Professional Documents
Culture Documents
Inertial Rotation of A Rigid Body
Inertial Rotation of A Rigid Body
Eugene Butikov
St. Petersburg State University, St. Petersburg, Russia
E-mail: [email protected]
Introduction
Any motion of a body occurring in the absence of external forces can be called a free or in-
ertial motion. Inertial motion is certainly the simplest possible motion. For a mass point the iner-
tial motion is indeed a very simple one – it is a uniform rectilinear motion. However, for a rigid
body only translational free motion, during which the body does not rotate, is actually simple
enough. If the body rotates, its motion can be rather complicated even in the absence of external
forces. The problem of torque-free rotation has many applications including any object freely
falling in a gravitational field, as well as a space station or a non-spherical satellite orbiting the
earth. For an axially symmetrical body, kinematics of such inertial rotation corresponds to pre-
cession. Being applied to a gyroscope, this free precession is called nutation.
The problem of a rigid body rotation has been studied over several centuries, and the equa-
tions governing the behaviour of a body have been well known since Euler. And nowadays most
textbooks in physics and introductory mechanics (see, for example, [1] – [3]) still treat the prob-
lem of torque-free rotation of a rigid body on the basis of Euler equations which are referred to
the non-inertial reference frame rotating together with the body. Euler equations are defined for
projections of the angular velocity vector onto the coordinate axes fixed to the rotating body, and
hence these equations can tell us how the momentary angular velocity behaves itself with respect
to the body. However, in this problem we are interested first of all how the body moves with re-
spect to the inertial space. For students, transition to the inertial frame of reference from a frame
which rotates in a complicated manner may be rather confusing. From a pedagogical point of
view, it seems expedient to find a solution to this problem directly in the inertial space.
We discuss in this paper such a solution for an axially symmetrical body in an attempt to
develop some clarity regarding the problem with the help of an obvious geometrical interpreta-
tion of the investigated motion. We suggest also a simple simulation program (Java-applet) “Free
rotation of an axially symmetrical body” [4] that visualizes the inertial rotation and hence can
facilitate understanding about its principal features. The program is available in the Internet and
is executed directly in any web browser with the Java plug-in installed.
1
have constant directions in space. In this frame of reference an arbitrary motion of a rigid body is
a rotation about a fixed point (about of the centre of mass).
Kinematics of rotation of a rigid body about a fixed point is characterized by a vector ω of
momentary angular velocity. Any point of the rotating body has a (linear) velocity, which at
every moment of time is exactly the same as if the body were rotating around an axis directed
along the angular velocity vector. However, for the general case of free rotation, the vector of
angular velocity and hence the momentary axis of rotation change continuously their direction in
space. Even in the absence of external torques, that is, during the inertial rotation, behaviour of
the momentary axis of rotation (sometimes this behaviour is called “wobbling”) seems to be very
complicated and counterintuitive. Moreover, the trajectories of different points of a freely rotat-
ing body seem even more complicated. The simulation program “Free rotation of an axially
symmetrical body” illuminates various unexpected features of the inertial rotation, showing be-
haviour of the momentary angular velocity and the trajectory of an arbitrarily chosen point of the
body. The program visualizes also the suggested geometrical interpretation of inertial rotation.
For a rotating body, vector L of the total angular momentum is proportional to the momen-
tary angular velocity ω, but generally deviates from ω in direction. Both vectors L and ω have a
common direction only if the body rotates about one of the three mutually orthogonal axes called
principal axes of inertia. Principal axes of inertia exist for every body. For symmetrical bodies
manufactured of a homogeneous material, the principal axes of inertia coincide with the axes of
symmetry. For example, the principal axes of inertia of a rectangular parallelepiped pass through
its centre parallel to the edges. Moments of inertia, calculated with respect to the principal axes
of inertia that pass through the centre of mass, are called central principal moments of inertia.
