JIMMA UNIVERSITY

JIMMA INSTITUTE OF TECHNOLOGY

Post Graduate Study

School Of Mechanical Engineering

M.Sc. In Design of Mechanical Systems

Course name: Dynamics and Control of Mechanical Systems

Course code: DMS6021

Review on:

A review of general rigid body dynamics simulation with examples

By: Fakada Dabalo

Id. No.: RM0969/09

Submitted to: Prof. Hirpa Lamu

Oct, 7/2017, Jimma, Ethiopia

 ACKNOWLEDGEMENT
First of all foremost, my gratitude goes to Professor, Hirpa Lamu for delivering me this
vital course, Dynamics & Control of Mechanical Systems (DMS6021) and sharing his rich
experience and knowledge without tired through all contents of cource. His encouragement,
lecturing excellent, and critical comments have greatly contributed to this review project. By
letting I to work on such interesting topic, he enabled me to have interesting of reading and
working on a broaden my thought to read a lots of available materials in dynamics & control
mechanical systems. Thanks

DMS6021 - Dynamics and Control of Mechanical Systems Page i

 ABSTRACT
The formulae that allow the description of the dynamics motion of the rigid body are recalled.
This paper is used to acquire knowledge about the combined translational and rotational
(general) dynamics of rigid bodies. To describe the orientations of the rigid body the use of both
Euler’s angles and Euler’s parameters (normalized quaternions) are very important. The rigid
body simulation is an important of mechanical system in order to understand the behaviour of the
important part of
motion of the dynamics system and performance of the system. Knowing the basics differential
equation of rigid body dynamics is the core to do the simulation. Simulation gives clue for new
possibilities for changing how designers, engineers, modellers and animators work with their
design problems. In this paper we will not get full general rigid body which has been done, it
also seeks to give the reader deeper insights to help guide their future research.
Keywords: Rigid body, rigid body dynamics, kinematics and kinetics rigid body, simulation.

Good abstract

Try to avoid use of first person in a technical report like this or any scientific publication.

DMS6021 - Dynamics and Control of Mechanical Systems Page ii

Table of Contents
ACKNOWLEDGEMENT ....................................................................................................................................i
ABSTRACT ...................................................................................................................................................... ii
1. INTRODUCTION ........................................................................................................................................ 1
If you do not have section 1.2, why do you need 1.1?
1.1 Background ......................................................................................................................................... 1
2. GENERAL DYNAMICS OF A RIGID BODY.................................................................................................... 2
2.1 Basic Equations of Rigid Bodies Dynamics ......................................................................................... 2
2.1.1 Dynamics Equations in General Form ......................................................................................... 3
2.2 Precession of a Body .......................................................................................................................... 5
2.2.1 Precession of Symmetric Top ...................................................................................................... 6
2.2.2 Torque-Free Motion of Spherical Top ......................................................................................... 7
2.3 Orientation of a Rigid Body ................................................................................................................ 8
2.3.1 Euler Angles and Other Conventions .......................................................................................... 9
2.3.2 Euler’s Parameters .................................................................................................................... 10
2.5 Modelling of Bodies Impact.............................................................................................................. 12
3. RIGID-BODY SIMULATION....................................................................................................................... 13
4. CONCLUSION .......................................................................................................................................... 15
REFERENCES ............................................................................................................................................... 16

LIST OF FIGURES - preferably on a new page, not on the same page as TOCs

Figure 1. Body point position (r) and velocity (v) vectors ........................................................................... 3
Figure 2.Body mass center velocity vector (vC) and external force F in the inertial frame ........................ 4
Figure 3.The precessing non-symmetric body ............................................................................................. 5
Figure 4.Torque-free precession of symmetric top ..................................................................................... 7
Figure 5.The precessing symmetric top……………………………………………………………………………………………………8
Figure 6.Torque-free motion of spherical tops ............................................................................................ 8
Figure 7.Point A position vectors ................................................................................................................. 9
Figure 8.Initial position of a body and rotation.......................................................................................... 10
Figure 9.The body rotation about given angle. .......................................................................................... 11
Figure 10.Velocity vector ........................................................................................................................... 13
Figure 11. Motion simulation for pendulum .............................................................................................. 14

DMS6021 - Dynamics and Control of Mechanical Systems Page iii

 1. INTRODUCTION

1.1 Background
A rigid body defined as a large (infinite) number of particles where all distances between the
particles are fixed i.e. which doesn’t deform. (No real bodies are of course absolutely rigid).
General rigid body dynamics is a part of mechanics and can be divided in to kinematics and
kinetics. Kinematics is concerned with the motion without regard to its causes (forces). Kinetics
is concerned with the interplay between forces and motion.

We need to describe the motion of rigid bodies for two important reasons. First, we frequently
need to generate, transmit, or control certain motions by the use of cams, gears and linkage of
various types. Here we must analyse the displacement, velocity and acceleration of the motion to
determine the design geometry of the mechanical parts. Second, we must often determine the
motion of a rigid body caused by the forces applied to it. Calculation of the motion of a rocket
under the influence of its thrust and gravitational attraction is an example of such a problem.

One free rigid body has six degrees of freedom in space (and only three degrees of freedom in
plane). Consequently the motion of the body will be defined by six scalar independent kinematic
parameters (only three in plane). The six relations can be obtained from the two vector general
theorems: the theorem of the linear momentum and the theorem of the angular momentum.

A rigid body dynamics is undergoing translational and rotational motion about a fixed direction.
Because the body is translating, the axis of rotation is no longer fixed in space. We shall describe
the motion by a translation of the centre of mass and a rotation about the center of mass. By
choosing a reference frame moving with the center of mass, we can analyze the rotational motion
separately and discover that the torque about the center of mass is equal to the change in the
angular momentum about the centre of mass [1], [2]. ok, but better as [1, 2]

The general rigid body dynamics is focussed on many motion of system like absolute motion,
relative motion, angular momentum, inertia matrix and inertia dyadic, principal axes [3], [4],[5],
general plane motion, torque-free motion of an axisymmetric rigid body, general torque-free
motion [6], rigid body in in a circular orbit, equation of motion, linear stability analysis [7].

