - z)/h = r/a, or z = h[(a - r)/a]. Thus the total mo-
ment of inertia about the z axis is

I   0 r 0 z0
2 a h ( a  r )/ a 1 3
 r 3drd dz 
 a 4  h  Ma 2
10 10

where the total mass M of the cone is M   0 r 0 z0


2 a h ( a  r )/ a 1
 rdrd dz   a 2  h .
3
Since K = I/M, we have the radius of gyration K  a 3 / 10.
2

12.4 Use the parallel axis theorem to find the moment of inertia of a solid circular cy-
linder about a line on the surface of the cylinder and parallel to the axis of the cy-
linder.
Solution:

Fig.12.4a Fig.12.4b
Fig.12.4a shows the cross section of the cylinder, the axis of the cylinder passes
through point C. The line on the surface of the cylinder and parallel to the axis of the
cylinder is represented by point A. Then by the parallel axis theorem we have
IA = IC + Ma2
where IC is the moment of inertia about the central axis, and M is the mass of the cylin-
der. To calculate IC, we subdivide the cylinder, a cross section of which is shown in Fig.
12.4b, into concentric rings, one of which is the element shown shaded. The volume of
this element is
(Area)(thickness) = (2rdr)(h) = 2rhdr,
and the element of mass is dm = 2rhdr. The moment of inertia of dm is r2dm =
2r3hdr, where  is the mass density. Thus the total moment of inertia is

I  r 0 2  r 3hdr 
a 1 1
  ha 4  Ma 2
2 2
where the mass M is
M  0 2  rhdr    a 2 h.
a

Thus
1 3
IA  Ma 2  Ma 2  Ma 2 .
2 2

175
11.5 Consider a rigid body that is in the form of a plane lamina of any shape. If the lam-
ina is in the xy-plane and Ix, Iy, and Iz denote the moments of inertia about the x, y,
and z axes, respectively, prove that
Iz = Ix + Iy
This is known as the perpendicular axis theorem for a plane lamina.
Solution:

Fig.12.5
Let the position vector of the particle with mass ma in the xy plane be

ra  xa i  ya j
The moment of inertia of ma about the z axis is mara2. Then the total moment of inertia of
all particles about the z axis is
N N
I z   ma ra   ma ( xa  ya )
2 2 2

a 1 a 1
N N
  ma xa   ma ya  I x  I y
2 2

a 1 a 1

where Ix and Iy are the total moments of inertia about the x axis and y axis respectively.

11.6 (a) Find the inertia tensor for a uniform square plate of side a and mass m about a
diagonal in a coordinate system 0xyz, where 0 is at one corner and the x and y
axes are along two edges.
(b) Find the angular momentum about the origin for the plate just mentioned in (a)
when it is rotating with an angular speed  about (1) the x-axis, and (2) about
the diagonal through the origin.
Solution:

Fig.12.6a
(a) Let us choose coordinate axes as shown in Fig.12.6a with the square plate
lying in the xy plane with a corner at the origin. The moment of inertia of an element
dxdy of the plate about the x axis is y2dxdy, where  is the mass density. Then the
moment of iner-tia of the entire plate aboutthe x axis is
I xx  x 0 y 0 y 2 dxdy  a 4  ma 2
a a 1 1
3 3

176
where m = a2
Similarly, the moment of inertia of the plate about the y axis is

I yy  x 0 y 0  x 2 dxdy  a 4  ma 2
a a 1 1
3 3
The moment of inertia of the element dxdy about the z axis is (x2 + y2), and so the mo-
ment of inertia of the entire plate about the z axis is

I zz  x 0 y 0  ( x 2  y 2 )dxdy  ma 2  ma 2  ma 2
a a 1 1 2
3 3 3
This also follows from the perpendicular axes theorem.
The product of inertia of the element dxdy of the plate about the x and y axes is
xydxdy, so the product of inertia of the entire plate about these axes is

I xy  I yx   x 0 y 0  xydxdy   a 4   ma 2
a a 1 1
4 4
The product of inertia of the element dxdy of the plate about the x and z axes is the pro-
duct of dxdy by the distances to the yz and xy planes, which are x and 0 respectively.
Thus we have
Ixz = Izx = 0
and similarly Iyz = Izy = 0. Hence the inertia tensor is
ma 2 / 3 ma 2 / 4 0 1 3 / 4 0
ma 2
I = ma / 4
2 2
ma / 3 0  3 / 4 1 0
2
3
0 0 2ma / 3 0 0 2
(b)

Fig.12.6b
For rotation about the x-axis we have
x   ,  y  z  0
and the angular momentum is
 1 3 / 4 0    
  ma    ma  3 
2 2

L  I  3 / 4 1 0 0    
3    3  4 
 0 0 2  0   0 
ma 2 ma 2
  j
 i 
3 4
Notice that the angular momentum vector does not point in the same direction as the an-
gular velocity vector, but points downward. The magnitude of the angular momentum is
given by

177
5
L  ( Lx 2  Ly 2  Lz 2 ) 1/ 2  ma 2 .
12
For rotation about the diagonal we have

Fig.12.6c
x   y   cos 450   / 2 , z  0
or
 
  (i  j )
2
therefore the angular momentum is
 1 3 / 2 0 1  1
  ma 2
    ma 2
  
L  I   3 / 2 1 0 1  1
3 2    12 2  
 0 0 2  0  0
ma 2 
= (i  j )
12 2

and it is in the same direction as the angular velocity vector  . The magnitude of the
angular momentum in this case is equal to
L  ma2 / 12 .

12.7 (a) Find the principal moments of inertia of a square plate about a corner.
(b) Find the directions of the principal axes for part (a).
Solution:
An important property of a principal axis is that if a rigid body rotates about it the
direction of the angular momentum is the same as that of the angular velocity. Thus
 
L  I (1)
where I are the principal moments of inertia to be determined later. Now
 
L  Lx i  Ly j  Lz k,    x i   y j  z k
and
Lx  I xxx  I xy y  I xzz , Ly  I yxx  I yy y  I y zz Lz  I zxx  I zy y  I z zz
Substituting these into Eq.(1) we obtain
( I xx  I ) x  I xy y  I xz z  0
I yz x  ( I yy  I ) y  I yz z  0 (2)
I xzx x  I y z y  ( I z z _ I ) z  0
The principal moments of inertia are found by setting the determinant of the coefficients
of x, y, z in Eq.(2) equal to zero, i.e.,

178
Ixx  I Ixy Ix z

Iyx Iyy  I Iyz 0 (3)