Inertial rotation of a rigid body about one of the principal axes of inertia is very simple. In-
deed, during this rotation directions of vectors L and ω coincide, and since in the absence of ex-
ternal torques vector L of the angular momentum is conserved, vector ω of the angular velocity
is also conserved, i.e., its magnitude and direction in space remain constant. For this reason the
principal axes of inertia are otherwise called the axes of free rotation. If a rigid body is set spin-
ning about one of these axes and then released, the body simply continues spinning uniformly
around the axis whose direction in space and in the body does not change. All points of the body
synchronously trace circles whose centres are located on this axis.
When such bodies are set into rotation about the axis of symmetry, vector L of the angular
momentum is also directed along this axis. If vector ω of the angular velocity deviates from the
axis of symmetry through some angle, directions of vectors L and ω do not coincide anymore,
but vector L necessarily lies in the same plane with ω and the axis of symmetry. Mutual disposi-
tion of these vectors for an axially symmetrical body is shown in figure 1.
2
For a body whose central principal moment of inertia about a transverse axis is greater than
the moment of inertia about the axis of symmetry, vector L of the angular momentum deviates
from the axis of symmetry through a greater angle than vector ω of the angular velocity does.
Such mutual disposition of vectors L and ω with respect to the axis of the body is characteristic
of prolate, stretched bodies (left-hand side of figure 1). For an oblate body, squeezed along the
axis, vector L of the angular momentum deviates from the axis of symmetry through a smaller
angle than vector ω does (see right-hand side of figure 1).
z z
ω L
n n
L ω
y y
x x
Figure 1. Mutual disposition of vector ω of the angular velocity, vector L of the angular momen-
tum, and the axis of symmetry (vector n) for symmetrical bodies of prolate (left) and oblate
(right) shapes.
Next we introduce a unit vector n to show in space the direction of the symmetry axis of
our body, i.e., n is the vector starting at the origin (at the centre of mass) and directed along the
axis of the body. At any moment of time all the three vectors n, L, and ω lie in one plane, and
their mutual disposition doesn’t change during the motion of the body. The plane that contains
the three vectors n, L, and ω, is revolving uniformly in space about vector L of the angular mo-
mentum whose direction in space remains constant in the absence of external torques. Indeed,
linear velocity v = dn/dt of the end point of vector n, in terms of the momentary angular velocity
ω and vector n itself, is given by the formula dn/dt = ω × n. This means that at any time moment
the end point of vector n is moving perpendicularly to the plane of vectors n and ω, involving
this plane together with these vectors into a uniform rotation about L. During this rotation of the
plane, vector L remains constant both in direction and magnitude, while vectors n and ω move
synchronously preserving their magnitudes: they describe in space circular cones with the com-
mon axis directed along L. The steady motion described above is commonly called regular pre-
cession of vectors n and ω about L. The motion of the body itself, which consists of spinning
about the own axis with simultaneous precession of this axis (revolution of the axis about L) is
also called a uniform or regular precession.
Regular precession of vectors n and ω about L can be characterized by vector Ω of the an-
gular velocity of precession. Angular velocity of precession is directed along and proportional in
magnitude to the total angular momentum L, and inversely proportional to the transverse mo-
ment of inertia I⊥ (moment of inertia about an axis, perpendicular to the axis of symmetry):
Ω = L/I⊥. This will be proved below (see section Angular velocity of precession). With the help
of Ω, for the time rate of vector n variation in space, alongside the expression dn/dt = ω × n, we
can write also dn/dt = Ω × n. Similar expression dω/dt = Ω × ω is valid for the time rate of vec-
3
tor ω variation – both vectors n and ω synchronously execute uniform precession about a con-
stant vector L with the same angular velocity Ω.