An analysis of motion ought to begin with the Ancient Greeks [8]. The beginning of a rational
understanding of dynamics is credited to Galileo (1564-1642) who made careful observations
concerning bodies in free fall, motion on an inclined plane, and motion of the pendulum. Newton
(1642-1727), guided by Galileo’s work, was able to make an accurate formulation of the laws of
motion and, thus to place dynamics on a sound basis. Following Newton’s time, important
citations?
contributions to mechanics were made by Euler, D’Alembert, Lagrange, Laplace, Poinsot,
Coriolis, Einstein, and many others. Discussions of dynamics almost invariably begin by citing

DMS6021 - Dynamics and Control of Mechanical Systems Page 1

the work of Newton (1687) in his Principia, and seldom proceed further than this opus. It is as if
classical mechanics arose from the genius of Newton alone [9].The field of dynamics begins
with a brief overview of dynamics’ origins in Ancient Greece, and progresses through the
Middle Ages and the Renaissance, stopping at the times, places, and people that the author
considers most relevant. The dominant figure in the ancient development of dynamics was
Aristotle (384 BC- 322 BC) [10]. The dynamics and control of rigid-body motion permeates
fields of computer graphics, biomechanics, robotics, and numerical analysis. Rigid body
dynamics has a long history in computer graphics. Dynamics is a part of mechanics and can be
divided in to kinematics and kinetics. Kinematics is concerned with the motion without regard to
its causes (forces). Kinetics is concerned with the interplay between forces and motion [11]. For
free motion of rigid body the existence of dynamics and uniqueness for rigid-body systems of
normal-force-dependent friction are included; such observation has shown that the rigid-body
equations remain well behaved. However, starting Painleve, examples of rigid-body systems
with Coulomb friction were published which produced either no solution or multiple solutions to
the forward dynamics problem [12], [13], [14], [15], [16]

A rigid body is a non-naturally occurring volume of particles where each particle within the
volume retains its position during the lifetime of the rigid body. Each of the particles in the given
volume are also position-relative to the centre of mass of the given 3D rigid body. Therefore
mass within a rigid body is evenly distributed. These element properties are used to determine
how much the given rigid body rotates around a given axes, relative to an angular force applied
to it [17], [18], [19]. An inertia tensor for a 3D solid cube volume can be defined in terms of its
bounding width, height and depth. (For a default 3x3 symmetrical inertia tensor model)[19].
The last paragraph of the interoduction section is expected to give statement of the objective of the article/report

2. GENERAL DYNAMICS OF A RIGID BODY

2.1 Basic Equations of Rigid Bodies Dynamics

The basic equation general rigid body dynamics concerned with the tools that allow the
description of the dynamics of rigid body at all (kinematics and kinetics). In this section we deal
with linear and angular momentum.

In addition in the general rigid body motion analysis of a coin, a die, and roulette ball free
(unconstrained) body motion and constraint body motion are analyzed. Free body motion is
described by dynamics equations of a rigid body [20], [21], [22], [23], [24], [25]. At the instant
of the body (coin, die, ball) contact with surroundings (base, table, wheel) some additional
equations constraint equations are necessary.

DMS6021 - Dynamics and Control of Mechanical Systems Page 2

2.1.1 Dynamics Equations in General Form

The general form of body dynamics equations can be derived from body linear momentum and
angular momentum laws
𝑑𝑄 𝑑𝐾
= F and =M (2.1)
𝑑𝑡 𝑑𝑡

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 𝑉𝜌𝑑𝑉 𝑎𝑛𝑑 𝐾 = ∫𝑉 𝑟 × 𝑉𝜌𝑑 v (2.2)

Figure 1. Body point position (r) and velocity (v) vectors: (a) in the inertial frame xyz, and (b) in
the body embedded frame Oξηζ

In (1.2) V denotes a body volume (Fig.1), r⃗ position vector of a body point (volume dV ), v
velocity vector of the point described in inertial frame xyz, and ρ body material density.
Using the velocity vector vO represented by the components in the body embedded frame (ξηζ )
to describe body point velocity vector (v = vO + ω × re) on the basis of (2.1), (2.2) we obtain the
general form of body dynamics equations:
𝑇
𝑚v𝑜̇ + 𝑚𝑅𝑜𝑐 ̃̇ + 𝑚Ω
+ 𝑚𝜔 ̃ 𝑣𝑜 + 𝑚Ω
̃𝑇 Ω
̃ 𝑅𝑜𝑐 = 𝐹 (2.3)

̃̇ + 𝑚𝑅𝑜𝑐 Ω
𝑚𝑅𝑜𝑣 𝑣̇𝑜 + 𝐽𝑜 𝜔 ̃ 𝑣𝑜 + Ω
̃ 𝐽𝑜 𝜔
̃ = 𝑀𝑜 (2.4)

In the mentioned equations m is the mass of body, v𝑂̇ and 𝜔

̃ are column matrices of vectors v𝑂̇ ω
components, ROC and Ω ̃ are anti-symmetric matrices representing the vectors ROC and ω, JO is
inertia matrix of the body including mass moments of inertia of the body with respect to Oξηζ
frame. F and MO are column matrices of vectors F, MO components in Oξηζ frame.
We get a special form of body dynamics equations introducing the body embedded frame Cξηζ ,
where C is the body mass center (Fig.2), and using the mass center velocity vector
(vC = [ VCξ VCη VCζ ]T ) instead of vO:

𝑚𝑣̇𝑐 + 𝑚Ω ̃ 𝑣𝑐 = 𝐹 (2.5)
𝐽𝑐 𝜔
̃ + Ω ̇
̃ 𝐽𝑐 𝜔̃ = 𝑀𝐶 (2.6)
Where, ω- is used for the angular velocity vector and
̃ - For its matrix representation, i.e., the column matrix of the angular velocity
𝜔
components.
̃ - used for the square matrix representing the angular velocity vector (ω).
Ω

DMS6021 - Dynamics and Control of Mechanical Systems Page 3

Figure 2.Body mass center velocity vector (vC) and external force F in the inertial frame xyz,
angular velocity and acceleration vectors in the body embedded frame (Cξηζ
2.1.2 Newton–Euler Equations

The simplest equations describing dynamics of a rigid body are the well-known form of the
Newton–Euler equations based on body linear momentum and angular momentum laws
𝑑𝑄 𝑑𝐾𝐶
𝑑𝑡
= F and 𝑑𝑡
= MC (4.7)

Where KC and MC are vectors of angular momentum of a body and body force moments with
respect to the body mass center (C). Newton–Euler equations have the form:

mv̇ C = F (2.8)

̃̇ + Ω
𝐽𝑐 𝜔 ̃̇ = 𝑀𝑐
̃ 𝐽𝑐 𝜔 (2.9)

where the mass center velocity and external force are expressed by vector components with
respect to the inertial frame (xyz), i.e., VC = [ 𝑋̇ 𝑌̇ 𝑍̇ ]T , F = [ fx fy fz]T, whereas JC, 𝜔 ̃,
̃, Ω
and MC are described in the frame Cξηζ .

On the other hand, it is more convenient to describe the rotations of a body (4.4) by their
components with respect to the body embedded frame (Cξηζ ).

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 (4.8) and (4.9)
can be written in the well-known form of Newton–Euler equations:

m ẍ = 𝑓𝑥 , m ÿ = 𝑓𝑦 , m z̈ = 𝑓𝑧 , (2.10)

𝐽𝜉 𝜔̇ 𝜉 + ( 𝐽𝜁 − 𝐽𝜂 )𝜔𝜂 𝜔𝜁 − 𝐽𝜉𝜁 𝜔̇ 𝜁 − 𝐽𝜉𝜂 𝜔̇ 𝜂 + 𝐽𝜂𝜁 ( 𝜔𝜁2 − 𝜔𝜂2 ) + ( 𝐽𝜉𝜂 𝜔𝜁 − 𝐽𝜉𝜁 𝜔𝜂 )𝜔𝜉 = 𝑀𝐶𝜉 ,

𝐽𝜂 𝜔̇ 𝜂 + ( 𝐽𝜉 − 𝐽𝜁 )𝜔𝜉 𝜔𝜁 − 𝐽𝜉𝜂 𝜔̇ 𝜉 − 𝐽𝜂𝜁 𝜔̇ 𝜁 + 𝐽𝜉𝜁 (𝜔𝜉2 − 𝜔𝜁2 ) + (𝐽𝜂𝜁 𝜔𝜉 − 𝐽𝜉𝜂 𝜔𝜁 )𝜔𝜂 =

𝑀𝐶𝜂 , (2.11)

DMS6021 - Dynamics and Control of Mechanical Systems Page 4

 𝐽𝜁 𝜔̇ 𝜁 + (𝐽𝜂 − 𝐽𝜉 )𝜔𝜉 𝜔𝜂 − 𝐽𝜂𝜁 𝜔̇ 𝜂 − 𝐽𝜉𝜁 𝜔̇ 𝜉 + 𝐽𝜉𝜂 (𝜔𝜂2 − 𝜔𝜉2 ) + (𝐽𝜉𝜁 𝜔𝜂 − 𝐽𝜂𝜁 𝜔𝜉 )𝜔𝜁 = 𝑀𝐶𝜁 ,

The moments of inertia in this case are determined with respect to the body embedded frame
Cξηζ with the origin in the body mass center and in the general case, for a non-symmetric or
nonhomogeneous body, the inertia products are nonzero terms, because the axes Cξηζ will not be
principal axes.
In the case of symmetric body the axes Cξηζ are principal axes (Jξη = 0, Jηζ = 0, Jζξ = 0) and
from (2.11) we obtain
𝑑𝜔𝜉
𝐽𝜉 𝑑𝑡
− ( 𝐽𝜂 − 𝐽𝜉 )𝜔𝜂 𝜔𝜉 = 𝑀𝐶𝜉 ,

𝑑𝜔𝜂
𝐽𝜂 − ( 𝐽𝜁 − 𝐽𝜁 )𝜔𝜁 𝜔𝜁 = 𝑀𝐶𝜂 , (2.12)
𝑑𝑡

𝑑𝜔𝜁
𝐽𝜁 − ( 𝐽𝜉 − 𝐽𝜂 )𝜔𝜉 𝜔𝜂 = 𝑀𝐶𝜁 ,
𝑑𝑡

Expressing ωξ , ωη, ωζ in terms of Euler angles (ϕ, ϑ, ψ) or in normalized quaternions (e0, e1, e2,
e3) one can obtain the motion equations in the chosen coordinates[ 27].

2.2 Precession of a Body

The term precession is used to indicate that the body angular velocity vector rotates during the
motion.
In general case of a body motion we use (2.11) to describe the body rotation. There is no
analytical solution of (2.11) in general case but we can solve these equations numerically for any
data and obtain the body motion as well as the changes in angular velocity vector (ω) and
angular momentum vector (KC). Some exemplary results of such solutions for torque-induced
body motion are shown in Fig.3 where vectors ω and KC are depicted for some time instant. We
can see that both vectors change their directions and values during the body motion. From
presented cases follow that these bodies undergo precession (i.e., the vector ω rotates).

