Part IB Physics B 2010-11

Answers to Classical Dynamics Examples 1

These worked answers to the Dynamics problems are oered to students and supervisors as
a guide, and I hope they contain some generally useful hints. In some cases, I have indulged
in a little research that goes beyond the stated problem. Where possible I have included
illustrative diagrams.
Health Warning
These solutions must be used with caution.
The worked answers will be useless to you unless you have already made a very
serious attempt to solve the problem yourself and have discussed it with your
supervisor. If you consult the solutions earlier, they will just act as spoilers.
You must remember that there is often more than one way to solve a physics problem.
I have sometimes suggested alternatives, but your supervisors will know many others.
The most dicult part of a physics problem is knowing where to start. It is therefore
quite possible that I taken too much for granted at the beginning of some problems. If
so, please let me know.
I have often given the numerical answers to several more decimal places than is really
justied, so that you can check your calculations (and mine) carefully. I have used
values of the physical constants taken from the formula book that you will use in the
Examination.
Please let me me know about any errors, typos and suggestions for improvement.
Steve Gull
email: steve@mrao.cam.ac.uk
http://www.mrao.cam.ac.uk/~steve/part1bdyn
1. The moment of inertia of the cylinder is
1
2
ma
2
, so its initial kinetic energy is
1
4
ma
2

2
.
The friction F(t) slows the rotation and increases the linear velocity v:
m v = F ; I = Fa so that I(t) +mav(t) is a constant (= I). The friction stops
at

when the cylinder rolls v

= a

, so
1
2
ma
2

+ma

=
1
2
ma
2
, so

=
1
3
. The
energy is evaluated as
1
4
ma
2
(

)
2
+
1
2
ma
2
(

)
2
=
1
12
ma
2

2
, i.e. 2/3 has been lost.
2. The coordinates of the centre of mass are
x =
1
2
l sin ; y =
1
2
l cos and the velocities are
x =
1
2
l cos

; y =
1
2
l sin

. The kinetic energy is
T =
1
2
m( x
2
+ y
2
) +
1
2
I

2
, where the moment of inertia
is I =
1
12
ml
2
. We nd T =
1
6
ml
2

2
.
The potential energy is V =
1
2
mgl cos .
Using the energy method
d
dt
(T +V ) = 0 we derive the
equation of motion

=
3g
2l
sin .
The story doesnt quite end there, because the reaction force at the wall must
eventually go negative, so the ladder will actually lose contact with the wall. This is
easy to prove by Newtonian methods, and indeed by Lagrangian methods if you know
what youre doing. . . We nd N =
3
4
mg sin (3 cos 2 cos
0
) where
0
is the initial
angle when the ladder was at rest.
R-r
r
q
w
3. The coordinates of the centre of mass are
x = (R r) sin ; y = (R r) cos . As varies
we see that the velocity of the centre of mass is (R r)

.
As the cylinder rolls, the velocity of its centre of mass
must just be r, so =
(R r)
r

as required.
The kinetic energy is
T =
1
2
I
2
+
1
2
mr
2

2
=
3
4
mr
2

2
=
3
4
m(R r)
2

2
.
The potential energy is V = mg(R r) cos .
Using the energy method the EoM is

+
2g
3(R r)
sin = 0.
The angular frequency of small oscillations is
2
= 2g/3(Rr).
w
q
q
g
w r
2
4. The centrifugal force per unit mass is
2
x and the
surface satises tan = dy/dx =
2
x/g.
We get y =
2
x
2
/2g and using the formula
given we get
2
= g/2f = 1.566 rad s
1
.
mg
a
h
A
B
A
B
R
R
2m v W l sin
5. Taking moments about the left-hand wheel:
R
B
a =
1
2
mga + 2hmvsin
so that R
B

1
2
mg = 2hmvsin /a.
The reaction R
A
= mg R
B
so the dierence is
4hmvsin /a. Comparing this to the average
reaction
1
2
mg we get a fractional dierence
R
B
/R
A
1 + 8hvsin /ag.
6. Method 1 In coordinate system used in lecture v = gt and m x = 2mgtcos and so
x =
1
3
gt
3
for a point on the equator. Using h =
1
2
gt
2
we get
x =
1
3