When such torque-free regular precession occurs with a gyroscope and superimposes on its
forced precession, it is usually called nutation. We note again that a torque-free precession of an
object occurs if the angular velocity deviates from the axis of symmetry. The axis of a freely ro-
tating symmetrical body preserves its direction in space (does not precess) if the angular velocity
is directed along the axis of symmetry, i.e., if the body is spinning about its axis of symmetry.
z
z Ω
ω
L
ω
Ω n
ω0
y
x
x
ω0
4
momentary axis of rotation. In other words, the lateral surface of this cone associated with the
body is also formed by the set of momentary axes of rotation at different moments of time, but,
contrary to the space cone, this “body cone” shows how all these axes are located inside the body
(relative to the body) at different moments of time. For this reason, this imaginary cone, associ-
ated with the rotating rigid body, is called a moving axoid.
The moving and immovable cones touch one another by their lateral surfaces (outwardly
for a prolate body, whose transverse moment of inertia is greater than longitudinal, as shown in
the left-hand part of figure 2) along the momentary axis of rotation ω. All the points of the body,
which are located at a given moment on the momentary axis of rotation, have zero linear veloci-
ties. This means that the moving cone (attached to the body) is rolling without slipping over the
surface of the immovable cone, the line of contact being ω. A complicated motion of an arbitrary
point of the body corresponds to addition of two rather simple motions of the body superimposed
upon one another, namely, a uniform rotation about its axis of symmetry with angular velocity
ω0 and a uniform precession of this axis in space along a cone with angular velocity Ω. This
geometrical interpretation of kinematics of the torque-free precession (of the inertial rotation) is
clearly shown in figure 2.
We can also compare this evident geometrical interpretation of the torque-free precession
(as rolling without slipping of the body cone over the surface of the space cone) with the decom-
position of vector ω of the total momentary angular velocity onto the vector sum of two compo-
nents ω0 and Ω, corresponding to the individual rotations (see figure 2):
ω = ω0 + Ω.
One of these components ω0 corresponds to the axial rotation (spinning) of the body about
its axis of symmetry. This component of the angular velocity is directed along the axis of sym-
metry, that is, its direction inside the body does not change during the motion. In space, this
component generates a circular cone together with the axis of the body, changing with time ac-
cording to the expression dω0/dt = Ω × ω0. The other component Ω, on the contrary, does not
change its direction in space: this component corresponds to precession of the body axis of
symmetry about angular momentum L, whose direction in space (vertical on the screen) is pre-
served in the absence of external torques. This uniform precession occurs simultaneously with
the axial rotation of the body.
5
L
ω
n
Ω
L
ω ω0
Figure 3. Relationship between vector L of the total angular momentum and vector Ω of the an-
gular velocity of precession.
We note that projections of both ω and Ω onto the perpendicular axis are equal to ω⊥.
Considering similar right-angle triangles with a common angle, whose hypotenuses are formed
by vectors L and Ω, while legs by L⊥ and ω⊥ respectively (see figure 3), we can write a propor-
tion L/L⊥ = Ω/ω⊥. Taking into account that L⊥ = I⊥ω⊥, we get from this proportion the desirable
expression for angular velocity Ω of precession: Ω = L/I⊥. For small deviations of ω from n
(when ω⊥ << ω) this exact expression yields for the angular velocity of precession a simple ap-
proximate relationship, according to which Ω equals ω times the ratio of longitudinal and trans-
verse moments of inertia. This means that for prolate bodies Ω < ω (precession goes slower than
the axial rotation), while for oblate bodies Ω > ω. In particular, for a thin disc the angular veloc-
ity of precession (of “wobbling”) is twice as great as the angular velocity of revolution: Ω ≈ 2ω.
Figure 4. Simulation of the inertial rotation of an elongated symmetrical body (left) and the geo-
metrical interpretation of this precession as rolling without slipping of the body cone over the
surface of the space cone.
6
The simulation shows also vectors ω and ω0, precessing with angular velocity Ω about the
fixed in space direction of the angular momentum. Choosing different values for the ratio of
transverse and longitudinal moments of inertia (in the program this parameter is called “Prolate-
ness”), and different deviations of the total angular velocity ω from the body axis (parameter
“Angle”), we can see clearly how vectors ω0 and Ω add in various cases to form vector ω of
momentary angular velocity.