Figure 3.The precessing non-symmetric body: changes of angular momentum (KC) and angular
velocity (ω) vectors direction due to nonzero moments of external forces: (a) for MCξ ≠ 0, MCη
≠0, MCζ ≠ 0 and (b) for MCξ ≠ 0, MCη ≠ 0, MCζ = 0
Beyond the torque-induced precession torque-free precession is also possible. In the case of
torque-free motion of a body the moment of external forces with respect to the body mass center
is zero (MC = 0) and from (4.7) we have dKC/dt = 0. It means that the body angular momentum
vector with respect to the mass center is constant vector and there are no changes in the value
neither in the direction of KC (KC = const). For some special cases for symmetric bodies and

DMS6021 - Dynamics and Control of Mechanical Systems Page 5

zero moments of external forces (MC = 0) – an analytical solution of (2.11) can be obtained [23].
We present such solutions for symmetric top (which is the model of ideal coin) and spherical top
(which is the model of ideal die and torque-free motion of a roulette ball).

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 Jξ = Jη < Jζ and prolate symmetric top (cigar
shaped) with Jξ = Jη > Jζ . (A body is an asymmetric top if all three moments of inertia are
different.)

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 (Jξ = Jη = Jζ ).

2.2.1 Precession of Symmetric Top

Taking into account the symmetry of body (symmetric top) Jξ = Jη =J1, Jζ = J3, Jξη = Jξζ = Jηζ = 0
and assuming zero value of forces moment acting on the body (MCξ = MCη = MCζ = 0), from
(2.12) we get the following equations:

𝐽𝜉 𝜔̇ 𝜉 − ( 𝐽1 − 𝐽3 )𝜔𝜁 𝜔𝜂 = 0 ,

𝐽1 𝜔̇ 𝜉 − ( 𝐽3 − 𝐽1 )𝜔𝜁 𝜔𝜉 = 0 , (2.13)

𝐽3 𝜔̇ 𝜁 = 0,

Solving (2.13) together with initial conditions (i.e., the initial angular velocity components

ωξ(0) = ωξ0, ωη(0) = ωη0, ωζ (0) = ωζ0) we obtain the components of

angular velocity of the body in the form
𝐽1 −𝐽3 𝐽 −𝐽3
𝜔𝜉 (𝑡) = 𝜔𝜂𝑜 sin ( 𝐽1
𝜔𝜁𝑜 𝑡) + 𝜔𝜉𝜊 𝑐𝑜𝑠 ( 1𝐽 𝜔𝜉𝜊 𝑡) ,
1

𝐽1 −𝐽3 𝐽1 −𝐽3
𝜔𝜂 (𝑡) = 𝜔𝜂𝜋 𝑐𝑜𝑠 ( 𝐽1
𝜔𝜉𝜊 𝑡) − 𝜔𝜉𝜊 𝑠𝑖𝑛 ( 𝐽1
𝜔𝜁𝜊 𝑡) , (2.14)

𝜔𝜁 (𝑡) = 𝜔𝜁𝜊 ,

For symmetric top the vector ω has constant value

2 2 + 𝜔2
𝜔 = √𝜔𝜉𝜊 + 𝜔𝜂𝜊 𝜁𝜊 (2.15)

From (2.14) arise that 𝜔𝜁 (t) is time independent and two other components can be transformed to
the following form:

𝜔𝜉 (𝑡) = 𝜔𝜏 sin(𝜔𝑝 𝑡 + 𝜑𝜏 ) , 𝜔𝜂 (𝑡) = 𝜔𝜏 cos (𝜔𝑝 𝑡 + 𝜑𝜏 ), (2.16)

With the constant coefficients

𝐽1 −𝐽3 𝜔
𝜔𝑝 = 2
𝜔𝜁𝜊 , 𝜔𝜏 = √𝜔𝜉𝜊 2 ,
+ 𝜔𝜂𝜊 𝜑𝜏 = 𝑎𝑟𝑐𝑡𝑔 𝜔 𝜉𝜊 (2.17)
𝐽1 𝜂𝜊

DMS6021 - Dynamics and Control of Mechanical Systems Page 6

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.

Kξ = J1𝜔𝜏 sin (𝜔𝑝 t + 𝜑𝜏 ) ,

Kη = J1ωτ cos ((𝜔𝑝 t + 𝜑𝜏 ) , (2.18)

Kζ = J3ωζ0 ,

and the constant value as

2
𝐾𝑐 = √𝐽12 (𝜔𝜉0 2
+ 𝜔𝜂0 2
) + 𝐽30 2
𝜔𝜁0 (4.19)

Figure 4.Torque-free precession of symmetric top : vectors of angular velocity (ω) and angular
momentum (KC).
We can also obtain the angle γ between the symmetry axis of the body (ζ ) and the angular
momentum vector (KC) and the angle α between ζ and the vector 𝝎 ̃ (see Fig.4) as
𝐽3 𝝎𝜁0 𝝎𝜁0
𝛾 = arccos 𝐾𝑐
= const. , 𝛼 = arccos 𝝎
= const. (2.20)

The angles γ and α have constant value, but they are measured from axis ζ which moves
(changes its direction). While the angular momentum vector KC has constant value and direction
in the spatial frame xyz we can use the angle β = α - γ to obtain the orientation of the vector ω
̃
with respect to the fixed vector KC. From (2.20) we have
(𝐽3 −𝐽1 )𝝎𝜁0 ωτ
𝛽 = 𝛼 - 𝛾 = arcsin = const. (2.21)
𝐾𝐶𝜔

The vectors of angular velocity and angular momentum for symmetric top are also presented in
Fig. 2.5, where ω and KC are shown in the spatial frame xyz as well as in the body embedded
frame ξηζ . The angular velocity vector ω in the spatial frame xyz rotates around the angular
momentum vector (Fig. 2.5a). Both vectors rotate in the body embedded frame ξηζ (Fig. 2.5b).

