322 CHAPTER 10 Dynamics of Rotational Motion

Figure 10.32 A gyroscope supported at 10.7 GYROSCOPES AND PRECESSION

one end. The horizontal circular motion of
the flywheel and axis is called precession. In all the situations we’ve looked at so far in this chapter, the axis of rotation either has
The angular speed of precession is Ω. stayed fixed or has moved and kept the same direction (such as rolling without slipping).
Circular motion But a variety of new physical phenomena, some quite unexpected, can occur when the axis
of flywheel axis Ω
Flywheel of rotation changes direction. For example, consider a toy gyroscope that’s supported at one
(precession)
Flywheel end (Fig. 10.32). If we hold it with the flywheel axis horizontal and let go, the free end of
axis the axis simply drops owing to gravity—if the flywheel isn’t spinning. But if the flywheel
is spinning, what happens is quite different. One possible motion is a steady circular mo-
Pivot tion of the axis in a horizontal plane, combined with the spin motion of the flywheel about
Path followed
the axis. This surprising, nonintuitive motion of the axis is called precession. Precession
by free end of v Rotation of is found in nature as well as in rotating machines such as gyroscopes. As you read these
axis flywheel words, the earth itself is precessing; its spin axis (through the north and south poles) slowly
When the flywheel and its axis are stationary, changes direction, going through a complete cycle of precession every 26,000 years.
they will fall to the table surface. When the To study this strange phenomenon of precession, we must remember that angular ve-

g T that acts on an object Sand the rate of

flywheel spins, it and its axis “float” in the air
locity, angular momentum, and torque are all vector quantities. In particular, we need
while moving in a circle about the pivot.

change of the object’s angular momentum L, given by Eq. (10.29), g T = dL>dt. Let’s first
S
the general relationship between the net torqueS S

apply this equation to the case in which the flywheel is not spinning (Fig. 10.33a). We
take the origin O at the pivot and assume that the flywheel is symmetrical, with mass M
Figure 10.33 (a) If the flywheel in Fig. 10.32 and moment of inertia I about the flywheel axis. The flywheel axis is initially along the x-
S
is initially not spinning, its initial angular axis. The only external forces on the gyroscope are the normal force n acting at the pivot
S
momentum is zero. (b) In each successive (assumed to be frictionless) and the weight w of the flywheel that acts at its center of mass,
time
S
interval dt, the torque produces a change a distance r from the pivot. The normal force has zero torque with respect to the pivot, and
S
dL = T dt in the angular momentum. TheS S
the weight has a torque T in the y-direction, Sas shown in Fig. 10.33a. Initially, there isS no
flywheel acquires an angular momentum L
S
in the same direction as T, and the rotation, and the initial angular momentum Li is zero. From Eq. (10.29) the change dL in
flywheel axis falls. angular momentum in a short time interval dt following this is
S S
(a) Nonrotating flywheel falls dL = T dt (10.32)
S
z This change is in the y-direction because T is. As each additional time interval dt elapses,
S
y S S S the angular momentum changes by additional increments dL in the y-direction because
S
n T r:w
the direction of the torque is constant (Fig. 10.33b). The steadily increasing horizontal an-
Pivot gular momentum means that the gyroscope rotates downward faster and faster around the
O x
S
r Axis y-axis until it hits either the stand or the table on which it sits.
Now let’sSsee what happens if the flywheel is spinning initially, so the initial angular
S momentum Li is not zero (Fig. 10.34a). Since the flywheel rotates around its symmetry
w Path of free end S S
axis, Li lies along this axis. But each change in angular momentum dL is perpendicular to
When the flywheel is not rotating, its weight S S S
the flywheel axis because the torque T = r : w is perpendicular to that axis (Fig. 10.34b).
creates a torque around the pivot, causing it
to fall along a circular path until its axis rests
on the table surface.
S
Figure 10.34 (a) The flywheel is spinning initially with angular momentum Li. The forces (not
(b) View from above as flywheel falls shown) are theSsame as those in Fig. 10.33a. (b) Because the initialSangular momentum is not zero,
S
each
S
change dL = T dt in angular momentum is perpendicular to L . As a result, the magnitude of
y L remains the same but its direction changes continuously.
S
S
dL 1L i 02
S (a) Rotating flywheel (b) View from above
dLS
S
dLS Lf When the flywheel is rotating, the Ssystem Now the effect of the torque is to cause
Pivot dL
S starts with an angular momentum Li parallel the angular momentum to precess around
dL x to the flywheel’s axis of rotation. the pivot. The gyroscope circles around
O
its pivot without falling.
Flywheel Rotation of flywheel y S
z dL S
In falling, the flywheel rotates about the S
T Torque due to weight
S
Lf dL S
y dLS
pivot and thus acquires an angular momentum force (as in Fig. 10.33)
S S
L. The direction of L stays constant. S dL
Li dL
S

