General Motion of Rigid Body
General Motion of Rigid Body
General Motion of Rigid Body
Abstract The tools that allow the description of the motion of the rigid body
are recalled. We used both Eulers angles and Eulers parameters (normalized
quaternions) to describe the orientations of the body. Precession of the rigid body
and air resistance and the dissipation of the energy at successive collisions are
discussed.
and
dK
= M,
dt
(2.1)
where F, M are vectors of body external force and moments whereas Q, K are
vectors of linear momentum and angular momentum of body defined by
Q=
v d V
V
and
K=
r v d V .
(2.2)
Strzako, J. et al.: General Motion of a Rigid Body. Lect. Notes Phys. 792, 2339 (2010)
c Springer-Verlag Berlin Heidelberg 2010
DOI 10.1007/978-3-642-03960-7 2
24
(b)
(a)
rO
dV
v
dV
r rO
Fig. 2.1 Body point position (r) and velocity (v) vectors: (a) in the inertial frame x yz, and (b) in
the body embedded frame O
In (2.2) V denotes a body volume (Fig. 2.1), r position vector1 of a body point
(volume d V ), v velocity vector of the point described in inertial frame x yz, and
body material density.
Using the velocity vector vO represented by the components in the body embedded frame ( ) to describe body point velocity vector (v = vO + re ) on the
basis of (2.1), (2.2) we obtain the general form of body dynamics equations:
ROC = F ,
vO + m
T
m v O + mRT
+ m
OC
(2.3)
vO +
JO = MO .
mROC v O + JO
+ mROC
(2.4)
In the mentioned equations m is the mass of body, v O and are column matrices
are antisymmetric matrices representing
of vectors2 v O , components, ROC and
the vectors rOC and , JO is inertia matrix of the body including mass moments of
inertia of the body with respect to O frame. F and MO are column matrices of
vectors F, MO components in O frame.
We get a special form of body dynamics equations introducing the body embedded frame C , where C is the body mass center (Fig. 2.2), and using the mass
center velocity vector (vC = [ vC vC vC ]T ) instead of vO :
vC = F ,
m v C + m
(2.5)
JC = MC .
JC
+
(2.6)
In most of figures the vector quantities are denoted with an arrow above the symbol.
The symbol is used for the angular velocity vector, and for its matrix representation, i.e., the
column matrix of the angular velocity components. For the square matrix representing the angular
is used.
velocity vector () the symbol
2.1
25
F
V
vC
dV
and
dKC
= MC ,
dt
(2.7)
where KC and MC are vectors of angular momentum of a body and body force
moments with respect to the body mass center (C). NewtonEuler equations have
the form:
m v C = F ,
(2.8)
JC = MC ,
JC +
(2.9)
where the mass center velocity and external force are expressed by vector components with respect to the inertial frame (x yz), i.e., vC = [ x y z ]T , F =
, and MC are described in the frame C .
[ f x f y f z ]T , whereas JC , ,
On the other hand, it is more convenient to describe the rotations of a body (2.4)
by their components with respect to the body embedded frame (C ).
Equations (2.8) and (2.9) are general equations of motion and are suitable for the
dynamics analysis of any body, i.e., a nonsymmetric and nonhomogeneous body. In
the case in which the forces F are independent of the angular velocity of the body
and moments MC are not the functions of body mass center accelerations, (2.8) and
(2.9) are uncoupled. This happens when the air resistance of a body is neglected.
The scalar form of a body dynamics equations obtained from the general equation
(2.8) and (2.9) can be written in the well-known form of NewtonEuler equations:
m x = f x ,
m y = f y ,
m z = f z ,
(2.10)
26
J + J J J J + J 2 2 +
+ J J = MC ,
J + J J J J + J 2 2 +
+ J J = MC ,
(2.11)
J + J J J J + J 2 2 +
+ J J = MC .
The moments of inertia in this case are determined with respect to the body embedded frame C with the origin in the body mass center and in the general case, for
a nonsymmetric or nonhomogeneous body, the inertia products are nonzero terms,
because the axes C will not be principal axes.