Ixz Iyz Iz z  I
Substituting the moments and products of inertia from Problem 12.6 (a) we obtain
ma 2 / 3  I ma 2 / 4 0
ma / 42
ma / 3  I
2
0 0
0 0 2ma 2 / 3  I
Expansion gives
 I 2  (2ma 2 / 3) I  7m2 a 4 / 144 2ma 2 / 3  I   0
from which we find the three principal moments of inertia
1 7 2
I1  ma 2 , I2 ma 2 , I3  ma 2 .
12 12 3
To find the direction of the principal axis corresponding to I1, we let I = I1 in Eq.(2), and
substitute the moments and products of inertia from Problem 12.6 (a):
1 2  1
 ma  I  x  ma 2 y  0 z  0
3  4
1 1 
 ma 2 x   ma 2  I  y  0 z  0
4 3 
2 
0 x  0 y   ma 2  I  z  0
3 
The first two equations yield y = z while the third gives z = 0. Thus the direction of
the principal axis is the same as the direction of the angular velocity vector

   x i   y j  z k   x i   x j   x (i  j )
Then the principal axis corresponding to I1 is in the direction i  j .
Similarly, by letting I = I3 = ma2/12 in Eq.(2) we find y = -x, z = 0 so that the

direction of the corresponding principal axis is    x i   x j   x (i  j ) , or i  j . By
letting I = I3 = 2ma2/3 in Eq.(2) we find x = y = 0, while z is arbitrary. This gives

  z k which shows that the third principal axis is in the direction k as shown in Fig.
12.7.

Fig.12.7
12.8 Find
(a) the principal moments of inertia at the center of a uniform rectangular plate of

179
sides a and b, and
(b) the torque needed to rotate such a plate about a diagonal with constant angular
velocity .
Solution:

Fig.12.8a
(a) The principal axes lie along the directions of symmetry and thus must be along
the x axis, y axis, and z axis as shown in Fig.12.8a. The z axis is perpendicular to the xy
plane.
Let us first calculate the moment of inertia of the plate about a side, say about the
Y axis. The mass of the shaded element is adX, its moment of inertia about the Y axis is
aX2dX. Thus the total moment of inertia is

I Y  X 0  aX 2 dX   ab 3  mb 2
b 1 1
3 3
where  is the mass density and m = ab. By the parallel axis theorem, we have IY = Iy +
m(b/2)2 = mb2/3, or Iy = mb2/12. Similarly, we find Ix = ma2/12. Then by the perpendicu-
lar axes theorem, we have Iz = Ix + Iy = m(a2 + b2)/12. Accordingly, the principal
moments of inertia of the plate at the center 0 are given by
1 1 1
I1  ma 2 , I2  mb 2 , and I3  m(a 2  b 2 ) .
12 12 12
(b) The torque needed to rotate the plate can be found from the
 Euler’s equations,
Eq.(12.82). Let us first find the components of the angular velocity  . By an inspection
of Fig.12.8b, we find

Fig.12.8b
   a b
  (  i)i  (  j ) j   i  j
a b
2 2
a b
2 2

Thus
 a  b
1  , 2  , 3  0 Fig. 1
a 2  b2 a 2  b2
Substituting these and I1, I2, I3 (from part a) into Eq. (12.82), the Euler’s equations:

180
I 11  ( I 3  I 2 )23  N 1
I 2 2  ( I 1  I 3 )31  N 2
I 3 3  ( I 2  I 1 )12  N 3
we find
m(b 2  a 2 )ab 2
N 1  N 2  0, N3 
12(a 2  b 2 )
Thus the required torque about 0 is
 m(b 2  a 2 )ab 2
N k
12(a 2  b 2 ) 
Note that if the rectangular plate is a square (i.e., a = b), the N  0 .

12.9 A solid cylinder of radius a and mass M rolls down an inclined plane of angle .
The coefficient of friction between the cylinder and inclined plane is . Discuss the
motion for various values of .
Solution:
See Example 12.2.

12.10 A cube of edge s and mass M is suspended vertically from one of its edges as
shown in Fig.12.9,
(a) Show that the period of small oscillations is P  2 4 2 s / 3g .
(b) What is the length of the equivalent pendulum?
Solution:

Fig.12.10
(a) Our system constitutes a physical (or compound) pendulum. Its equation of
motion is given by Eq.(12.38), with b (the distance from the suspension point to the
center of mass) replaced by s / 2 . This is easy to see. Since the diagonal of a square of
side s has length s 2  s 2  s 2 . The distance 0C from the axis to the center of mass is
1
s 2  s / 2 . Now the moment of inertia of a cube is the same as that of a square plate
2
1 2
about a side. Thus I  M ( s 2  s 2 )  Ms 2 . Then the period for small oscillations is
3 3
2 Ms 2 / 3
P  2  2 4
2 s / 3g .
Mgs / 2
2
(b) The length of the equivalent simple pendulum is   2s .
3

181
12.11 A dumbbell (two equal masses connected by a massless rigid rod of length 2a) ro-
tates at a fixed inclination  with a constant angular velocity  about a pivot point
at the center of the rod. Find the angular momentum and torque of the system.
Solution:

Fig.12.11
In Fig.12.11 the dumbbell is in the plane of the paper and is of length 2a. We take
the origin fixed at 0, and the axes are fixed in the inertial system. From Eq.(12.28) we
have         
L  L1  L2  mr1  (  r1 )  mr2  (  r2 )
  
and we note that L1 & L2 both point in the same direction, that of L (see Fig.12.11). It is
easy to see that the magnitude of the angular momentum is
L  ma 2 sin  ma 2 sin  2ma 2 sin (1)
 
However, the direction of L is continually changing (it constrained to rotate about  .), a
  
torque N is necessary to maintain the motion. This is determined by Eq.(12.3), L  N .
 
Now L is a vector denoting the direction in which the end of the  vector is moving, and,
L
 
in analogy to Eq.(12.19), may be written   L . Thus the torque N applied to the rotating
masses is directed perpendicularly out of the paper for the instantaneous config-uration
(see Fig.12.11). pictured by Fig. 1 and is of magnitude
N  2 2 a 2 m sin cos (2)
By Newton’s third law, the torque which the rotating dumbbell exerts on the bearings
(say at A and A’) is equal and opposite to this. Thus it points into the paper at the instant
under consideration and is of the same magnitude as that given by Eq.(2).

12.12 Let T be a symmetric tensor, and 1, 2, and 3 are its three real eigenvalues with
the three corresponding eigenvectors A1, A2, and A3. Show that the three eigen-
vectors A1, A2, and A3 are orthogonal if the three eigenvalues 1, 2, and 3 are
distinct.
Solution:
We start with the eigenvalue equation
TAi = I Ai
~
and multiply this equation from the left by A j and obtain
~ ~ ~
A j TAi = A j iAi = i ( A j Ai)
The left-hand side is also equal to

182
~
(TAj)Ai = (jAj)Ai = j( A j Ai)
from which follows that
~
(i - j) ( A j Ai) = 0.
~
Because i  j, we conclude that A j Ai = 0, which means orthogonality.