_
8h
3
/g = 0.245455 m to the East. Thats using the sidereal day of
23.9344696 hr, of course.
Method 2 The stone falls from height H above the Earths surface at radius R,
rotating at angular velocity wrt the inertial frame. Its angular momentum is
(R+H)
2
and is constant. At height h its angular velocity is = (R+H)
2
/(R+h)
2
and its relative velocity wrt an observer rotating with the Earth is
( )(R +h) = R(1 +H/R)
2
(1 +h/R)
1
(R +h) 2(H h)
Using H h =
1
2
gt
2
we recover the above formula x =
1
3
gt
3
.
7. The winds circulate anticlockwise, approximately
parallel to the isobars. The force on a unit volume
of air due to the pressure gradient dP/dr (inwards)
and the Coriolis force 2vsin is outwards.
The equation of motion for a unit volume of air
at radius r in a frame rotating with the Earth is

_
r
v
2
r
_
=
dP
dr
+ 2vsin
Setting r = 0 we get a quadratic v
2
+ 2vr sin
r

dP
dr
= 0. The data given imply a
density for air of 1.16 kg m
3
(using a molecular mass of 29 g) and the resulting wind
speed is 27 km hr
1
.
You can also formulate the problem in the accelerated frame moving with the air, in
which there is a centrifugal term v
2
/r, conrming the above result.
It is interesting to note that, if you tried to do the same problem for an anticyclone,
there are some sign changes and the resulting quadratic has no real solutions. The
reason is that cyclones tend to wind themselves up and increase the pressure gradient,
whereas anticyclones tend to disperse and decrease the pressure gradient. Hence, no
anticyclone can have a pressure gradient as high as that given in the question.
r/a
V(r)
8. The eective potential is V (r) =
k
r
exp(r/a) +
J
2
2mr
2
.
This is illustrated in the gure opposite for the cases
where r/a is less than, equal to and greater than the
critical value r/a =
1
2
(1 +

5).
The exponential fall-o eventually causes the potential
to be insuciently attractive to permit bound states.
If there is a minimum in the potential it occurs at r = r
0
where
dV
dr

r
0
=
k
r
2
exp(r/a)
_
1 +
r
a
_

J
2
mr
3
= 0
We now form d
2
V/dr
2
to see if there is a minimum there
d
2
V
dr
2
=
k
r
3
exp(r/a)
_
2 +
2r
a
+
r
2
a
2
_
+
3J
2
mr
4
Using the value of J
2
at r
0
above we get
d
2
V
dr
2

r
0
=
k
r
3
exp(r/a)
_
1 +
r
a

r
2
a
2
_
which conrms that the orbit is stable below the critical value of r/a =
1
2
(1 +

5).
q
q
0
V( ) q
9. The potential energy V and kinetic energy T are
V () = mgl(1 cos ) , T =
1
2
ml
2
_

2
+ sin
2

2
_
The angular momentum about the polar axis is
J = ml
2
sin
2

and is conserved.
We can therefore substitute for

to get the energy
equation as required: E = U
e
+
1
2
ml
2

2
,
where U
e
= mgl(1 cos ) +
J
2
2ml
2
sin
2

.
For the circular orbit, we locate the bottom of the U
e
potential well
dU
e
d
= 0 = mgl sin
J
2
cos
ml
2
sin
3

. (0.1)
Solving this for J
2
= m
2
l
3
g sin
4
/ cos and comparing with our earlier form for J
yields the steady angular velocity =

2
=
g
l cos
. (0.2)
Dierentiating again, we nd
d
2
U
e
d
2
= mgl cos +
J
2
ml
2
sin
4

_
sin
2
+ 3 cos
2

_
. (0.3)
and substituting for J
2
as above yields
d
2
U
e
d
2
=
mgl
cos
_
1 + 3 cos
2

_
=
2
ml
2
_
1 + 3 cos
2

_
. (0.4)
This is the required result, but many students are unsure how to nish it o
rigorously. We write a quadratic approximation for the potential
U
e
() = V (
0
) +V

(
0
) +
1
2
()
2
V

(
0
) . (0.5)
Now take the energy equation
E = U
e
() +
1
2
ml
2

2
= constant (0.6)
and dierentiate with respect to time
0 =

_
V

+ml
2

_
, (0.7)
which is SHM at the required frequency
2
=
2
(1 + 3 cos
2
).
m
2
l sin

2ml cos

2ml sin

Optional part
A proper solution of this using Newtonian methods
(whether or not you are in the rotating frame) is rather
dicult, as it requires the use of the formulae for
the velocity and acceleration in spherical polar coordinates.
For this case the particle is constrained to the sphere
r = l, so it simplies somewhat:
velocity : l