2.2.2 Torque-Free Motion of Spherical Top

For torque-free motion of spherical top (Jξ = Jη = Jζ = J1), (4.13) simplify to the form

DMS6021 - Dynamics and Control of Mechanical Systems Page 7

Figure 5.The precessing symmetric top Figure 6.Torque-free motion of
spherical tops

𝐽1 𝜔̇ 𝜉 = 0 , 𝐽1 𝜔̇ 𝜂 = 0 , 𝐽1 𝜔̇ 𝜁 = 0 , (2.22)

Fig.5 The precessing symmetric top: angular momentum (KC) and angular velocity (ω) vectors
in (a) spatial frame xyz and (b) body embedded frame ξηζ . Fig.6 Torque-free motion of
spherical tops: constant angular momentum (KC) and angular velocity (ω) vectors’ directions
and values for a ball, tetrahedron die, and cubic die (no precession motion)

And we get the solution: ωξ = ωξ0, ωη = ωη0, ωζ = ωζ 0. It means that the angular momentum
vector (KC) and angular velocity vector (ω) have constant values and directions during the body
motion. We will call such a motion as no precession motion. Illustration of this case is shown in
Fig. 2.6.

2.3 Orientation of a Rigid Body

The orientation of a rigid body (orientation of a body embedded frame ξηζ ) with respect to the
local reference frame x y z is described by [21], [22], [26], [29].

r = R r’ (2.23)

where,
r – the column matrix representing the vector r of the position coordinates of an arbitrary point A
(Fig.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 ξηζ ,with respect to the
frame x y z )

DMS6021 - Dynamics and Control of Mechanical Systems Page 8

Figure 7.Point A position vectors: (a) r– in the initial position and (b) r’– in the final position.
The inverse transformation is defined by
r’ = R-1 r = RT r (2.24)

(R-1 is the inverse matrix of R, and R-1 = RT, whereas RT is the transpose matrix of R).
The matrix R can be expressed in different ways. Its form depends on the coordinates that are
chosen for the body orientation description. The elements of the rotation matrix (R) are cosines
of the angles between the axes of body reference frames x y z and ξηζ (Fig. 4.7b)

cos ∢(𝑥 , , 𝜉) 𝑐𝑜𝑠∢(𝑥 , , 𝜂) 𝑐𝑜𝑠∢(𝑥 , , 𝜁)

[𝑐𝑜𝑠∢(𝑦 , , 𝜉) cos ∢(𝑦 , , 𝜂) cos ∢(𝑦 , , 𝜁)] (2.25)
cos ∢(𝑧 , , 𝜉) cos ∢(𝑧 , , 𝜂) cos∢(𝑧 , , 𝜁)

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 (Ω ̃ 𝜉 = 𝑅̇ 𝑅 𝑇 ) which is
necessary in the dynamic analysis of the body. Matrix Ω ̃ is obtained in the xyz reference frame.
We use the symbol Ω ̃ 𝜉 for the body angular velocity matrix defined by the components in the
̃ 𝜉 = 𝑅𝑇 Ω
body embedded frame ξηζ . This matrix is expressed as Ω ̃ 𝑅.

2.3.1 Euler Angles and Other Conventions

In what follows, we adapt the conventions of [20], [25], [27], [29] which are different from the
ones used in [24]. Using Euler angles (or any other from 12 possible conventions of specifying
the relative orientation of a body) the rotation matrix R is the composition of three consecutive
rotations: ϕ1, ϕ2, ϕ3, around axes ξηζ of the frame embedded and fixed in the body

𝑅 = 𝑅1 (𝜑1 )𝑅2 (𝜑2 )𝑅3 (𝜑3 ) (2.26)

Whereas R1(ϕ1), R2(ϕ2), R3(ϕ3) are the matrices of successive rotations around ξ, η, and ζ axes
The definition of Euler angles is not unique. In works of various authors different sets of angles
describe the body orientations and other naming conventions for the same angles are used. These
conventions depend on the axes about which the rotations are carried out and on the rotation
sequences. Basic definitions, names, and expressions used in this work are briefly presented
below. For the rotation of the body of value ϕi around the ξ axis (Fig.2.8) the rotation matrix has
the following form:

1 0 0
𝑅𝜉 (𝜑𝑖 ) = [ 0 𝑐𝑜𝑠𝜑𝑖 − 𝑠𝑖𝑛𝜑𝑖 ] (4.27)
0 𝑠𝑖𝑛𝜑𝑖 𝑐𝑜𝑠𝜑𝑖

DMS6021 - Dynamics and Control of Mechanical Systems Page 9

For the rotation of value ϕi around the η axis

𝑐𝑜𝑠𝜑𝑖 0 𝑠𝑖𝑛𝜑𝑖
(𝜑 )
𝑅𝜂 𝑖 = [ 0 1 0 ] (2.28)
−𝑠𝑖𝑛𝜑𝑖 0 𝑐𝑜𝑠𝜑𝑖

Figure 8.Initial position of a body and rotation; (a) the rotation sequence: ϕ1 – around the ξ (b), ϕ2
– around the η (c), ϕ3 – around the ζ (d)
The rotation of value ϕi around the ζ axis leads to

𝑐𝑜𝑠𝜑𝑖 − 𝑠𝑖𝑛𝜑𝑖 0
𝑅𝜁 (𝜑𝑖 ) = [𝑠𝑖𝑛𝜑𝑖 𝑐𝑜𝑠𝜑𝑖 0] (2.29)
0 0 0
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 ξηζ of the
body embedded frame by symbols 1, 2, 3 (ξ → 1, η → 2, ζ → 3), possible rotation sequences can
be represented as follows: 121 (ξ-η-ξ),
123 (ξ-η-ζ ), 131 (ξ-ζ -ξ), 132 (ξ-ζ -η), 212 (η-ξ-η), 213 (η-ξ-ζ ), 231 (η-ζ -ξ), 232 (η-ζ -η), 312 (ζ
-ξ-η), 313 (ζ -ξ-ζ ), 321 (ζ -η-ξ), 323 (ζ -η-ζ ).