12.13 An automobile wheel has its axle (axis of rotation) slightly bent relative to the sym-
metry axis of the wheel. To remedy the situation counterbalance weights can be
suitably located on the rim so as to make the axle a principal axis for the total sys-
tem (weights and wheel). For simplicity, we consider the wheel as a thin uniform
circular disk of radius a & mass M as shown in Fig. 1. The axle inclines by a small
angle  relative to the x-axis (symmetry axis of the disk). Two balancing weights
(m’) lie in the xy-plane. Show that if
 aM
bm'
8
the wheel will be dynamically balanced.
Solution:

Fig.12.13
It is easy to see, from the symmetry relative to the xy plane, that the z axis is a
principal axis for the wheel plus weights: z is zero for the weights, and the xy plane
divides the wheel into two equal parts having opposite signs for the products zx and zy.
Accordingly we can use Eq.(12.76) to find the relationship between  and the other
parameters.
Now, for the wheel alone the moments of inertia about the x and the y axes are
1 2 1
ma and ma 2 , respectively.
2 4
Thus, for the wheel plus weights
1 2 1 2
I xx  ma  2m' a 2 , I yy  ma  2m' b 2
2 4
The xy product of inertia for the wheel alone is zero, so we need only consider the
weights for finding Ixy for the system, namely
I xy    xi yi m'i   (b)am'b(a)m'  2abm'
Eq.(12.76) then gives the inclination of the axle:

183
2 I xy 4abm'
tan 2  
I xx  I yy 1
ma 2  2m'(a 2  b 2 )
4
If  is very small and also m’ is small compared to m, then we can express the above
rela-tion in approximate form by neglecting the second term in the denominator and
using the fact that tan   for small . The result is
bm'
 8 .
am

12.14 A coin is initially set spinning rapidly on its edge on a table. As the spin decreases
due to frictional forces, a point is reached where the coin begins to precess (wob-
ble).With further loss of energy the coin will gradually “lie down” with an increase
of the precession rate, while at the same time the actual turning rate, as seen above,
decreases. Explain the curious behavior of a spinning coin just described with your
calculations in detail.

Fig.12.14
Solution:
We can understand the curious behavior of a spinning coin with a calculation sim-
ilar to the treatment of top motion in the text. We shall neglect frictional torques in com-
parison with the gravitational torque which is equivalent to the assumption that the lying
down rate is small compared to the precession rate. Since there is no unaccelerated point
in the body we must compute the torque and angular momentum about the center of mass.
Now it is seen, from Fig.12.14, that the torque is perpendicular to the plane formed by the
vertical and the symmetry axis. The components of angular momentum along the symme-
try and the vertical axes are conserved and give
3     cos  cons tan t (1)
L z  I sin 2   I 33 cos (2)
where I3 and I are the moments of inertia about the symmetry axis and perpendicular to
the symmetry axis, respectively. The energy of the coin, due to motion and position of the
c.m. and to the rotation about the center of mass, is given by
1 1 1
E  I ( 2   2 sin 2  )  I 33 2  Mgz  Mz 2 (3)
2 2 2
where z = a sin.
Subtraction of the time derivative of Eq.(2) from the time derivative of Eq.(3)
yields the following equations of motion

184
( I  Ma 2 cos2  )  Ma 2 2 cos sin   Mga cos
(4)
 I 2 sin  cos  I 33 sin   0
The time derivative of Eq.(2) gives
  sin cos  I 33 sin  0
I sin 2   2 I (5)
We first examine the stability of the initially spinning coin. Since the coin is spun at  =
/2 and the center of mass remains fixed (hence   0 or the coin would skid), we shall
adopt the initial condition
3   cos    0 . (6)
By Eq.(4) with    / 2   we have, keeping only linear terms in  and its derivatives
    2   Mga / I   0 (7)
Thus if  (the total spin at 900) is large enough,
   Mga / I 
1/ 2
 0 (8)
the motion will be stable. As  decreases below 0 the coin will wobble and begin to roll
and precess.
The coin’s motion is now assumed to exhibit regular precession (  constant)
with no skidding. To see what the rolling constraint implies we must calculate the rim
velocity of the disk. Since there is no c.m. motion, the rim velocity is just 3a, where 3
is the angular velocity normal to the coin, and the rolling constraint is
3   cos    0 (9)
We have seen that 3 was initially zero and is conserved so we expect a smooth
transition from spinning to precession. For pure precession we find from Eq.(4), by set-
ting  = constant,
  0 / (sin ) 1/ 2 (10)
The vertical component of the angular velocity is
s     cos (11)
Using the rolling constraint of Eq.(9) we find
s   sin 2   0 (sin ) 3/ 2 (12)
Thus, as the coin lies down (  0) the precession becomes very rapid while the spin
slows to stop.

12.15 A gyroscopic compass is used for navigation on ships and airplanes. In its simplest
form such a compass consists of a rapidly rotating circular disk whose axis is free
to turn in a horizontal plane. Show that owing to the earth’s motion the system is in
equilibrium when the axis of the rotating disk lies along a geographical north-south
line or along a meridian.
Solution:

185
Fig.12.15a Fig.12.15b
Fig.12.15a shows the earth rotating about its axis with angular velocity . P is a
point on its surface at which angle of latitude is . PV points vertically upward and PN
denotes the meridian direction towards the North Pole. A gyroscopic compass whose axis
is free to rotate horizontally is placed at P. Fig.12.15b shows the axis of the compass at an
angle  to PN,  being the angular velocity of the wheel about its axis PA and PE the
eastern direction through P.
The total angular velocity vector of the gyrocompass has components (sin - )
about PV, cos about PN and  about PA. An equivalent representation is
sin   , cos  sin ,   cos  cos 
about the mutually perpendicular (principal) axes PV, an axis through P mutually perpen-
diccular to PV and PA, and PA respectively. Suppose I and I’ are the principal moments
of inertia about these axes, then the kinetic energy T is
1
T
2
 I ( sin   ) 2  I2 cos2  sin 2   I '(   cos  cos ) 2 
This is also the expression for the Lagrangian L if we take the zero level of the potential
energy through P. Suppose the wheel rotates through an angle  in time t so that    .
Lagrange’s equation of motion for coordinate  is
d
  cos  cos   0
dt
or
  cos  cos  n (1)
Lagrange’s equation of motion for coordinates  is
I  I2 cos2  sin cos  I ' cos  sin (  cos  cos )  0 (2)
Substituting Eq.(1) into Eq.(2), assuming  to be small and neglecting 2, we obtain
I  ( I ' n cos  )  0 (3)
Eq.(3) shows that  = 0 is a position of equilibrium of the gyrocompass and that the pe-
riod of small oscillation about it is 2 I / ( I ' n cos  ) .