+l sin

e

acceleration : l(

2
+ sin
2

2
) e
r
+l(

cos sin

2
) e

+l(sin

+ 2 cos

) e

where we recognise, for example, the rst term as the centripetal acceleration.
In the frame rotating at the steady state
2
= g/l cos
0
we have a centrifugal force
m
2
l sin e

, and components of the Coriolis force 2ml(sin

e

cos

).
The full equations of motion are then
ml(

2
+ sin
2

2
) = T +mg cos +ml
2
sin
2
+ 2mlsin
2

ml(

cos sin

2
) = mg sin +ml
2
sin cos + 2mlsin cos

ml(sin

+ 2 cos

) = 2mlcos

Phew! Since were going to expand to rst order about

0
we didnt really need to
worry about the terms that are second-order in the velocities, and we dont actually
need the rst equation at all, since it simply tells us what the tension in the string
is. . .
We now set sin sin
0
+ cos
0
and cos cos
0
sin
0
and use
mg = ml
2
cos
0
to get the much simplied set of equations
ml

= ml
2
sin
2

0
+ 2mlsin
0
cos
0

ml sin
0

= 2mlcos
0

Integrating the second equation to get sin

0

= 2cos
0

1
, and substituting in the
rst, we recover

=
2
(1 + 3 cos
2

0
)
10. This is a very instructive short question.
(a) If the impulse is in the direction of motion, then the
velocity is still perpendicular to the radius vector, but
the energy is increased. Assuming that the orbit is still
bound, the satellite moves to an elliptical orbit with a
larger semi-major axis. The only points on the orbit
where the velocity is perpendicular to the radius is at the
points of furthest and closest approach. Here it is clearly
the point of closest approach.
(b) Again, assuming that the velocity is not reversed, the
point of impact must be at the point of furthest approach.
(c) The impulse is directed away from the central mass, so
the angular momentum and hence the semi-latus rectum
are unchanged. The only two points at which the orbit
crosses the distance of the semi-latus rectum are at
= /2 as illustrated.
11. The satellite starts o in a circular orbit with a = r potential energy GMm/r and
kinetic energy GMm/2r. After the burn it has kinetic energy GMm/8r, making a
total of 7GMm/8r, so that the new semi-major axis is 4r/7. we want this to be
equal to (1 +)r/2, so that the satellite grazes the surface. Hence = 7 and e =
3
4
.
The alternative burn actually increases the kinetic energy to 5GMm/8r, but doesnt
aect the angular momentum or the semi-latus rectum, which remains r
0
= r. The
new total energy is 3GMm/8r and the semi-major axis has increased to 4r/3, so
that (1 e
2
)a = r
0
1 e
2
= 3/4, e =
1
2
, r
min
= 2r/3, r
max
= 2r and the satellite
does not hit the surface.
12. This question refers to pulsar formation, and explains their very wide distribution
about the Galactic plane compared to the normal stellar population.
Denote the masses m and m and the angular velocity of the original circular orbit as
. Consider the balance of gravitational and centrifugal forces on either star to get
the formula

2
=
Gm(1 +)
r
3
. (0.8)
After the explosion, the initial velocities do not change, but the centre of mass of the
remaining stars is dierent, so that they each have a speed r/2 relative to the new
centre of mass. (This is the trickiest step in the problem.) To see if the new orbit is
bound, we simply evaluate the total energy
E =
Gm
2
r
+ 2
1
2
m(r/2)
2
=
Gm
2
4r
( 3) . (0.9)
The orbit is unbound if > 3.
1
We have set a constant of integration to zero here. If it were non-zero, the resulting small oscillations in
would be oset from
0
in the appropriate way.
13. The radius and velocity are correct, hence the energy and orbital period are correct
too, and the distance a given in the question is indeed the semi-major axis of the
orbit. From the polar equation of the ellipse we conclude that a(1 e
2
) = r
0
and that
the maximum and minimum distances are rmax
min
= r
0
/(1 e) = a(1 e). However, the
angular momentum is incorrect J
2
= GMm
2
a cos
2
= GMm
2
r
0
= GMm
2
a(1 e
2
).
Hence we conclude that e = sin as required.
The new orbit has a semi-major axis a