In the case of symmetric body the axes C are principal axes (J = 0, J =
0, J = 0) and from (2.11) we obtain
J
d
(J J ) = MC ,
dt
d
(J J ) = MC ,
dt
d
(J J ) = MC .
dt
(2.12)
Expressing , , in terms of Euler angles (, , ) or in normalized quaternions (e0 , e1 , e2 , e3 ) one can obtain the motion equations in the chosen coordinates.
2.2
Precession of a Body
27
Fig. 2.3 The precessing nonsymmetric body: changes of angular momentum (KC ) and angular
velocity () vectors direction due to nonzero moments of external forces: (a) for MC = 0, MC =
0, MC = 0 and (b) for MC = 0, MC = 0, MC = 0
3 A symmetric top is a body in which two moments of inertia are the same. There are two classes of
symmetric tops, oblate symmetric top with J = J < J and prolate symmetric top (cigar shaped)
with J = J > J . (A body is an asymmetric top if all three moments of inertia are different.)
4 A spherical top is a special case of a symmetric top (although it need not be spherical) with equal
moment of inertia about all three axes (J = J = J ).
28
J (J1 J3 ) = 0 ,
J1 (J3 J1 ) = 0 ,
(2.13)
J3 = 0 .
Solving (2.13) together with initial conditions (i.e., the initial angular velocity components (0) = 0 , (0) = 0 , (0) = 0 ) we obtain the components of
angular velocity of the body in the form
(t) = 0 sin
(t) = 0 cos
J1 J3
0
J1
3
t + 0 cos J1 J
t
,
0
J1
J1 J3
0
J1
3
t 0 sin J1 J
0 t ,
J1
(2.14)
(t) = 0 .
For symmetric top the vector has constant value
=
2
20 + 0
+ 20 .
(2.15)
From (2.14) arise that (t) is time independent and two other components can be
transformed to the following form:
(t) = sin p t + ,
(2.16)
(t) = cos p t + ,
with the constant coefficients
p =
J1 J3
0 ,
J1
2
20 + 0
,
= arctg
0
.
0
(2.17)
From presented solutions (2.14), (2.15), (2.16), and (2.17) it follows that for the
symmetric top its angular momentum vector (KC ) has components obtained as
K (t) = J1 sin p t + ,
K (t) = J1 cos p t + ,
(2.18)
K (t) = J3 0 ,
and the constant value as
KC =
2
J12 (20 + 0
) + J32 20 .
(2.19)
2.2
Precession of a Body
29
We can also obtain the angle between the symmetry axis of the body ( ) and
the angular momentum vector (KC ) and the angle between and the vector (see
Fig. 2.4) as
= arccos
J3 0
= const ,
KC
= arccos
0
= const .
(2.20)
The angles and have constant value, but they are measured from axis which
moves (changes its direction). While the angular momentum vector KC has constant
value and direction in the spatial frame x yz we can use the angle = to
obtain the orientation of the vector with respect to the fixed vector KC . From
(2.20) we have
= = arcsin
(J3 J1 ) 0
= const .
KC
(2.21)
The vectors of angular velocity and angular momentum for symmetric top are
also presented in Fig. 2.5, where and KC are shown in the spatial frame x yz as
well as in the body embedded frame . The angular velocity vector in the
spatial frame x yz rotates around the angular momentum vector (Fig. 2.5a). Both
vectors rotate in the body embedded frame (Fig. 2.5b).
J1 = 0 ,
J1 = 0 ,
(2.22)
30
(a)
(b)
Fig. 2.6 Torque-free motion of spherical tops: constant angular momentum (KC ) and angular
velocity () vectors directions and values for a ball, tetrahedron die, and cubic die (no precession
motion)
(2.23)
where
r the column matrix representing the vector r of the position coordinates of
an arbitrary point A (Fig. 2.7) of the body before its rotation (in the initial
position),
r
the column matrix of coordinates of the same point of the body after its
rotation (in the final position),
R the rotation matrix (representing the orientation of the local frame ,
with respect to the frame x
y
z
).