12.16 Consider a uniform bar of cylindrical cross section, mounted shown in Fig.12.6, ro-
tat ing about a horizontal shaft with constant angular velocity  . The supports for
the shaft are mounted
 on a turntable that revolves about a vertical axis at constant
angular velocity  . Discuss the motion of this rotating bar by using Euler’s equa-

186
tions.
Solution:

Fig.12.16a

Fig.12.16b Fig.12.16c
We shall consider the mass of the shaft to be negligible compared to the mass of
the bar. Let the angular speed of the bar about the horizontal axis be constant and denoted
by , and let  be the constant angular speed of rotation of the turntable (see Fig.12.16a).
We choose the principal axes of the bar as shown by (1), (2), and (3) in
Fig.12.16b The triad of unit vectors i, j, k are chosen to be along the principal axes and
they rotate with the bar as it turns. The moments of inertia about the principal axes are
denoted by I1, I2, and I3. By virtue of the symmetry and shape of the bar, I2 = I3 and these
are larger than I1. At any instant the bar makes an angle  with a horizontal y axis in its
plane of rotation, as shown in Fig.12.16b.
As the turntable rotates, the z-axis remains vertical while the y-axis moves into
the plane of the paper. The x-axis points towardthe reader, and we may represent the

angular velocities at any instant by the vectors  and  , as shown in Fig.12.16c. By

direct projection of  and  on the principal axes of the bar, we find the angular velo-
city components
1  sin , 2  cos , 3   (1)
where  and  are constant. From these equations we obtain the accelerations
1   cos    cos
 2   sin     sin  (2)
 3  0
since    . Hence, by Eqs.(12.82), Euler’s equations are
I 1 cos   cos ( I 3  I 2 )  N 1
 I 2  sin    sin  ( I 1  I 3 )  N 2
2 sin  cos ( I 2  I 1 )  N 3
These simplify to

187
I 1 cos  N 1
( I 1  2 I 2 ) sin   N 2 (3)
( I 2  I 1 )2 sin  cos  N 3
Hence it is apparent that the torque components relative to the principal axes are periodic
functions. If  is very small compared to , we note that N3 is the smallest of these
torque.
We are more interested, however, in the resulting moments about the x-, y- and z-
axes shown in Fig.12.16c. They are
N x  N 3 , N y  N1 cos  N 2 sin , N 3 = N1sin + N 2 cos (4)
Substituting Eqs.(3) into Eq.(4) and simplifying, we have
1
Nx  ( I 2  I 1 )2 sin 2 (5a)
2
N y  I 2  ( I1  I 2 ) cos 2 (5b)
N z  ( I1  I 2 ) sin 2 (5c)
Note that these are all periodic functions with period one half that of the rotation of the
bar. The physical interpretation of these equations is the following. The term I3 in Eq.
(5b) represents a constant torque tending to depress the positive x-axis, i. e., the k direc-
tion of the rotational axis. Since I1 < I2, the second term in Eq.(5b), represents an uplifting
torque on the positive x-axis when cos2 > 0 and a depressing moment when cos2 < 0,
this action results in a periodic thrust downward on the left bearing in Fig.12.16a. Simi-
larly, Eq. (5c) represents a torque tending to rotate the shaft in a periodic way horizontal-
ly, and consequently can be used to calculate the side thrust on the bearings. Lastly, Eq.
(5a) represents the torque about the shaft axis itself. We call the torques Nx, Ny, Nz the
rolling, pitch-ing, and yawing torques of the systems.

12. 18 A uniform thin circular disk is constrained to spin with an angular velocity 
about
an axis passing through the center of the disk but making an angle  with the axis
of symmetry of the disk. It is suddenly released. Find the half-angle of the cone in
space described by the symmetry axis and the time in which this cone is described.
Solution:

Fig.12.18
Let us choose axes rigidly attached to the disk, with the instantaneous axis of  in

188
the yz plane and making an angle  with 0z (fig. 1). Since there is no external torque ac-
ting, the direction in space of the angular momentum L will not vary. Let it make an
angle  with 0z, where  is to be determined. Note that the axes chosen are principal axes
and since  has no x component, neither, then, has L.(Since Ix = Iy, the axis 0z and the
vectors  and L are all coplanar.) We have

  jI   kI
L  jLy  kL  
z y y z z

Now Ix = Iy = ma2/4, and Iz = ma2/2. Accordingly L becomes


 ma 2 ma 2

L j 
 sin   k  cos
4 2
Therefore we have
Ly (ma 2 / 4) sin  1
tan     tan 
Lz (ma 2 / 2) cos 2
in which  is the desired half angle of the cone described by 0z about the L vector. The
period of precession is given by
P  2 / 
L (ma 2 / 4) sin 2   4 cos2 
where      1  3 cos2 
Iy ma 2 / 4
Substituting this into the expression for P we have
2
P .
 1  3 cos2 
With reference to the space and body cones mentioned in the text, the half angles of these
are, respectively,  -  and .

12.19 An electron of mass m is revolving around the nucleus in an atom along a circular
orbit of radius r. An external magnetic field B acts in a direction that makes an an-
gle  with the normal to the electron’s orbital plane. The magnetic field produces a
torque
 1    
N   e(r  v )   B
2 

where e is the electronic charge, v is the velocity of the electron, and the quantity
within brackets is the magnetic moment of the effective circular current loop. Show
that the normal to the electron’s orbital plane precesses with an angular velocity
eB/2m about the direction of the field. This is known as Larmor precession.
Solution:

Fig.12.19

189
We choose the coordinate systems as shown in Fig.12.19. The axis 0z0 is the di-
rection of B. The axes 0xyz are moving with the precessional velocity of the orbit, which
is in the xy plane, the 0x axis being constructed to remain in the x0y0 plane. Since the an-
gular momentum due to the orbital velocity is
   
L  mr  v  kmrv
the total angular momentum is

L  I x x i  I y y j  ( I zz  L ) k
in which  is the angular velocity of the moving axes. The first of Euler’s equations has
the form
1
I x x  ( I z  I y ) z y  I z y   ervB sin 
2
where   v / r is the angular velocity of the electron. We assume it is very much larger
than any of the components of . Accordingly we shall neglect terms of the type xy
and moreover shall neglect  x . Thus
1
I z y   ervB sin 
2
from which
1 
 ervB sin  
 
y  2 I z 
But Iz = mrv, and so
eB sin 
y 
2m
The angular velocity of precession  is given by
y eB
  
sin  2m

This gives the rate at which the normal to the plane of the orbit at 0 ( L vector) precesses
about the direction of the magnetic field B. The frequency is known as Larmor frequency.

12.20 For forces that act only for a very short time, Eq.(12.76) can be rewritten in integral
form
 
dL   N ( e) dt
The time integral on the right-hand side is called the angular (rotational) impulse. For
rotation about a fixed z axis, the preceding equation becomes
dLz  I z z z   N z ( e) dt
We now apply this result to the dynamics of billiard shots. Consider a stationary
billiard ball of radius a on a pool table. If the cue stick hits the ball in its vertical
median plane in the horizontal direction at a vertical distance h above the table,
show that the condition for the ball to roll without slipping is h = 7a/5.