= a(1 +e), so by Keplers third law

T

= T(1 +e)
3/2
1 day + 37.6 s.
The only purpose of putting a satellite in a geostationary orbit is that is has a period
of 1 day. Since it now doesnt have this period, it was a big mistake to make the
correction. . .
Alternative
Most of these orbit problems can also be done by using the eective potential. Here
the eective potential of the orbit is
U
e
=
GMm
r
+
J
2
2mr
2
. (0.10)
Dierentiate this to nd the radius of the circular orbit a, and evaluating the angular
momentum J
0
and the energy E
0
J
2
0
= GMm
2
a , E
0
=
GMm
2a
.
Now suppose that, as stated, the velocity v is misaligned by , so that the energy is
unchanged, but the new angular momentum is J = J
0
cos . We can now solve the
equation U
e
= E
0
to nd the new turning points:
E
0
=
GMm
r
+
J
2
0
cos
2

2mr
2
.
Rearranging, we get r
2
2ra +a
2
cos
2
, so that, as required, r = a (1 sin ) .
14. The polar equation of the ellipse r
0
= r(1 +e cos ) is truly marvellous. For e > 1 we
have hyperbolic orbits and the major axis is now negative, but is still related to the
semi-latus rectum r
0
= a(1 e
2
) and the energy (E = A/2a, now positive) in the
same way. When you set >

in the formula, r becomes negative, the orbit

switches to the repulsive branch, but all the physical results remain valid. At = ,
which is the distance of closest approach c on the repulsive branch, we still get
c = r
0
/(1 e) = a(1 +e).
We get e from sec

= e and relate this to the scattering = 2

(in lectures
I dened the scattering angle = = 2

since it is a negative angle, but here

is clearly intended to be positive), so e = cosec/2.
The semi-major axis a is related to the energy in the usual way (see above), so
c =
zZe
2
8
0
E
(1 + cosec/2)
where Z = 82 is the charge on the nucleus and z = 2 is the charge of the
alpha-particle. The distance of closest approach for = 60

is 1.4 10
14
m.
F
impact parameter
f
b
r
Alternative
Since the use of negative r has a slightly
Alice through the looking-glass quality to it, I oer
an alternative. This ab initio derivation is known as
the the linear momentum trick.
The revised geometry is shown in the diagram.
Take the initial velocity as v

along the x-direction

and the impact parameter as b, so that the angular
momentum is J = mbv

and energy is
1
2
mv
2

.
Note that the angular momentum at angle and radius r is J = mr
2

The change in
x-momentum scattering to angle is
mv

(1 cos ) =
zZe
2
4
0
_

0
dt
cos
r
2
=
zZe
2
m
4
0
J
_

0
d cos =
zZe
2
m
4
0
J
sin .
Using 2 sin
2
= (1 cos 2) and 2 sin cos = sin 2, we can quickly rearrange this
to get
b =
zZe
2
8
0
E
cot /2 .
The next trick is to note that the velocity v at the radius c of closest approach satises
cv = bv

,
1
2
mv
2

=
1
2
mv
2
+
zZe
2
4
0
c
.
This neatly unwraps to give a quadratic for c
c
2
2c
zZe
2
8
0
E
b
2
= 0 = c
2
2cb tan /2 b
2
,
so that c/b = tan /2 + sec /2 (positive root) and
c =
zZe
2
8
0
E
(1 + cosec/2) .
15. The disc has moments of inertia I
1
= I
2
=
1
4
ma
2
= 10
3
kg m
2
and
I
3
=
1
2
ma
2
= 2 10
3
kg m
2
. It is rotating at 45

to its axis, so let

= (1/

2, 0, 1/

2) wrt the body axes. Using J = (I

1

1
, I
2

2
, I
3

3
) we get
J = (1/

2, 0,

2) 10
3
kg m
2
s
1
and T =
1
2
J = 0.75 10
3
J.
16. The vertical motion is completely decoupled from the rotation and the horizontal
motion so that, on bouncing elastically, u
2
= u
1
.
The oor gives a horizontal impulse P (P negative as dened) so that the momentum
and angular momentum equations are
P = m(v
2
v
1
) ; Pa = I(
2