x S
Li
Initial angular
v momentum due to x
O
rotation of flywheel
 10.7 Gyroscopes and Precession 323

S S
This causes the direction of L to change, but not its magnitude. The changes dL are al-
ways in the horizontal xy-plane, so the angular momentum vector and the flywheel axis
with which it moves are always horizontal. That is, the axis doesn’t fall—it precesses.
If this still seems mystifying to you, think about a ball attached to a string. If the ball is
initially at rest and you pull the string toward you, the ball moves toward you also. But if
the ball is initially moving and you continuously pull the string in a direction perpendicu-
lar to the ball’s motion, the ball moves in a circle around your hand; it does not approach
S
your hand at all. InSthe first case the ball has zero linear momentum p to start with; whenS
S
you apply a force F toward you for a time dt, the ball acquires a momentum d p = F dt,
S
which is also toward you. But if the ball already has linear momentum p, a change in
S S
momentum d p that’sS
perpendicular
S
to p changes the direction of motion, not the speed.
S S
Replace p with L and F with T in this argument, and you’ll see that precession is simply
the rotational analog of uniform circular motion. S
At the instant shown in Fig. 10.34a, the gyroscopeS
has
S
angular momentum L. A short
time interval dt later, the
S
angular momentum is L dL; theS infinitesimal change in an-
S
gular momentum is dL = T dt, which is perpendicular to L. As the vector diagram in
Fig. 10.35 shows, this means that theS flywheel
S
axis of the gyroscope has turned through
a small angle df given by df = 0 dL 0 > 0 L 0 . The rate at which the axis moves, df>dt, is
called the precession angular speed; denoting this quantity by Ω, we find

S S
df 0 dL 0 > 0 L 0 tz wr
Ω = = = = (10.33)
dt dt Lz Iv

Thus the precession angular speed is inversely proportional to the angular speed of spin
about the axis. A rapidly spinning gyroscope precesses slowly; if friction in its bearings
causes the flywheel to slow down, the precession angular speed increases! The precession
angular speed of the earth is very slow 11 rev>26,000 yr2 because its spin angular momen-
tum L z is large and the torque tz , due to the gravitational influences of the moon and sun,
is relatively small.
As a gyroscope precesses, its center of mass moves in a circle with radius r in a hori-
S
zontal plane. Its vertical component of acceleration is zero, so the upward normal force n
exerted by the pivot must be just equal in magnitude to the weight.
S
The circular motion
of the center of mass with angular speed Ω requires a force F directed toward the center
of the circle, with magnitude F = M Ω 2r. This force must also be supplied by the pivot.
One key assumptionS that we made in our analysis of the gyroscope was that the angu-
lar momentum vector L is associated with only the spin of the flywheel and is purely hori-
zontal. But there will also be a vertical component of angular momentum associated with
the precessional motion of the gyroscope. By ignoring this, we’ve tacitly assumed that the
precession is slow—that is, that the precession angular speed Ω is very much less than
the spin angular speed v. As Eq. (10.33) shows, a large value of v automatically gives a
small value of Ω, so this approximation is reasonable. When the precession is not slow,
additional effects show up, including an up-and-down wobble or nutation of the flywheel
axis that’s superimposed on the precessional motion. You can see nutation occurring in Sa
gyroscope as its spin slows down, so that Ω increases and the vertical component of L
can no longer be ignored.

Figure 10.35 Detailed view of part of In a time dt, the angular momentum
Fig. 10.34b. vector and the flywheel axis (to which
it is parallel) precess together through
an angle df.
y

S S
L dL S
dL
df
x
O S
L