190
Solution:

Fig.12.20
As shown in Fig.12.20, the cue imparts an impulse to the stationary ball at a ver-
tical distance h above the table. The linear impulse is

MVx  MVx 1  t Fx dt
t1
(1)
0

where Vx1 is the velocity of the CM just after impact. The angular impulse about the z
axis which passes through the CM of the ball is

Lz  I z zz 1  t (h  a ) Fx dt
t1
(2)
0

where a is the radius of the ball. By elimination of the force integral between Eqs.(1) and
(2) and substitution of the moment of inertia (2Ma2/5), we arrive at the following relation
between the spin and velocity of the ball immediately after the impulse:
5 ha 1
z 1  V
2 a2 x
The velocity of the ball at the point of contact with the table is
7a  5h
Vc  Vx 1  az 1  Vx 1
2a
If the ball is to roll without slipping (so Vc = 0), we find that
7
h a.
5
For a “high shot” with h > 7a/5, Vc is opposite in direction to Vx1. Since the friction at the
billiard cloth opposes Vc, it causes Vx to increase and z1 to decrease until pure rolling
sets in. For a “low shot”, Vc is in the direction of Vx1, and the friction force decrease Vx1
and increases z1 until rolling occurs.

191
Chapter 13

13.1 Two inertial frames S and S’ have their respective x-axes parallel, with S’ moving
with constant velocity v along the positive x-direction of the system S. A rod makes
an angle of 300 with respect to the x’-axis. What is the value of v if the rod makes
an angle of 450 with respect to the x-axis?
Solution:
We have L' y  L'sin 300  05 . L', and L' x  L'cos 300  0866
. L'
Since there will be a length contraction only in the x-x’ direction,
Ly  L' y  0.5L', Lx  L' x 1  v 2 / c 2  0.866 L' 1  v 2 / c 2
Now  = Ly/Lx, we have
0.5
tan 450  1 
0.866 1  v 2 / c 2
Solving for v
v  0816
. c.

13.2 The wave equation


1  2
  2
2
0
c  t2  
is satisfied in a vacuum for any component of E or B , and it is a consequence of
Maxwell’s equations. Show that this wave equation is invariant under a Lorentz
transformation. For simplicity, you can restrict yourself to the case in which the
wave propagates along the x-axis.
Solution:
If the wave propagates along the x-axis, the above wave equation reduces to
 2 1  2
 0
 t 2 c2  t 2
This equation will be invariant if it retains the same form when expressed in terms of the
new variables x’, y’, z’, t’. Now we find from the Lorentz transformation that
 x' 1  x' v
  
x 1  v 2 / c2 t 1  v 2 / c2
 t' v / c2  t' 1
 
x 1  v 2 / c2 t 1  v 2 / c2
From the chain rule and using the above results, we obtain
     x'    t ' 1  v 
   
 x  x'  x  t '  x' 1  v 2 / c 2  x' 1  v 2 / c2  t '
Differentiating again with respect to x, we have
 2 1   2 v 2  2  2v  2
  2  4 
 x2 1  v 2 / c 2   x' c  t ' 2  c 2  v 2  x'  t '
Similarly we have
 1  1 
 
t 1  v 2 / c 2  x' 1  v 2 / c2  t '

192
 2 1  2  2  2  2v  2
  v   
 t 2 1  v 2 / c 2   x' 2  t ' 2  1  v 2 / c 2  x'  t '
Substituting these in the wave equation, we obtain
 2 1  2  2 1  2
  
 x 2 c 2  t 2  x' 2 c 2  t ' 2
so that the one-dimensional wave equation is invariant under Lorentz transformations.
The generalization to three-dimensional case is an easy task, since
 y'  z'  x'  x'  y'
  1,      0
y z y z x
 2  2  2  2
and  ,  .
 y' 2  y 2  z 2  z' 2
Note that the wave equation is not invariant under Galilean transformations.

13.3 A light signal is emitted by observer 0 at exactly the moment when observer 0’

passes him with velocity v (relative to 0). Show that the velocity of light in any
direction is the same, as measured by 0 and 0’.
Solution:
We are required to prove that
x 2  y 2  z 2  c 2 t 2  x ' 2  y ' 2  z' 2  c 2 t ' 2  0 .
From the inverse Lorentz transformations we have
 x 'vt ' 
2
x ' 2 v 2 t ' 2 2vx' t '
x 
2
 
 1  v 2 / c2  1  v 2 / c2
 t 'vx ' /c 2 
2
v 2 x' 2 / c 4  t ' 2 2vx' t ' /c 2
t 
2
 
 1  v 2 / c2  1  v 2 / c2
and y 2  y' 2 , z 2  z' 2 .
Substituting, we find that
x 2  y 2  z 2  c 2 t 2  x' 2  y' 2  z' 2  c 2 t ' 2 .
Therefore, since x 2  y 2  z 2  c 2 t 2  0, we also have
x' 2  y' 2  z' 2  c 2 t ' 2  0,
so that the pulse as determined by 0’ is also spherical.

13.4 A track star on earth runs the 100 meter dash in 10.0 seconds, as measured on
earth. What is his “time” as measured by a clock on a spaceship receding from the
earth with speed 0.99c?
Solution:
Each clock measures the time interval between two events, namely the start and fin-
ish of the race. The proper time interval, , is that one measured by a clock traveling with
the runner. The runner moves at a speed of 10 m/s, that is very small compared with the
speed of light; accordingly, for all practical purposes, the time interval measured by a
clock stationary with respect to the earth is equal to . Thus the “time” as measured
aboard the space will be
s  

193
Now
  1  (0.99) 2   1  (1  0.01) 2   1  (1  0.02  0.0001)
1/ 2 1/ 2 1/ 2

  0.02  0.0001  10 2  0.01  10 1  0.005


1/ 2 1/ 2 1/ 2
2 1/ 2
 10 1  0.0025 / 1414
.  7.1
If the motion of the space ship is perpendicular to the length of the track, an observer in
the space ship will agree that runner did indeed run 100 meters and in 71 seconds. Should
the motion of the space ship be parallel to the track, then the observer in the space ship
will see the race was run over a shorter length of track as a result of the length contrac-
tion:
1
L  L0  100m / 7.1  14.1m .

13.5 A spaceship moves away from the earth with constant speed v  c / 2 . The astro-
naut observes that a rodlike external probe is 6 m long and makes an angle of 450
with the ship’s line of motion. To an observer on earth, how long is the probe and
what angle does it make with the line of motion?
Solution:
According to the earth observer, the component of the rod along the line of motion
(which we shall designate as the x-direction) will suffer a length contraction; whereas the
component of length of the rod perpendicular to the line of motion (which we designate
as the y-direction) will remain unchanged.
On the space ship:
: L0 x  L0 cos 450  L0 / 2 , L0 y  L0 sin 450  L0 / 2
On the earth::
Lx  1  v 2 / c 2  L0 x  (1  1 / 2) 1/ 2 L0 / 2  L0 / 2, Ly  L0 y  L0 / 2
1/ 2

Thus,
1 2 1 2 
 
1/ 2
1/ 2
L  Lx  L y
2 2
  L0  L0   4 / 3 L0  0.866  6 ft  519 . ft
4 2 
and
Lx L0 / 2 1
cos     0.577, and   550 .
L 3 L0 / 2 3

13.6 A spaceship moving away from the earth at a velocity v1 = 0.75c with respect to
the earth, launches a rocket with a velocity v2 = 0.75c in the direction away from
the earth. What is the velocity of the rocket with respect to the earth?
Solution:
The velocity of the rocket with respect to the earth is
v1  v2 0.75c  0.75c
V  2  0.96c .
1   v1v2 / c  1  (0.75c)(0.75c) / c
2

Thus, in spite of the fact that v1 + v2 = 1.5c, the actual velocity relative to the earth is less
than c.