1
) I(
2

1
) = ma(v
2
v
1
)
The energy equation can be written
1
2
I(
2
2

2
1
) =
1
2
m(v
2
2
v
2
1
) so that
v
2
+a
2
= (v
1
+a
1
), i.e. the velocity of the point of impact reverses as usual.
Substituting for
2
we nd v
2
(Ma
2
/I + 1) = 2
1
+ (ma
2
/I 1)v
1
and, using
I =
2
5
ma
2
, we get v
2
= (3v
1
4a
1
)/7,
2
= (10v
1
/a + 3
1
)/7.
17. The moments of inertia (wrt the CoM) about the x- and y-axes are I
x
=
1
3
mb
2
and
I
y
=
1
3
ma
2
respectively. Suppose that the impulse of the collision is P, the resulting
velocity of the centre of mass is v and the nal angular velocities are
x
and
y
.
Whilst this problem can be done in a variety of ways, I strongly advise that
everything is considered wrt the CoM, i.e. as an impulse P to the CoM and as angular
impulses Pb and Pa about the x- and y-axes.
The equations for the linear momentum, angular momentum and energy conservation
then become
P = m(v +u) (1)
Pb = I
x

x
(2)
Pa = I
y

y
(3)
1
2
mu
2
=
1
2
mv
2
+
1
2
I
x

2
x
+
1
2
I
y

2
y
(4)
Eliminating P between (1) and (2) we nd I
x

x
= mb(u +v) and similarly
y
is found
from (1) and (3): I
y

y
= ma(u +v). Substituting into (4) we get
m(u
2
v
2
) =
m
2
b
2
I
x
(u +v)
2
+
m
2
a
2
I
y
(u +v)
2
= 6m(u +v)
2
where we have used the formulae above for the moments of inertia.
Although this is a quadratic, we already know that v = u is a solution (i.e. P = 0)
so, cancelling the factor of u +v and rearranging, we see that the other solution is
v =
5
7
u. We also nd
x
=
6
7
u/b and
y
=
6
7
u/a, so that the instantaneous velocity
of the point of impact is v +b
x
+a
y
= u, i.e. it has reversed.
18. The collision is relatively straightforward again, with I =
1
4
ma
2
and impulse P,
velocities u

and v for the mass and the disc afterwards, and angular velocity , we get:
P = m(u u

) (1)
P = mv (2)
Pa/2 = I (3)
1
2
m(u
2
(u

)
2
) =
1
2
mv
2
+
1
2
I
2
(4)
We nd that = mva/2I I
2
= mv
2
for this case. Substituting u

from (1) into

(4) we nd (of course) v = 0 or, more interestingly, v =
2
3
u, u

=
1
3
u, =
4
3
u/a.
It is now supposed that the disc was rotating at
2
3
u/a. This implies that
J
3
=
1
3
mau = J
1
, since I
3
=
1
2
ma
2
. The total J =
1
3

2mau, and the space precession

frequency is J/I
1
=
4
3

2u/a. The disc returns to its original spatial orientation in a

time 2/
s
or in a distance a/

2.
19. I give a much more complete answer to this question in the Appendix. The coin has
radius a, mass m and moment of inertia I =
1
4
ma
2
. Suppose the impulse is P and
that the velocities before and after are u and v as in the previous question. If the
angular velocity afterwards is , we can write the equations of linear momentum,
angular momentum and energy conservation as
P = m(v +u) (1)
Pa cos = I (2)
1
2
mu
2
=
1
2
mv
2
+
1
2
I
2
(3)
where is the angle the coin makes to the horizontal Eliminating P between (1) and
(2) we nd I = ma(u +v) cos , and substituting into (3) we get
m(u
2
v
2
) =
m
2
a
2
cos
2

I
(u +v)
2
= 4m(u +v)
2
This again has u +v as a factor. For small we can set cos = 1 implying v =
3
5
u
and =
8
5
u/a. This means that the velocity of the point of impact is v +a = u (i.e.
it reverses) and that the velocity of the other edge is v +a =
11
5
u.
For small angle , the other edge is approximately 2a above the ground. By the time
the edge hits the ground the original point of impact has travelled
5
11
2a upwards, so
that the angle the coin makes is
5
11
.
20. The model for the equatorial bulge has a ring of mass m
r
, so that the two moments of
inertia are 2mr
2
/5 +m
r
r
2
/2 and 2mr
2
/5 +m
r
r
2
. The data given then imply that
m
r
m/375.
We now nd the torque on the ring. Let an element of mass m
r
d/2 be at r and the
Sun be at R, where
r = (r cos , r sin , 0) , R = (Rcos , 0, Rsin ) . (0.11)
The gravitational force on the element is
dF =
GMm
r
d
2
(Rcos r cos , r sin , Rsin )
(R
2
+r
2
2Rr cos cos )
3/2
. (0.12)
The couple is dG = r dF:
dG =
GMm
r
d
2
(Rr sin sin , Rr sin cos , Rr cos sin )
(R
2
+r
2
2Rr cos cos )
3/2
. (0.13)
Integrating around the ring, the G
x
and G
z
components are zero, and the G
y
we could
probably nd in a book if we wanted to. However, here we expand in r/R 1
G
y
=
GMm
r
r sin
2R
2
_
2
0
d cos
_
1 2r cos cos /R +r
2
/R
2
_
3/2