For an oblate axially symmetrical body whose longitudinal moment of inertia is greater
than the transverse one, the torque-free precession may occur even more surprising. In this situa-
tion vectors n and ω at any moment of time are deviated from vector L of the angular momentum
to the opposite sides, as we can conclude from the right-hand side of figure 1. In this case the
angle between vectors ω0 and Ω is obtuse.
In other words, vector ω0 of the spin angular velocity is directed oppositely to vector n, in
contrast to the case of a prolate body. This means that when the axis of the body is precessing
counterclockwise and the body cone is rolling inwardly over the space cone (touching the space
cone by its inner surface), the own rotation of the body occurs clockwise, that is, in the opposite
sense with respect to precession. This is a retrograde kind of precession. Figure 4 gives an im-
pression how the program simulates this behaviour.
Figure 5. Computer simulation of the torque-free precession of an oblate symmetrical body (left-
hand side) and the geometrical interpretation of this retrograde precession (right-hand side).
Long ago this kind of precession surprised the famous physicist Richard Feynman and
stimulated him to develop a deep insight into the phenomenon. He tells the following story [5]:
“I was in the cafeteria and some guy, fooling around, throws a plate in the air. As the plate went
up in the air I saw it wobble, and I noticed the red medallion of Cornell on the plate going
around. It was pretty obvious to me that the medallion went around faster than the wobbling. I
had nothing to do, so I start figuring out the motion of the rotating plate. I discovered that when
the angle is very slight, the medallion rotates twice as fast as the wobble rate – two to one. It
came out of a complicated equation! I went on to work out equations for wobbles. Then I
thought about how the electron orbits start to move in relativity. Then there’s the Dirac equation
in electrodynamics. And then quantum electrodynamics. And before I knew it ... the whole busi-
ness that I got the Nobel prize for came from that piddling around with the wobbling plate.”
7
As we already mentioned, for a thin disc the angular velocity of precession is approxi-
mately twice as great as the angular velocity of revolution: Ω ≈ 2ω. Also our simulation shows
clearly that at slight angles for a plate (i.e., for a thin disc) vector ω0 of the spin angular velocity
and vector ω of the momentary angular velocity are very nearly equal in magnitude (though di-
rected almost oppositely), while the precession (or “wobble”) rate Ω is twice as fast as the spin
ω0 of the plate. It’s just the opposite to Feynman’s description of the wobbling plate!
Points of the body located on its axis of symmetry trace circular paths whose centres lie on
the axis of the immovable cone, that is, on the axis of precession (vertical on the screen). We can
consider complicated motion of points that doesn't lie on the axis of symmetry as a superposition
of two relatively simple motions, namely of a rotation about the axis of symmetry with simulta-
neous motion of this axis in space (precession) along the surface of a vertical cone. The program,
alongside the simulation of rolling without slipping of a body cone over the space cone, can
show trajectories traced by such points in the course of inertial rotation. First we choose position
of a point by indicating the angle between the axis of the body and direction to the point in ques-
tion. For more convenient observation, the program shows the trajectory traced by the spike of
an imaginary long arrow fixed to the body. The spike of this arrow lies beyond the surface
bounding the body. All points of this arrow (starting at the centre of mass) trace similar trajecto-
ries, but the path of its end point shows their peculiarities in a larger scale.
In particular, we can choose a point on the lateral surface of a body cone, which at the ini-
tial moment of the simulation lies on the momentary axis of rotation. To do this, for the direction
to the point we should choose the angle, equal to deviation of the angular velocity from the body
axis. This point traces on a spherical surface a trajectory which looks like a cycloid, whose con-
secutive arcs join with one another like a sharp beak with a common tangent. For a prolate body,
if we choose a point located to the body axis closer than the surface of the body cone (i.e., a
point inside the body cone), the point will trace a wavy (sine-like) trajectory. Points of the body
located outside the body cone trace trajectories with loops. For an oblate body, on the contrary,
these points trace wavy trajectories, and vice versa.