194
13.7 An observer on earth sees two spaceships A and B approaching her along the same
straight line: A approaches from the left with speed c/2, and B from the right with
speed 3c/4. With what speed does each spaceship approach the other?
Solution:
Let us solve the problem from the point of view of the astronaut of A, who sees the
earth approaching him along his negative x direction with speed c/2. We shall denote his
system by S, and S’ the earth system which is moving with uniform velocity -c/2. Now
the spaceship B moves with respect to the earth with velocity -3c/4. Thus
v ' x v 3c / 4  c / 2 10
vx  2  2   c
1  v' x v / c 1  (3c / 4)(c / 2) / c 11
that is, B approaches A with speed (10/11)c along A’s negative x direction.
The problem can also be solved from the point of view of the astronaut of B. He
considers himself stationary and the earth approaching him along his positive x direction
with speed 3c/4. We now denote B’s system by S, and S’ denotes the earth system mov-
ing with constant velocity 3c/4. The space ship moves with respect to the earth with
velocity c/2. Thus
v ' x v c / 2  3c / 4 10
vx  2  2   c
1  v' x v / c 1  (c / 2)(3c / 4) / c 11
that is, A approaches B with speed (10/11)c along B’s positive x direction.

13. 8 A man on a station platform sees two trains approaching each other at the rate
7c/5, but an observer on one of the trains sees the other train approaching her with
a velocity 35c/37. What are the velocities of the trains with respect to the station?
Solution:
We use Eq.(13.24) to calculate the velocities of the trains with respect to the
station, with frame S at the station platform and frame S’ moving with train A. An
observer on train A sees train B approaching her with a velocity 35c/37; we shall denote
this velocity as v’. The velocity of train B with respect to frame S will be denoted by v.
Eq.(13.24) then gives
v 'u
v
1  uv ' /c 2
where u is the velocity of S’ with respect to S (It is also the velocity of train A with
respect to frame S.). Now u  v  7c / 5. Substituting yields
35C / 37  (7c / 5  v)
v
1  (35c / 37)(7c / 5  v) / c 2
Solving for v
v 2  3514
. cv  148. c2  0
from which we find
v  0.49c
and u  7c / 5  0.49c  0.91c
That is, the velocity of train A with respect to the station is 0.91c, and that of train B with
respect to the station is 0.49c.

13.9 Show that if two Lorentz transformations, with relative velocities given by v1 and

195
v2, respectively, are carried out consecutively, the result is the same as that of a
single Lorentz transformation with relative velocity v given by
  2
  1
1  1  2
where   V / c,  1  v1 / c, and  2  v2 / c .
Solution:
We now consider three inertial frames of references S, S’ and S” as shown in
Fig.13.9, in which frame S” has relative velocity v2 with respect to S’ and S’ has a

Fig.13.9
relative velocity v1 with respect to S. We want to find the relative velocity V of S” with
respect to S. We can use the velocity transformation equation, (13.24), again, with the
following replacements V  v, v1  u, v'  v2 :
v2  v1
V
1  v1v2 / c 2
where V is the relative velocity (in the x-direction) of S’ with respect to S. Dividing the
last equation by the speed of light c, we obtain the desired result,
  2
  1 .
1  1  2

13.10 Show that even if the relativistic mass is used, the expression mv2/2 is not the
correct relativistic energy.
Solution:
1 2 1 1
mv  m0 v 2  m0 v 2 (1   2 ) 1/ 2
2 2 2
(1)
1 2 1 2 
 m0 v 1      
2  2 
Now the correct relativistic expression for the kinetic energy is
T  mc 2  m0 c 2  m0 c 2 (  1)  m0 c 2  (1   2 ) 1/ 2  1
 3 
 m0 c 2 1   2    
 4 
1 2
which is different from Eq.(1). Hence, mv is not the correct relativistic kinetic energy.
2

13.12 A pion of mass  comes to rest and then decays into muon of mass  and a neutri-
no of mass zero. Show that the kinetic energy of the muon is c2( - )2/2.
Solution:
   
Ebefore   c 2 , pbefore  0; Eafter  E   E , pafter  p   p
 
Conservation of momentum requires that p    p , conservation of energy says that
E   E   c 2 .

196
But E   p c, E   2c4  p 2c2 , and p   p
Hence p c   2c4  p 2c2   c2
0r  2 c4  p 2c2   c2  p c
 2  2
from which we find p  c
2
T  E    c 2   2 c 4  p  2 c 2   c 2
 2   2 
Now   c2  p c   c2   c2   c c   c 2
 2 
 2  2  +  2 2 (   ) 2 2
 c  c .
2 2

13.13 Two identical bodies, each with rest mass m0, approach each other with equal velo-
cities u, collide, and stick together in a perfectly inelastic collision. (a) What is the
rest-mass of the composite body? (b) What is the rest-mass of the composite body
as determined by an observer who is at rest with respect to one of the initial bodies?
Solution:
(a) Since the initial velocities are equal in magnitude, the initial momentum of the
system is zero. Conservation of momentum requires that the final momentum must be
zero. Eq.(13.54) then gives
2m0 c 2
 M 0c2
1 u / c
2 2

2m0
from which we have M0   2m0
1  u2 / c2
(b) Consider the body, A, that moves in the +x-direction. The velocity v of the
observer 0’ at rest with respect to A is equal to A’s velocity, v = u. The second body, B,
has a velocity uB = -u as measured by 0. Its velocity as measured by 0’, u’B,, is obtained
from the Lorentz transformation:
uB  c 2u
u' B  2  
1  uB v / c 1  u2 / c2
Since the composite body, C, is at rest with respect to the laboratory (observer 0), its ve-
locity with respect to 0’ is u’C = -u. From conservation of momentum, as measured by 0’
m0 u' A m0 u' B M 0 u' C
 
1  uA2 / c2 1  u' B 2 / c 2 1  u' C 2 / c 2
But u’A = 0, so
2u
m0 M 0 (  u)
1  u2 / c2

 2u / c 
2
1  u2 / c2
1  
1  u2 / c2 
Solving for M0, we find the same result:

197
2m0
M0  .
1  u2 / c2

13.14 A particle of rest-mass m0 and velocity v = 3c/5 collides with a stationary particle
of rest-mass m0. Assuming that, after collision, the two particles stick together; find
the velocity and the rest-mass of the composite particle.
Solution:
From pfinal = pinitial, we have
M 0u f m0 ui m0 (0.6c)
   0.75m0 c
1 u f / c
2 2
1  ui / c
2 2
1  (0.6) 2
From Efinal = Einitial, we have
M 0c2 m0 c 2 m0 c 2
  m0 c 2
  m0 c 2  2.25m0 c 2
1 u f / c
2 2
1  ui / c
2 2
1  (0.6) 2

Solving the two equations simultaneously, we obtain


u f  0.333c, and M 0  212
. m0 .