2.3
31
z
(a)
(b)
z=
E
I
A
y=
z
x
D
F
x=
H
D
Fig. 2.7 Point A position vectors: (a) r in the initial position and (b) r
in the final position
(2.24)
cos <)(x
, ) cos <)(x
, ) cos <)(x
, )
R = cos <)(y
, ) cos <)(y
, ) cos <)(y
, ) .
cos <)(z
, ) cos <)(z
, ) cos <)(z
, )
(2.25)
On the basis of the rotation matrix (R) it is possible to define the matrix (or angular velocity tensor) that contains the components of the body angular velocity vector
T ), which is necessary in the dynamic analysis of the body. Matrix
is
= RR
(
obtained in the x yz reference frame. We use the symbol for the body angular
velocity matrix defined by the components in the body embedded frame . This
= RT R.
matrix is expressed as
32
R = R1 (1 )R2 (2 )R3 (3 ) ,
(2.26)
1
0
R (i ) = 0 cos i
0 sin i
0
sin i .
cos i
(2.27)
0 sin i
1
0 .
0 cos i
(2.28)
cos i
R (i ) = 0
sin i
(a)
(b)
z=
z
1
C
y=
x=
y
y
x=
x
(c)
(d)
C
x
2
y
x
x
y
3
2.3
33
cos i
R (i ) = sin i
0
0
0 .
1
sin i
cos i
0
(2.29)
Depending on the order of rotations around the sequence of chosen axes there
are 12 possible variants of this method determining the position of a body in space.
Denoting the axes of the body embedded frame by symbols 1, 2, 3 ( 1,
2, 3), possible rotation sequences can be represented as follows: 121 ( -- ),
123 ( -- ), 131 ( - - ), 132 ( - -), 212 (- -), 213 (- - ), 231 (- - ), 232
(- -), 312 ( - -), 313 ( - - ), 321 ( -- ), 323 ( -- ).
Classical Euler angles are: 1 = , 2 = , 3 = , that indicate consecutively
the rotations: by angle 1 = around the axis, by 2 = around the new position
of the axis, and by 3 = around the (rotation sequence is abbreviated as - -
or 313). The rotation matrix R for such angles is obtained by substituting into (2.26)
the following rotational matrices:
R1 (1 ) = R () ,
R2 (2 ) = R () ,
R3 (3 ) = R () ,
(2.30)
that means
R = R ()R ()R () .
(2.31)
In Table 2.1 the set of 12 possible rotation sequences and the singularity condition
for each case and values of angles causing singularities in numerical solutions are
presented. These singularities arise in the inversion process of the matrix B, which
is used to calculate the generalized velocities q on the basis of body angular velocity
(q = B1 ).
Table 2.1 Singularity condition in the numerical analysis of the rigid body dynamics (k =
0, 1, . . .)
Rotation sequence
Singularity condition
121
123
131
132
212
213
231
232
312
313
321
323
sin 2 = 0
cos 2 = 0
sin 2 = 0
cos 2 = 0
sin 2 = 0
cos 2 = 0
cos 2 = 0
sin 2 = 0
cos 2 = 0
sin 2 = 0
cos 2 = 0
sin 2 = 0
2
2
2
2
2
2
2
2
2
2
2
2
= k
= 2 k
= k
= 2 k
= k
= 2 k
= 2 k
= k
= 2 k
= k
= 2 k
= k
34
cos 2
e0
e1 v1 sin 2
p= =
,
e2
v2 sin 2
e3
v3 sin 2
(2.32)
or, shortly
cos 2
e0
p=
= o
.
e
v sin
(2.33)
The basis of this method of body orientation description is well known as the Euler
theorem stating that any rotation of the rigid body can be expressed as a single
rotation about some axis. The axis can be represented by 3D vector vo (Fig. 2.9).
The vector vo is a unit vector and it remains unchanged during the body rotation.