GMm
r
r sin
2R
2
_
2
0
d cos (1 + 3r cos cos /R+. . .)
=
3GMm
r
r
2
4R
3
sin 2 =
1
500
GMm
r
r
2
R
3
sin 2 .
(0.14)
To nd the precession rate look down from the pole of ecliptic plane and see the
projected angular momentum J
p
= (2mr
2
/5 +m
r
r
2
)sin . As always with gyroscope
problems this gives the precession rate as

= G
y
/J
p
= 1.6 10
4
rads yr
1
.
a
q
q

z

3
21. The diagram shows that the angular velocities along
the body axes are (
1
,
2
,
3
) = (0,

sin , +

cos ).
The condition of zero slip at the base is a( +

cos ) = 0,
so we must have
3
= 0.
As usual, look down from above, and see that that the
horizontal component of the angular momentum is
J
x
= J
2
cos = I
1

sin cos and that it precesses at

due to the couple G = mga cos . The precession rate

is =

and satises the gyroscope formula J
x
= G
so that, with I
1
=
1
4
ma
2
we get
2
= 4g/a sin.
The coin is rolling around a circle of radius a cos so does not make a complete
revolution is every precession cycle, The dierence being 1 cos
1
2

2
for small
angle . Using sin , we see that the head of the coin rotates at approximately
_
g
3
/a.
Appendix: more on problem 19
Figure 1: Final of the coin, plotted as a function of the initial . There are regions with 1,
2 and 3 impacts Yes, I know it says Problem 4. . . .
Problem 19 is easy enough, but invites the question as to what happens next. The coin is
left spinning, but still approaching the table, so another impact is inevitable, even without
gravity. Here I give the answer for the number of impacts it makes in the absence of gravity,
and also determine its nal linear and angular velocities as a function of the initial angle .
Coin has mass m and moment of inertia I =
1
4
ma
2
. For any one impact let the initial and
nal velocities and angular velocities be u
1
,
1
and u
2
,
2
, and have vertical impulse P
when the coin hits the table. Momentum equation:
P = m(u
2
u
1
) .
Angular momentum equation:
Pa cos = I (
2

1
) .
Eliminating P gives
ma cos (u
2
u
1
) = I (
2

1
) .
The conservation of energy equation can be written in the form
m
_
u
2
2
u
2
1
_
= I
_

2
2

2
1
_
,
which can be divided by () and rearranged to give
u
2
+a cos
2
= (u
1
+a cos
1
) ,
i.e. the vertical velocity of the point of contact reverses, as expected.
Solving for u
2
,
2
gives answer for the general case:
u
2
=
u
1
(ma
2
cos
2
I) +
1
2aI cos
ma
2
cos
2
+I
;

2
=
u
1
2ma cos
1
(ma
2
cos
2
I)
ma
2
cos
2
+I
.
At this point I resorted to solving the problem numerically, in the absence of gravity, and
for arbitrary moment of inertia. The solution of the impact (given above) is easy, its not
hard to work out whether another impact will happen, but nding the time of the next
impact (and hence the next ) does not have a closed form solution. I used a binary chop,
of course. The plot shows the nal as a function of the initial .
There are regions with 1 , 2 and 3 impacts, and its particularly interesting to see how fast
it spins at the transition between 1 and 2 impacts. The general pattern is as follows: for
near /2, the coin will only make 1 impact, but as the angle is reduced the linear velocity
after the impact decreases and the angular velocity increases. Eventually, the other side of
the coin will make another impact with the table. For I =
1
4
ma
2
the limiting case = 0 has
3 impacts, and the coin is left moving away from the table with a non-zero angular velocity.