Classical Euler angles are: ϕ1 = ψ, ϕ2 = ϑ, ϕ3 = ϕ, that indicate consecutively

the rotations: by angle ϕ1 = ψ around the ζ axis, by ϕ2 = ϑ around the new position of the ξ axis,
and by ϕ3 = ϕ around the ζ (rotation sequence is abbreviated as ζ -ξ-ζ or 313). The rotation matrix
R for such angles is obtained by substituting into (2.26) the following rotational matrices:

𝑅1 (𝜑1 ) = 𝑅𝜁 (𝜓) , 𝑅2 (𝜑2 ) = 𝑅𝜉 (𝜗), 𝑅3 (𝜑3 ) = 𝑅3 (𝜑) , (4.30)

That means,

𝑅 = 𝑅𝜁 (𝜓)𝑅𝜉 (𝜗)𝑅𝜁 (𝜑) . (2.31)

The set of 12 possible rotation sequences and the singularity condition for each case and values
of angles causing singularities in numerical solutions are obtained.

2.3.2 Euler’s Parameters

An alternative to Euler’s angles and similar conventions of body orientation description are
Euler’s parameters (also called Euler symmetric parameters and known in mathematics as
normalized quaternions) [27], [28], [30].They are very useful in representing the rotations due to
some advantages in comparison to other representations. The main advantage of Euler

DMS6021 - Dynamics and Control of Mechanical Systems Page 10

parameters is that they do not produce any singularities in numerical solutions of body motion
equations.

In the matrix notation Euler’s parameters are represented by the column matrix.
𝜙
𝑐𝑜𝑠 2
𝑒0 𝜙
𝑒1 𝜈1 𝑠𝑖𝑛 2
𝑃 = [𝑒 ] = 𝜙
(2.32)
2 𝜈2 𝑠𝑖𝑛
𝑒3 2
𝜙
[𝜈3 𝑠𝑖𝑛 ] 2

𝜙
𝑒 𝑐𝑜𝑠
2
Or, shortly 𝑃 = [ 0 ] = [ ] (2.33)
𝑒 𝜊
𝑉 𝑠𝑖𝑛
𝜙
2

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 φ is a scalar value. The rotation matrix (2.25) can be expressed by Euler
parameters as [28]

𝑅 = (2𝑒02 − 1)𝐼 + 2𝑒𝑒 𝑇 + 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 ] (2.35)
−𝑒2 𝑒1 0

Figure 9.The body rotation about given angle; by φ – around the vector vo.
The expanded form of the matrix R (2.34) – expressed by unit quaternions (e0, . . . , e3) –
has the following form:

DMS6021 - Dynamics and Control of Mechanical Systems Page 11

 −1 + 2𝑒02 + 2𝑒12 2𝑒1 𝑒2 − 2𝑒0 𝑒3 2𝑒0 𝑒2 + 2𝑒1 𝑒3
𝑅 = [2𝑒1 𝑒2 + 2𝑒0 𝑒3 − 1 + 2𝑒02 + 2𝑒22 − 2𝑒0 𝑒1 + 2𝑒2 𝑒3 ] (2.36)
−2𝑒0 𝑒2 + 2𝑒1 𝑒3 2𝑒0 𝑒1 + 2𝑒2 𝑒3 − 1 + 2𝑒02 + 2𝑒32

The antisymmetric matrix Ω ̃ ξ (Ω

̃ 𝜉 = 𝑅 𝑇 R ˙ ) containing scalar components of the coin angular
velocity vector – in the body embedded frame ξηζ – has the form

0 − 𝑒̇3 𝑒0 + 𝑒̇2 𝑒1 − 𝑒̇1 𝑒2 + 𝑒̇0 𝑒3 𝑒̇2 𝑒0 + 𝑒̇3 𝑒1 − 𝑒̇0 𝑒2 − 𝑒̇1 𝑒3

̃𝜉 = 2 [
Ω 0 − 𝑒̇1 𝑒0 + 𝑒̇0 𝑒1 + 𝑒̇3 𝑒2 − 𝑒̇2 𝑒3 ] (2.37)
𝑎𝑠𝑦𝑚. 0

The angular velocity vector of the coin 𝜔

̃𝜉 in the body embedded frame ξηζ is expressed by the
column matrix
𝜔𝜉 𝑒̇1 𝑒0 − 𝑒̇0 𝑒1 − 𝑒̇3 𝑒2 + 𝑒̇2 𝑒3
𝜔
̃𝜉 = [ 𝜂 ] = 2 [𝑒̇2 𝑒0 + 𝑒̇3 𝑒1 − 𝑒̇0 𝑒2 − 𝑒̇1 𝑒3]
𝜔 (2.38)
𝜔𝜁 𝑒̇3 𝑒0 − 𝑒̇2 𝑒1 + 𝑒̇1 𝑒2 − 𝑒̇0 𝑒3

Where as

ω̇ξ −e1 ë 0 + e0 ë 1 + e3 ë 2 − e2 ë 3
ω̇
̃ ξ = [ω̇η ] = 2 [−e2 ë 0 − e3 ë 1 + e0 ë 2 + e1 ë 3 ] (2.39)
ω̇ζ −e3 ë 0 + e2 ë 1 − e1 ë 2 + e0 ë 3

The column matrix containing xyz scalar components of the coin angular velocity vector (i.e., the
components in fixed spatial frame) has the form
𝜔𝑥 𝑒̇1 𝑒0 − 𝑒̇0 𝑒1 + 𝑒̇3 𝑒2 − 𝑒̇2 𝑒3
𝜔
̃ = [ 𝑦 ] = 2 [𝑒̇2 𝑒0 − 𝑒̇3 𝑒1 − 𝑒̇0 𝑒2 + 𝑒̇1 𝑒3 ]
𝜔 (2.40)
𝜔𝑧 𝑒̇3 𝑒0 + 𝑒̇2 𝑒1 − 𝑒̇1 𝑒2 − 𝑒̇0 𝑒3

Using Euler’s parameters provide to avoid singularities in numerical solutions of body rotation
problems. The coin dynamics is one of such problems. The equation of coin dynamics and
special attention is paid to the coin dynamics description in quaternions (Air Resistance Forces
and Moments) [31], [32], [33], [34], [35].