13.15 Consider a system of non-interacting particles. Show that its rest-mass M exceeds
the sum of rest-masses of constituent particles by the total kinetic energy of the par-
ticles (divided by c2):
M   m j  2  Tj .
1
j c j
Solution:
Energy and momentum of the system are additive:
n n
E system   Ei psystem   pi
i 1 i 1
From these sums the rest mass of the system, M0, can be evaluated by Eq.(13.40):
( E system ) 2 ( psystem ) 2
M 2

c4 c2
This relation can be simplified when the system of particles in free motion is viewed from
an inertial frame so chosen as to make the total momentum zero. Then we have
E system
M .
c2
Now, energy of each particle can always be expressed as sum of rest energy and kinetic
energy:
Ei  mi c 2  Ti (i  1, 2, 3,  , n)
So the rest mass of a system exceeds the sum of rest masses of its individual particles by
an amount equal to the total kinetic energy of all particles (as seen in frame in which total
momentum is zero!):
n
1 n
M   mi  2  Ti .
i 1 c i 1

13.16 A particle of rest-mass m0 moves under the action of a constant force. Find the time

198
dependence of the particle’s velocity.
Solution:
The equation of motion takes the form
d 1
(m0 v)  F ,   ,  v/c
dt 1  2
where F = constant. Integrating with the initial condition v = 0 at t = 0, we obtain
m0 v
m0 v   Ft
1  v 2 / c2
A velocity as a function of time is found from the above equation algebraically:
Ft / m0
v .
1  ( Ft / m0 c) 2
Note that Ft/m0 is the non-relativistic velocity. The non-relativistic equation of motion is
of the form
d
(m v )  F
dt 0 NR
where vNR is the non-relativistic velocity. Integration with the same initial condition v = 0
at t = 0 gives
m0 v NR  Ft , or v NR  Ft / m0
Accordingly, we can v as
v NR
v .
1  (v NR / c) 2

13.17 A particle of rest-mass
 M0 moves with an instantaneous velocity v under the
action of a force F . Show that

(a) if F is parallel to v , then
    1
F  M 0 3a , where a  dv / dt , and   ;
  1  v 2 / c2

(b) if F is perpendicular to v , then F  M 0 a .
Solution:
 m0 v
(a) p   m0 v 
1  v2 / c 2
 
  m0 (dv / dt ) d
F  dp / dt   m 0 v 1 / 1  v 2 / c2
1 v / c
2 2 dt
   
  m0 a  m0 v (1 / 2)(2v / c ) 3 (dv / dt )   m0 a   3 ma (v 2 / c 2 )
2

 
  3 m0 a  1  v 2 / c 2  v 2 / c 2    3 m0 a
(b) Let the instantaneous momentum be in the x direction and the force be in the y
   
direction. Then dp  Fdt   m0 a and dv will also be in the y direction. So we have
  
F   m0 (dv / dt )   m0 a .

13.18 As viewed by an observer in the laboratory, a proton collides with another proton
initially at rest. After the collision, a proton and an antiproton come off in addition

199
to the original protons. Find the minimum kinetic energy that the incoming proton
must have to make this reaction energetically possible.
Solution:
We first view this collision process from the point of view of the center-of-mass
sys-tem, where the total momentum is zero. Before the collision, the two original protons
approach each other with momenta of equal magnitude but opposite directions; in
addition, since the two protons have equal rest masses, Eq.(13.40) tells us that their total
energies are equal as well. After the collision, we have four particles: three protons and
one anti-proton.
Conservation of energy says that the total energy of these four particles must equal the
total energy of the two protons before the collision. Now the least energy that the four
particles can have in the center-of-mass system is just their rest mass energies; that is,
each of the four particles comes off with zero kinetic energy (and hence each with zero
momentum) in the center-of-mass system. The laboratory observer sees all four particles
move off with the speed of the center-of-mass. Then, calling E’1 and E’2 the total energies
of the two original protons in the center-of-mass, we have
E '1  E ' 2  4 M 0 c 2
since each of the four final particles has a rest mass energy of M0c2.
To obtain the energy in the laboratory system, we make use of the Lorentz transformation
for momentum and energy. If the center-of-mass moves along the positive x-direction in
the laboratory, then
E '1   ( E1  vp1 x ), E ' 2   ( E2  vp2 x )
where primes denote the center-of-mass quantities, v is the velocity of the center-of-mass
and 1 and 2 denote the incoming and stationary protons. It is more convenient to use the
inverse transformations obtained by interchanging primed and unprimed quantities and
reversing the sign of v, thus
E1   ( E '1 vp'1x ), E2   ( E ' 2 vp' 2 x )
Adding these two equations, we obtain
E1  E2   ( E '1  E ' 2 )   v( p'1x  p' 2 x )   ( E '1  E ' 2 )  4 M 0 c 2
The second equality follows from the fact that p'1x  p' 2 x  0; the right-hand equality from
the first equation. We now make use of the velocity of the center-of-mass to evaluate .
The velocity of the center-of-mass is (to be proved later)
v  p1x c 2 / ( E1  E2 ) .
Thus,
  (1  v 2 / c 2 ) 1/ 2  1  p1x 2 c 2 / ( E1  E 2 )
1/ 2

 ( E1  E 2 ) / ( E12  2 E1 E 2  E 2 2  p1x 2 c 2 ) 1/ 2
Since the motion of the incoming proton is taken along the x-direction, E12-p1x2c2 =
M02c4; the second proton is at rest, so E2 = M0c2. Thus,
  ( E1  E2 ) / (2 M 0 c 2 E1  2 M 0 2 c 4 )1/ 2
Substituting this value into the equation relating the energies in the laboratory and center-
of-mass systems, we have
E1  E2  4 M 0 c 2 ( E1  E2 ) / (2 M 0 c 2 E1  2 M 0 2 c 4 )1/ 2
or 2 M c0
2
E1  2 M 0 2 c 4 
1/ 2
 4 M0c2

200
Squaring and solving for E1, we find
E1 = 7M0c2
Finally, the kinetic energy, T1, is given by
T1 = E1 - M0c2 = 6M0c2 = 5.6 GeV.
The incoming proton must have at least this kinetic energy for this process to be energeti-
cally possible.
The velocity of the center-of-mass with respect to the laboratory system: The di-
rection of the incoming proton is taken along the x-axis, then as measured by an observer
in the laboratory system, the incoming proton and the stationary proton have respective
components of momentum
p1x  M 0 v / 1  V 2 / c 2 , p1y  p1z  0
p2 x  p2 y  p2 z  0
Since the center-of-mass system moves with speed v along the postive x-direction of the
laboratory system, the Lorentz transformations of momentum and energy give
p'1y  p1y  0, p' 2 y  p2 y  0; p'1z  p1z  0, p' 2 z  p2 z  0
and p'1x   ( p1x  E1v / c 2 ), p' 2 x   ( p2 x  E2 v / c 2 )
where primes denote the center-of-mass quantities, and   1  v 2 / c 2 . Now the total
momentum in the center-of-mass system is zero:
 
p'1  p' 2  i( p'1x  p' 2 x )  j ( p'1 y  p' 2 y )  k( p'1z  p' 2 z )
 i p1x  ( E1  E 2 )v / c 2   j (0)  k(0)  0
Solving for v, we obtain
p1x c 2
v q.e.d.
E1  E 2