An intermediate position between the cases of prolate and oblate bodies corresponds to the
spherical top – a body, whose all three central principal moments of inertia are equal. The shape
of a spherical top shouldn’t necessarily be spherical. For example, a cube manufactured of a ho-
mogeneous material is characterized by equal principal moments of inertia about the axes paral-
lel to its edges. This means that with respect to rotation the cube is dynamically equivalent to a
spherical ball. All other regular polyhedrons (tetrahedron, octahedron, dodecahedron, icosahe-
dron) are also spherical tops – in a torque-free rotation about the centre of mass they all behave
themselves in the same way.
For a spherical top, directions of the principal axes of inertia can be chosen arbitrarily –
any three mutually orthogonal axes (with the origin at the centre of mass) can be regarded as
principal axes. In particular, for a cube these axes should not be necessarily parallel to its edges.
This means that any axis passing through the centre is an axis of free rotation. In other words, for
an arbitrary direction of vector ω of the angular velocity, vector L of the angular momentum will
have the same direction: for a spherical top an inertial torque-free rotation about any axis is sim-
ply a uniform rotation, and the direction of this axis of rotation in space is constant.
The described above geometrical interpretation of inertial rotation of an axially symmetri-
cal body is certainly applicable to the special case in which the longitudinal and transverse mo-
ments of inertia are equal, e.g., to the case of a spherical top. During a free rotation of a spherical
top vector ω of the angular velocity and the momentary axis of rotation are directed along L and
preserve their direction in space – they do not precess. This means that the space cone degener-
ates into a ray (a half-line) directed along vector L of the total angular momentum. Illustration of
this behaviour of the spherical top in the simulation program is shown in figure 6. The axis of the
body cone is the axis which we have chosen (arbitrarily) in the body as its axis of symmetry.
8
(For a cube this axis can be directed, say, in parallel to one of its edges, or along one of its space
diagonals, or can have any other fixed direction.) Rolling of the body cone over the degenerated
space cone reduces to a uniform revolution of this body cone about vector ω that passes from the
vertex of the cone (centre of mass of the body) along its lateral surface. Any point of the body
(e.g., the spike of an arrow fixed to the body, see figure 6) traces a circle with the centre located
on the axis of rotation.
Concluding remarks
In this paper we have considered the important old problem of torque-free rotation of a
symmetrical rigid body. Our treatment is appropriate for teaching introductory mechanics and
general physics to undergraduate students. A small simulation program (Java-applet) [4] is de-
veloped to visualize the investigated motion. Simultaneously with simulating the body motion,
the program presents a clear geometrical interpretation of the inertial rotation. The suggested ap-
proach is free from difficulties inherent to traditional treatment of the problem which has many
applications including objects falling in a gravitational field, non-spherical satellites orbiting the
earth, etc. An understanding about the inertial (torque-free) rotation of an axially symmetrical
body is also an important prerequisite for the study of a counterintuitive behaviour of a gyro-
scope, whose torque-induced precession is generally complicated by a nutation [6].
References
1. Kittel Ch, Knight W D, Ruderman M A 1965 – 1971 Mechanics. Berkeley Physics
Course (New York: McGraw-Hill).
2. Goldstein H 1980 Classical Mechanics. 2nd ed (Addison – Wesley).
3. Goldstein H, Poole Ch P, Safko J L 2002 Classical Mechanics. 3rd ed (Addison –
Wesley).
9
4. Butikov E 2006 Free Rotation of an Axially Symmetrical Body
http://www.ifmo.ru/butikov/Applets/Precession.html
5. Feynman R P 1985 Surely You Are Joking, Mr. Feynman! (W W Norton). See pp. 157 –
158 for a discussion of the rotating plate motion.
6. Butikov E 2006 Forced Precession of a Gyroscope
http://www.ifmo.ru/butikov/Applets/Gyroscope.html
10