2.5 Modelling of Bodies Impact

Let us consider that the body collides with a floor when the point A on its surface touches the
floor as shown in Fig. 2.10. In the rigid body dynamics the most common model of collision of
bodies is based on Newton’s hypothesis [24], [25].
𝑣́
−𝑥 = 𝑣𝐴𝑧 , (2.41)
𝐴𝑧

Where χ is the coefficient of restitution (0 < χ < 1), A stands for the body point which makes
contact with the floor at the instant of impact (Fig. 2.10), vAz and vAz are projections of the
velocity of the point A on the direction (z) normal to the impact surface, respectively, before and
after the impact [36], [37].

DMS6021 - Dynamics and Control of Mechanical Systems Page 12

Figure 10.Velocity vector; vA of the point A after the collision and its scalar components vAτ, vAν
, vAz

3. RIGID-BODY SIMULATION
In this section we will have the implementation of rigid body dynamics with simulation.
Knowing the basic differential equation of rigid body dynamics is the core to do the simulation.
A system of one or more rigid bodies is described by a generalized state vector whose
components are identified [38].Simulation results for rigid body dynamics is done with
MATLAB.

Sample Problem 1: (Fixed-Axis Rotation)

The pendulum has a mass of 7.5 kg with a mass center at G and a
radius of gyration about the pivot O of 295 mm. If the pendulum is
released from rest when 𝜃= 0, plot the total force supported by the
bearing at O along with its normal and tangential components as a
function of 𝜃. Let 𝜃 range between 0° and180° .

Free Body Diagram

The free body and mass acceleration diagrams are identical to those in the
sample problem. The main difference in approach is that we will obtain results
at an arbitrary angle 𝜃rather than at 60° .

Problem Formulation
𝑔𝑟̅
[∑ 𝑀𝑂 = 𝐼𝑜 𝛼] ; 𝑚𝑔𝑟̅ 𝑐𝑜𝑠𝜃 = 𝑚𝑘𝑜2 𝛼 ; 𝛼= 𝑐𝑜𝑠𝜃
𝑘𝑜2

𝜔 𝑔𝑟̅ 𝜃 𝑔𝑟̅ 2𝑔𝑟̅

[𝜔𝑑𝜔 = 𝛼𝑑𝜃] ; ∫0 𝜔𝑑𝜔 = 2 ∫0 𝑐𝑜𝑠𝜃𝑑𝜃 = 2 𝑠𝑖𝑛𝜃 ; 𝜔2 = 𝑠𝑖𝑛𝜃
𝑘 𝑜 𝑘 𝑜 𝑘𝑜2

[∑ 𝐹𝑁 = 𝑚𝑟̅ 𝜔2 ] ; 𝑂𝑛 − 𝑚𝑔𝑠𝑖𝑛𝜃 = 𝑚𝑟̅ 𝜔2 ; [∑ 𝐹𝑡 = 𝑚𝑟̅ 𝜔2 ] ; −𝑂𝑡 + 𝑚𝑔𝑐𝑜𝑠𝜃 = 𝑚𝑟̅ 𝜔2

After substituting for 𝛼 and 𝜔 we have,

DMS6021 - Dynamics and Control of Mechanical Systems Page 13

 𝑟̅ 2 𝑟̅ 2
𝑂𝑛 = 𝑚𝑔 (1 + 2 ) 𝑠𝑖𝑛𝜃 ; 𝑂𝑡 = 𝑚𝑔 (1 − ) 𝑐𝑜𝑠𝜃
𝑘𝑜2 𝑘𝑜2

The magnitude of the force at O is,

O = √(𝑂𝑛 )2 + (𝑂𝑡 )2

After substituting r = 0.25 m, k0 = 0.295 m, m = 7.5 kg, and 𝑔 = 9.81 m/s2 all forces will be
functions of 𝜃 only.

MATLAB Script

%%%%%%%%%%%%%%%%%% Script %%%%%%%%%%%%%%%%%%%

% This script plots the force at O and its

% components as functions of theta

≪rb = 0.25;
≪k0 = 0.295;
≪mg = 7.5*9.81;
≪theta = 0:0.01:pi;
≪On = mg*(1+2*(rb/k0)^2)*sin(theta);
≪Ot = mg*(1-(rb/k0)^2)*cos(theta);
≪O=sqrt(On.^2+Ot.^2);
≪td=180*theta/pi; % converts to degrees
≪plot(td,O,td,On,td,Ot)
≪xlabel('theta (degs)')
≪title('Force at O (Newtons)')

Explanation of this figure is missing

Figure 11. Motion simulation for pendulum

DMS6021 - Dynamics and Control of Mechanical Systems Page 14

 4. CONCLUSION
A general rigid body dynamics system is analyses in translation and rotational motion about
fixed axes. Because when the body is translating, the axis of rotation is no longer fixed in space.
In this paper the orientations of the body describe we both with Euler’s parameters and Euler’s
angles. And some of precession of the rigid body and the energy dissipation at successive
collisions are discussed. Relay 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. The dynamic
equations of motion in rigid body form are derived using a Euler technique and the momentum obtained
using MATLAB fixed axis-rotation. The plot of simulation diagram gives detail concepts general
motion and mechanical behaviour of the systems and it serves as whether acceptable or not for
the given design.

