the analysis of r-s-s-r spatial four-bar mechanism
TRANSCRIPT
Scholars' Mine Scholars' Mine
Masters Theses Student Theses and Dissertations
1969
The analysis of R-S-S-R spatial four-bar mechanism The analysis of R-S-S-R spatial four-bar mechanism
David Perng Chyi
Follow this and additional works at: https://scholarsmine.mst.edu/masters_theses
Part of the Mechanical Engineering Commons
Department: Department:
Recommended Citation Recommended Citation Chyi, David Perng, "The analysis of R-S-S-R spatial four-bar mechanism" (1969). Masters Theses. 5380. https://scholarsmine.mst.edu/masters_theses/5380
This thesis is brought to you by Scholars' Mine, a service of the Missouri S&T Library and Learning Resources. This work is protected by U. S. Copyright Law. Unauthorized use including reproduction for redistribution requires the permission of the copyright holder. For more information, please contact [email protected].
T1-iE ANALYSIS OF R-S-S-R SPATIAL
FOUR-BAR MECHANISM
BY
DAVID FERNG CHYI, 1944-
A
THESIS
submitted to the faculty of the
UNIVERSITY OF MISSOURI - ROLLA
in partial fullfillment of the work required for the
Degree of
MASTER OF SCIENCE IN MECHANICAL ENGINEERING
Rolla, Missouri
Approved by
T2315 c. I 55 pages
183294
ABSTRACT
Although many methods are known for the kinematic
analysis of spatial mechanisms, such as descriptive
geometry, vector analysis, dual number theory, matrices,
screw calculus, and various other methods, a better way
to analyze the specific R-S-S-R four-bar mechanism is
to use tensor methods in order to eliminate the apparent
ly formidable and tedious tasks of mathematically for
mulating problems and obtaining solutions.
The analysis described in this paper is developed
basically by using tensor notations and operations, and
the calculations take advantage of the capabilities of
modern digital computers. A FORTRAN program for the IBM
360 Data Processing System which will analyze the dis
placement of each joint pairs, angular velocity, and
angular acceleration of the output link is included.
This program is available upon request.
Several example problems were solved with the
program and only one is presented in this thesis.
ii
iii
ACKNOWLEDGEMENT
The author is greatly indebeted to Dr. Chung-you Ho,
of the Department of Mechanical Engineering at the Univer
sity of Missouri at Rolla, for his advice in the formu
lation of this thesis, and for his guidence and valuable
criticism throughout this research program.
He also wishes to thank Mr. Kuo-Chien Hsei for his
help with problems regarding this study.
Also thanks to the Computer Center of the Univer
sity of Missouri at Rolla for the use· of their IBM 360
digital computer.
iv
TABLE OF CONTENTS
Page
ABSTRACT ••••••••••••••••• o •••••••••••••••••••••••••••••••••• ii
ACKNOWLEDGEMENT ••••••••••••••••••••••••••••••••••• o •••••• , • iii
TABLE OF CONTENTS. , ••••••••••••••••••••••• , ••••••••••••••••• i v
LIST OF ILLUSTRATIONS ••••••••••••••• o•••••••••••••••••••••••vi
LIST OF TABLES. o ••••••••••••••••••••••••••••••••••••••••••• vii
CHAPTER
CHAPTER
I. INTRODUCTION ••••••••••••••••••••••••••••••• o •••• 1
II. LITERATURE REVIEW•••••••••••••••••••••••••••••••5
A. DEFINITIONS AND BACKGROUND MATERIALS ••••••••• 5
1. KINEMATIC CONCEPTS ••••••••••••••••••••••••• 5
a. MECHANISM••••••••••••••••••••••••••••••o•5
b, KINEMATIC LINK •••••• , ••••••••••••••••• o •• 6
c, KINEMATIC PAIR ••••••••••••••••••••••••••• 6
d, MOTION ••••••••••••••••••• o•••••••••••••••8
e. GRUEBLER'S CRITERION OF MOBILITY ••••••••• 9
2. PRELIMINARY MATHEMATICS•••••••••••••••••••l2
a. NOTATIONAL CONVENTION ••••••••••••••••••• l2
b. DEFINITION OF A TENSOR••••••••••••••••••l4