The rotation angle is a scalar value.
The rotation matrix (2.25) can be expressed by Euler parameters as [10]
R = (2e02 1)I + 2eeT + 2e0 E ,
(2.34)
in which I is the identity matrix of dimensions (3 3) and the matrix E has the form
0 e3 e2
E = e3 0 e1 .
e2 e1 0
(a)
(2.35)
(b)
z
z=
y=
x=
y
x
2.4
35
The expanded form of the matrix R (2.34) expressed by unit quaternions (e0 , . . . , e3 )
has the following form:
(2.36)
0 e3 e0 + e2 e1 e 1 e2 + e0 e3 e2 e0 + e3 e1 e0 e2 e1 e3
= 2
0
e1 e0 + e0 e1 + e3 e2 e2 e3 . (2.37)
as ym.
0
The angular velocity vector of the coin in the body embedded frame is
expressed by the column matrix
e1 e0 e0 e1 e3 e2 + e2 e3
= = 2 e2 e0 + e3 e1 e0 e2 e1 e3 ,
e3 e0 e2 e1 + e1 e2 e0 e3
(2.38)
e1 e 0 + e0 e 1 + e3 e 2 e2 e 3
= = 2 e2 e 0 e3 e 1 + e0 e 2 + e1 e 3 .
e3 e 0 + e2 e 1 e1 e 2 + e0 e 3 .
(2.39)
whereas
The column matrix containing x yz scalar components of the coin angular velocity vector (i.e., the components in fixed spatial frame) has the form
e1 e0 e0 e1 + e3 e2 e2 e3
x
= y = 2 e2 e0 e3 e1 e0 e2 + e1 e3 .
e3 e0 + e2 e1 e1 e2 e0 e3
z
(2.40)
36
The determination of the air resistance forces acting on the body is cumbersome
due to its variable velocity and a change in the Reynolds number (Re = 2rv ) that
follows during the toss [13]. (A range of changes in the Reynolds number is wide
because of variations in the body velocity: Re << 1 in the initial stage of its motion
(during its fall with a zero initial velocity), Re
= 104 at the velocity of the mass
center v = 10 m/s and the coin radius r = 0.01 m.)
For detailed information on the aerodynamics and fluid mechanics pertinent to
this section, see [1317].
On the basis of kinematic relations for a rigid body, a distribution of velocities on
outer surfaces of the body is determined. It has been assumed that the air resistance
occurs only on this part of the body surface on which the velocity vectors of its
points have a sense compliant with the normal vector to this surface and directed
outward from the body. The air resistance force vector fr will be determined as a
sum of resistance forces (fi ) on all planes of a body
fr =
ns
fi .
(2.41)
i=1
(2.42)
(2.43)
Si
fin = n
|vAin |b vAin si d Si .
(2.44)
Si
1
T
T
sign(v Ain
io ) 1 + (v Ain
io )
2
(i = 1, 2, 3) ,
(2.45)
2.5
37
where oi are unit vectors, normal to the body surfaces under consideration, directed
outward from the body.
The functions si allow for the determination of regions in which the air resistance
components normal to the surface act. It has been assumed that normal components
of resistance forces do not occur in these regions where the velocity vector of points
situated on the surface is directed inward the body. The air resistance forces perpendicular to the body surface occur in these body points in which the following
condition is satisfied:
v TAin oi > 0
(i = 1, 2, 3) .
(2.46)
ns
mi ,
(2.47)
i=1
(2.48)
Si
RBAi in (2.48) denotes antisymmetrical matrix including the components of the vector5 rBAi .
v Az
,
v Az
(2.49)
where is the coefficient of restitution (0 < < 1), A stands for the body point
which makes contact with the floor at the instant of impact (Fig. 2.10), v Az and
5
In many textbooks for example, in [10] following notation is used RBAi = rBAi .
38
C
y
z
vAz
A
vA
vAy
y
vA x
x
v Az are projections of the velocity of the point A on the direction (z) normal to the
impact surface, respectively, before and after the impact.