Genrally good report, but the scientific depth as a review article is not strong enough

DMS6021 - Dynamics and Control of Mechanical Systems Page 15

 REFERENCES
[1] J.L. Meriam and L.G. Kraige, Engineering Mechanics, DYNAMICS, 5th Edition 5/1, 5/2,
5/3, 5/4 (review) , 5/5, 5/6 (review).

[2] R. C. Hibbeler, Engineering Mechanics, Dynamics, Twelfth Edition Springer, New York
(1989)

[3] Leim, E., The General Problem of the Motion of Coupled Rigid Bodies About a Fixed Point,
Springer- Verlag, New York, 1965.

[4] Grubin, C., “On Generalization of the Angular Momentum Equation,” Journal of Engineering
Education, Vol. 51,1960, pp. 237-238.

[5] Grubin, C., “Dynamics of a Vehicle Containing Moving Parts,” Journal of Applied
Mechanics, Vol. 29, 1962, pp. 486-488.

[6] Roberson, R. E., “Torques on a Satellite Vehicle from Internal Moving Parts,” Journal of
Applied Mechanics, Vol. 25, 1958, pp. 196-200.

[7] Aristotle. Physics. Kessinger Publishing Co., 2004 edition, 330 B.C.

[8] Landon, V. D., and Stewart, B., “Nutational Stability of an Axisymmetric Body Containing a
Rotor,” Journal of Spacecraft and Rockets, Vol. 1, No. 6, 1964, pp. 682-684.

[9] J. L. Meriam, L.G. Kraige Engneering Mechanics,DYNAMICS, 6th edition Wiley.

[10] Mach, E. Die Mechanik in ihrer Entwicklung: Historisch{kritisch dargestellt. 1886.

[11] ARMSTRONG W. W., GREEN M. W.: The dynamics of articulated rigid bodies for
purposes of animation. The Visual Computer 1, 4 (1985), 231–240.

[12] P. Dupont, “The effect of coulomb friction on the existence and uniqueness of the forward
dynamics problem,” in Proc. 1992 IEEE Int. Conf. Robot. Automat., Nice, France, May 1992,
pp. 1442- 1447.

[13] P. Lotstedt, “Coulomb friction in two-dimensional rigid body system,” ZAMM, vol. 61, no.
12, pp. 605-615, Dec. 1981.

[14] M. T. Mason and Y. Wang, “On the inconsistency of rigid-body frictional planar
mechanics.” in Proc. 1988 IEEE Int. Conf. Robot. Automat., Philadelphia, PA, Apr. 1988.

[15] P. Painleve, “Sur les lois du frttement de glissement,” Comptes Rendus de l’Acad. Des Sci.,
vol. 121, pp. 112-115, 1895.

[16] V. T. Rajan, R. Burridge, and J. T. Schwartz, “Dynamics of a rigid body in frictional contact
with rigid walls: Motion in two dimensions,” in Proc. 1987 IEEE Int. Conf. Robot.
Automat.,Raleigh, NC, 1987, pp. 671-677.

DMS6021 - Dynamics and Control of Mechanical Systems Page 16

[17] DeLoura et al. 2000. Game Programming Gems. Charles River Media.

[18] Eberly, H. D. 2010. Game Physics, Second Edition. Morgan Kaufmann.

[19] Arnold, V.I.: Mathematical Methods of Classical Mechanics. Springer, New York (1989)
[20] Beatty, M.F., Jr.: Principles of Engineering Mechanics. Plenum Press, New York (1986)
[21] Landau, L., Lifschitz, E.: Mechanics. Pergamon Press, Oxford (1976)
[22] Kittel, C., Knight, W.D., Ruderman, M.A.: Mechanica. Berkeley Physics Course –
Volume1. McGraw-Hill, New York (1965)
[23] Roth, E.J.: Dynamics of a System of Rigid Bodies. Dover, New York (1960)
[24] Wittaker, E.T.: Analytical Dynamics of Particles and Rigid Bodies. Cambridge University
Press, Cambridge (1937)
[25] Goldstein, H.: Classical Mechanics. Addison-Wesley, Reading (1950)

[26] Ken Shoemake, Animating Rotation with Quaternion Curves Proc. SIGGRAPH 1985, pp.
245 - 254.
[27] Mongomery, R.: How much does a rigid body rotate? Amer. J. Phys. 59, 394 (1991)
[28] Diebel, J.: Representing Attitude: Euler Angles, Unit Quaternions, and Rotation Vectors.
http://ai.stanford.edu/∼diebel/. Cited 15 May 2007
[29] Nikrevesh, P.E.: Computer-Aided Analysis of Mechanical Systems. Prentice-Hall,
Englewood Cliffs, NJ (1988)
[30] Marsden, J.E., Ratiu, T.S.: Introduction to Mechanics and Symmetry. Springer, New York
(1994)
[31] Long, L., Weiss, H.: The velocity dependence of aerodynamic drag: A primer for
mathematicians. Amer. Math. Monthly 106, 127 (1999).

[32] Landau, L., Lifschitz E.: Fluid Mechanics. Butterworth and Heinemann, New York (1987)
[33] McCormick, B.: Aerodynamics, Aeronautics and Flight Mechanics. Wiley, London (1995)
[34] Thompson, P.A.: Compressible Fluid Dynamics. McGraw-Hill, London (1972)
[35] White, F.M.: Viscous Fluid Flows. McGraw-Hill, London (1991)
[36] Mosekilde, E.: Topics in Nonlinear Dynamics: Applications to Physics, Biology and
Economic Systems. World Scientific, Singapore (1996)
[37] Feldberg, R., Szymkat, M., Knudsen, C., Mosekilde, E.: Iterated-map approach to die
tossing. Phys. Rev. A 42/8, 4493 (1990)

[38] D Baraff. Analytical methods for dynamic simulation of non-penetrating rigid bodies.
Computer Graphics, (Siggraph ’89 Proceedings), Jul 1989, pp 223-232.

DMS6021 - Dynamics and Control of Mechanical Systems Page 17