c. SOME TENSOR PROPERTIES, ••••••••••••••••• 15
CHAPTER III. POSITION ANALYSIS OF THE SPATIAL FOUR-BAR
LINKAGE ••••••••••••••••••••• 0 •• I •••••• 0 •••••••• 17
A. GENERAL DESCRIPTION•••••••••••••••••••••••••l7
B. COORDINATE RELATIONSHIP•••••••••••o•••••••••20
CHAPTER IV. METHOD OF APPROACH AND DERIVATION ••••• o.o••••••22
CHAPTER
v
A. GOVERNING EQUATION OF R-S-S-R MECHANISM ••• 22
B. DERIVATION OF THE OUTPUT FOLLOWER ANGLE
e111 , AND THE FOLAR, AZIMUTHAL ANGLE OF
THE SECOND LINK e11 , %11 ••••••••••••••••••22
C. OUTPUT FOLLOWER ANGULAR VELOCITY, aiii •••• 27 .. III D. OUTPUT ANGULAR ACCELERATION, e ••••••••• 29
V. CONCLUSION ••.••••••••••••••• , •••••••• o ••••••• 31
BIBLIOGRAPHY •••••••••••••••••••••••••••••••••••••••••••••• 32
APPENDICES ••• •••••••• , •••••••••••••••••••••••••••••••••••• 35
APPENDIX A. COMPUTER PROGRAM ••••••••••••••••• 36
APPENDIX B. FIGURES OF SOME VARIABLES OF
R-S-S-R SPATIAL MECHANISM •••••••• 42
VITA. e e a • • a e a • a • a G a • a a • • • • • 1 e • • e e • • • e • • • • e a a • • e a • • • • • 1 1 t a • 48
vi
LIST OF ILLUSTRATIONS
Figures Page
1. General diagram of four-bar spatial linkage ••••••• l8
2. Coordinate relationshiP•••••••••••••••••••••••••••20
3. R-S-S-R four-bar mechanism••••••••o•••••••••••••••23
4. The azimuthal angle of the second link versus
5. The polar angle of the second link versus input
angle •••••••••••••• o ••••••• o ••••••••••• o •••••••••• 44
6. The output angle versus input angle ••••••••••••••• 45
7. The output angular velocity versus input angle •••• 46
8. The output angular acceleration versus input
angle •••••••••••••••••••••••••••••• o •••••••••••••• 47
CHAPTER I
INTRODUCTION
The broad category of devices known as mechanisms
may be divided into the two major categories of planar
mechanisms and spatial mechanisms.
In planar mechanisms, the motion of all links can
be completely described in a single plane, one single
graphical projection. Two common examples of planar
mechanisms are the four bar linkage and the slider
mechanism.
However, other mechanisms are known whose motion
can not be described in a single projection; they have
three-dimensional motion. Common examples of this type
are the Hooke or Universal Joint, the Bennett mechanism,
and the Bricard mechanism. These form the second broad
category, spatial mechanism.
Recently, more and more emphasis has been placed
on the three dimensional linkages. Because of this extra
dimension habit, the kinematicians have more design pa
rameters than the traditional two-dimensional mechanisms.
Due to the rapid development of high speed digital com
puters and various elegant mathematical methods, spatial
mechanisms can be performed with greater ease.
1
In Germany, analysis and synthesis techniques were
* performed by Altman(l) , Beyer(2), Hain(3), Keler(4),
an~ ot~ers, based on graphical and analytical geometry
of dual quantities. In the U.s.s.R., tensor analysis
was used by s. G. Kislitsyn(5) in the theory of spatial
mechanism. While in this country, Denavit and Hartenberg
(6) adopted the matrix calculus; Chace(?), Beyer and
Harrisberger(8) used the vector technique; Yang and
Freudenstein(9) chose matrices with dual-number qua
ternions and Ho(lO) developed the tensor method.
It seems very convenient if a spatial mechanism is
pursued by tensor methods. In spatial mechanism, almost
every quantity involved can be represented by tensor
notation; such as scalar quantity which is a tensor of
zeroth order, moment of inertia is the tensor of second
order, velo~ity and acceleration are the tensors of
first order, and other complicated quantities can be
expressed as the tensors of higher order.