13.19 The relativistic Doppler Effect. A light source flashes with period 0 = 1/0 in its
rest frame, and the source moves towards an observer with velocity v. Show that
the frequency of the pulses  received by the observer is
1  v / c 
1/ 2

  0
 1  v / c 
and that if the observer is at angle  from the line of motion, D is given by
(1  v 2 / c 2 ) 1/ 2
  0 .
(1  v cos / c)
Solution:
Let us consider the receiver fixed in system S and the light source in S’
moving to-wards the receiver with velocity v. The source emits n waves during the time
interval T. The length of wave train (i.e., the total distance between the front and rear of
the wave train emitted during the time interval T) is
length of wave train = cT-vT
Since there are n waves emitted during this time interval T, the wavelength  must be
  (cT  vT ) / n
and the frequency,  = c/, is
  cn / (cT  vT ) (1)

201
In its rest frame, the source emits n waves of frequency 0 during the proper time 0, and
n = 00 (2)
The proper time interval 0 measured on the clock at rest in the moving system S’ is
related to the time interval T measured on a clock fixed with the receiver in system S by
0 = T/ (3)
where   1 / 1  v / c .
2 2

We substitute (3) into (2) to determine the number of waves n. & then n is substituted
into (1) to determine the frequency
c T /  1 0 1 v / c
 0   0
cT  vT 1  v / c  1 v / c
1 v / c
 0 (Source & receiver approaching)
1 v / c
It is straightforward to show that the above result is also valid when the source is fixed
and the receiver approaches it with velocity v. It is the relative velocity v that is
important.
When the source and receiver are receding from each other with velocity v, we have
1 v / c
 0 (Source & receiver receding)
1 v / c
The derivation is similar to the one just done above, except that the distance between the
beginning and end of the wave train becomes
length of wave train =cT + vT
and this change in sign is propagated throughout the derivation.
When the source and receiver move at an angle with respect to one another, the above
results are not valid any more. We now give a new derivation for this general case. Before
we do so, let us first consider how the differential operators /x, /y, /z, /t trans-
form under Lorentz transformation. By chain rule, we see that they transform according to
the following rules
 v 
  x'   t'  
  
 x' c 2  t ' (4)
 x  x  x '  x  t' 1  v 2 / c2
 
  x'   t ''   v
  
 t'  x' (5)
 t  t  x'  t  t' 1  v / c2
2

   
  (6)
 y  y'  z  z'
Now we consider, in the system S’, a plane wave propagating in the direction given by
the direction cosines  ',  ' ,  ' . Such a wave may be described in S’ by the wave
function
   ' x'  ' y' ' z' 
  A exp2 i  0 t ' 
  c 
where ‘ +‘ +‘ = 1. This function is a solution of the wave equation of 13.2. Now, the
2 2 2

same plane wave is described in S by the function

202
   x   y   z 
  A exp2 i t  
  c 
By partial differentiation, we see that, symbolically, x = /x = -i2/c, y = -i2/c,
z = -i2/c, t = i2. Similar expressions also hold for ‘x, ‘y et cetera. Putting these
into formulas (4), (5), and (6) we obtain
1  (v / c)  v/c
0   ,  '
1 v / c
2 2
1  v 2 / c2
1  v 2 / c2 1  v 2 / c2
 '  ,  ' 
1  (v / c) 1  (v / c)
The first of these expressions is the Doppler shift of frequencies for the general case. It is
more useful to recast this as
1  v 2 / c2 1  v 2 / c2
  0  0
1  (v / c) 1  (v / c) cos
It is easy to see that for  = 0, the above formula reduces to the one for “approaching”,
and for  = 1800, it reduces to the one for receding.

13.20 Find the relativistic path of an electron moving in the field of a fixed-point charge.
Solution:
This is a relativistic two-body central force problem. The general case in which
both particles move around a common center of mass is very difficult. If, however, one of
the particles is very heavy so that it is stationary, the relativistic problem is hardly more
difficult than the non-relativistic problem.
The Lagrangian of our electron can be written as
qQx 0
L  m r 2  r 2 2  x 0 2  
1
(1)
2 rc
where the stationary charge Q, which is setting up the field, is fixed at the origin. Q/r is
the potential through which our electron moves and qQ/r is the electron’s potential ener-
gy. It is easy to see that  is a cyclic coordinate. Therefore, corresponding to , we find
d
p   L /   mr 2  mr 2    const . , (2)
dt
which is just the angular momentum. Now x0 = ct is also a cyclic coordinate;
corresponding to x0, we have
E  qQ 
p0   mx 0 2   (3)
c  rc 
There is a third integral of the motion, namely
r 2  r 2 2  x 0 2  c 2 (4)
The radial equation of motion follows from
d L L
 0
d  r  r
qQ
and is m(r  r 2 )  2 x 0  0 (5)
r c

203
As it turns out, all four of the equations (2), (3), (4) and (5) are not independent. Any one
of them can be obtained from the other three. The elimination of x 0 between (3) and (5)
gives the result
qQ  E qQ 
r  r 2 )  2 
m(  0
r c  mc mrc 
Through using the angular momentum equation (2), we can write an orbit equation in
terms of the variable y = 1/r:
 qQ 
2
d2y qQE
 y   2 2 ,
2
  1  
2
(6)
d 2  c  c 
The solution is
1 qQE
  2 2 2  A cos(  ) (7)
r  c
Since the parameter  is not unity, the orbit does not close on itself but is a precessing
fig-ure. The integration constant A can be determined. Let us first express Eq.(4) in the
form
 dy   E qQy   mc 
2 2 2

   y  2
    (8)
 d   c c    
By substituting the solution (7) into (8) and solving for A, we find
 E   mc 
2 2

A     .
 c 2    
Now let us return to Eq.(6). Notice that non-periodic solutions of Eq.(6) exist when 2 <
0. Let us suppose that an electron in an atom has an angular momentum of order h . Let Q
= Zq where Z is the atomic number of the atom. Now if
q2Z 
2

  1 
2
  0
 
 hc
then

hc
Z  2  137
q
Thus atoms of greater atomic number than 137 presumably could not exist because of a
kind of relativistic collapse. The quantity
q2 1


hc 137
is known as the fine structure constant in atomic physics.

204
K22556
ISBN: 978-1-4822-2780-2
90000

w w w. c r c p r e s s . c o m
9 781482 227802

You might also like