Assumption (2.49) means that we include energy losses during the collision and
only the momentum is conserved. (In a purely elastic collision of bodies ( = 1)
kinetic energy, linear and angular momentum are conserved.) On the basis of Newton hypothesis we can calculate z component of the reaction force impulse (Sz ).
Besides Newtons hypothesis to analyze the phenomena that accompany the
impact the laws of linear momentum and angular momentum theorems of a body,
as well as supplementary constraint equations, have been employed.
The position of the colliding point ( A) in the body embedded frame is described
by A , A , A . Assuming that the collision time is negligibly short the position and
the orientation of body after the impact are the same as before the impact.
To describe the motion after the impacts we will use the additional frame with an
origin at point A and axis: z
parallel to the fixed axis z and axes , in the ground
plane (Fig. 2.10).
Mosekilde [18] and Feldberg et al. [19] proposed to obtain the value of considering free vibrations of dumped oscillator described by equation m z + z + kz = 0.
It means that during the collision the body vibrates on visco-elastic foundation (
and k are foundation damping and elasticity coefficients). The illustration for such
a model is shown in Fig. 2.11 and sample results are presented in Fig. 2.12.
Comparing
the oscillations velocity vz for t0 = 0 and t1 = T /2 (where T =
4/
4km 2
)
m2
References
Fig. 2.12 Displacement and
velocity of dumped oscillator
39
z
vz (T/2)
0.2
0
0.02
0.1
0.05
0.2
z(t)*102
vz(0)
2
vz (T /2)
= e 4km 2 .
vz (0)
vz (t)
(2.50)
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
Arnold, V.I.: Mathematical Methods of Classical Mechanics. Springer, New York (1989)
Beatty, M.F., Jr.: Principles of Engineering Mechanics. Plenum Press, New York (1986)
Landau, L., Lifschitz, E.: Mechanics. Pergamon Press, Oxford (1976)
Kittel, C., Knight, W.D., Ruderman, M.A.: Mechanica. Berkeley Physics Course Volume 1.
McGraw-Hill, New York (1965)
Roth, E.J.: Dynamics of a System of Rigid Bodies. Dover, New York (1960)
Wittaker, E.T.: Analytical Dynamics of Particles and Rigid Bodies. Cambridge University
Press, Cambridge (1937)
Goldstein, H.: Classical Mechanics. Addison-Wesley, Reading (1950)
Mongomery, R.: How much does a rigid body rotate? Amer. J. Phys. 59, 394 (1991)
Diebel, J.: Representing Attitude: Euler Angles, Unit Quaternions, and Rotation Vectors.
http://ai.stanford.edu/diebel/. Cited 15 May 2007
Nikrevesh, P.E.: Computer-Aided Analysis of Mechanical Systems. Prentice-Hall, Englewood
Cliffs, NJ (1988)
Marsden, J.E., Ratiu, T.S.: Introduction to Mechanics and Symmetry. Springer, New York
(1994)
Altmann, S.L.: Rotations, Quaternions, and Double Groups. Oxford University Press, Oxford
(1986)
Long, L., Weiss, H.: The velocity dependence of aerodynamic drag: A primer for mathematicians. Amer. Math. Monthly 106, 127 (1999)
Landau, L., Lifschitz E.: Fluid Mechanics. Butterworth and Heinemann, New York (1987)
McCormick, B.: Aerodynamics, Aeronautics and Flight Mechanics. Wiley, London (1995)
Thompson, P.A.: Compressible Fluid Dynamics. McGraw-Hill, London (1972)
White, F.M.: Viscous Fluid Flows. McGraw-Hill, London (1991)
Mosekilde, E.: Topics in Nonlinear Dynamics: Applications to Physics, Biology and Economic Systems. World Scientific, Singapore (1996)
Feldberg, R., Szymkat, M., Knudsen, C., Mosekilde, E.: Iterated-map approach to die tossing.
Phys. Rev. A 42/8, 4493 (1990)
http://www.springer.com/978-3-642-03959-1