Graphical and related methods have been important
because they avoid detail computation and provide a
visual perspective. With the advent of digital computers,
the computational advantage of graphical methods become
less significant. And as we know, it has always been
difficult and tedious to apply graphical methods to
* Numbers in the parenthesis refer to the Bibliography at the end of the thesis.
2
three-dimensional analysis.
Analysis and synthesis by conventional complex
mathematics has been very successful for two-dimen
sional problems. Extension to three-dimensional
analysis was suggested by Raven(ll), but nevertheless
conventional complex mathematics has remained a two
dimensional tool.
Matrix methods have been developed and applied
by Hartenberg, Denavit, and Uicker(l2). A computer
program, based on this method, will obtain position,
motion and force solutions for the complete motion
cycle of three-dimensional mechanism connected by
lower pairs in any single loopo The method has been
extended to general spatial mechanism. However, the
iteration technique required for the position solutions
is always subjected to the trunctional errors of the
computations, and iterpretation of the matrix equations
is also difficult. Nevertheless, it is a beautifully
formulated solution and it affords the available
numerical solution to an important category of mechanisms.
Dual numbers, quaternions, and vectors have been
applied to this kind of problems. These approaches have
advantages in the representation of spatial problems,
but also have some disadvantages especially in the
representation of complicated linkages not possessing
geometry symmetry.
3
The Instantaneous Screw Axis (ISA) theory adopted
by Skreiner(l3) is a newly developed method which has
a good approach to the problem. However, the positions
of the links relative to the ground link are not easily
determined through this method. It, therefore, is dis
carded in favor of the more direct tensor methods
introduced here.
The goal of machine design is to find a simple but
effective mechanism for transmitting a desired motion.
Spatial mechanisms have long been avoided by designers
because of their complexities. The construction of a
spatial mechanism is amazingly versatile. Sometimes,
an ingenious arrangement will be able to simplfy a
complex mechanism and accomplish the desired motion.
The spatial four-bar mechanism treated in this
paper for revolute-spherical-spherical-revolute joint
in the order given.
4
CHAPTER II
LITERATURE REVIEW
A. DEFINITIONS AND BACKGROUND MATERIALS
In order to analyze the R-S-S-R spatial four-bar
mechanism, it would be well to review a few important
definitions and concepts to form a general basis on
which to proceed.
1. KINEMATIC CONCEPTS
a. Mechanism
Probably the most important term requiring defini
tion in the study of kinematics is "mechanism". It is
a term having a definite meaning, but difficult to
define succinctly. A descriptive definition was given by
Reuleaux as "that imaginary concept of a device with
perfect geometry and rigidity••. Although this is true,
it is far from complete. The classical definition states
that a mechanism is an assemblage of kinematic links,
or machine parts, connected together by kinematic pairs,
or joints, used for the purpose of transforming one type
of motion into another. A mechanism is a closed, kine
matic chain of links. One link is a fixed frame or base,
and all motion is viewed with respect to this link as a
reference.
5
b. Kinematic Link
The term kinematic link was essential to the defi
nition of a mechanism and so it too must be defined. In
kinematics, the machine members of which a mechanism is
made are called links. In keeping with the purpose of
kinematics, all concepts of deformation due to load and
all stress-strain relationships are neglected. Kine
matically, the only purpose of the link is to hold a
specific spatial relationship between the several
kinematic pairs or joints, The particular physical
configuration, or the shape, size, weight, and material
are all only incidental in the study of how a mechanism
will move. A link is then a rigid body containing the
elements of at least two kinematic pairs or jointso
c. Kinematic Pair
In order to perform their function of motion,
transformation in a mechanism, the link must be con
nected by some sort of movable joints. The connections,
or joint, between pairs and links are called kinematic
pairs. Just as the link is the sole criterion of spatial
relationship between its two kinematic pair elements;
the pair is the only determining factor for the type
of motion permitted between the connected links. The
sole function of these pairs is to provide a connection
limiting the relative motion between two links to a
6
certain predetermined ~pe of motion.
Kinematic pairs can take on an infinite variety of
configurations. They are often difficult to recognize
by their physical appearence, but must be categorized
according to their characteristic motions. The categories
are wrapping connectors, lower pairs, and higher pairs.
The wrapping connectors usually take the form of belts
or chain and may be recognized by the fact that they
transmit motion in only one direction, tension. The
lower pairs, as identified by Reuleaux, are recognizable
by their clearly defined pair variable. They are six in
number: the revolute pair (R), which permits rotational
motion about an axis and is defined by a single variable
angle e; the prismatic pair (P), which permits translation
motion in one direction and is defined by a single dis
placement variableS; the screw pair (H), which permits
helical motion and is defined by either the rotation
angle e or translation displacement S, related through
9/2~ = S/h, where h is the lead of the screw (advance per
revolution); the cylindrical pair (C), which permits a
translation motion parallel to an axis and also a rota
tion motion about the same axis independently; the
spherical pair (S), which permits a globular rotation
about a point; and the last one is the planar pair (F),
7
which permits planar motion and is defined by two dis
placement variables and one rotation variable. Note that
the spherical pair may be viewed as wrapping pair for its
hollow element is wrapped around its full element, but it
8
is not a wrapping connector which transmits motion in only
one direction. The higher pairs, according to the definition
by Reuleaux, are the pairs with the surface elements so
shaped that only line or point contact are possible between
elem~nts.
d. Motion
ordinarily, there are two kinds of motion, one is
translation and the other is rotation. Both of them
include at least three distinctly different yet related
characteristics, i.e., displacement, velocity, and
acceleration, requiring a certain amount of time for
their completion.
Whenever a motion is described, the reference frame
with which the motion is related must be defined first,
In mechanism, it is tha very fact that motion exists that
implies reference frames of some sort on the moving parts
or a certain specified fixed coordinate. If the reference
frame of one machine part moves with respect to the
reference frame of another, we speak of relative motion.
When the reference frame of one part is fixed with respect
to the ground and the motions of the other parts are
referred to it, then these particular relative motions
may be termed absolute motions. Absolute motion is thus
a special case of relative motion, in which the reference
frame is fixed.
e. Gruebler's Criterion of Mobility
Before a spatial mechanism is established and motion
analysis started, the first question is whether or not
the mechanism will move. Gruebler's criterion (14) is a
quick and easy test for that. It is derived as follows:
since any rigid body has six degrees of freedom in space,
a mechanism of n links considered independently in space,
should have a total of 6n degrees of freedom. If these
links were combined to form a linkage, one link would
be taken to be a stationary frame of reference. Thus the
total number of degrees of freedom, f, with respect to
that stationary frame would become
f = 6(n - 1) (2-1)
when these links are connected by the various pairsf they
are imposed by certain restraints on the movability. The
total number of degrees of freedom which remains in the
mechanism is now k .
f = 6(n - 1) - ~ r~ ~1
(2-2)
where: ri is the number of restraints imposed by the i's
pair
9
k is the total number of pairs in the mechanism.
Equation (2-2) is essentially Gruebler's criterion
of movability. It shows the number of degrees of free
dom of a mechanism as a function of the number of links
and the number of restraints imposed by pairs.
The value of ri are different for different pairs
involved. Those for the lower pairs are listed in
Table I.
Table I
Values of ri for Lower Pairs
Pair Symbol ri
Revolute R 5
Prismatic p 5
Screw H 5
Cylindrical c 4
Planar F 3
Spherical s 3
It should be noted that this criterion has several
exceptions. In these cases some of the restraints become
redundent and the degrees of freedom may be more than
that predicted by equation (2-2), Some known exceptions
are the Bennett and the Bricard mechanism, Hooke's joint,
this R-S-S-R spatial mechanism, and all planar mechanisms.
In the case of planar mechanisms, another form of the
10
criterion should be used, that is k .
f = J(n - 1) - E (r1 - 3) i=l
The parameters in equation (2-J) are the same as
defined for equation (2-2).
11
(2-J)
Usually a mechanism is more interesting if it has one
degree of freedom. A single input will then completely
determine the motion of the entire mechanism. This
condition is known as constrained motion.
In a simple closed-loop chain, composed entirely
of binary links, k is equal to n and for constrained
motion n .
1 = 6(n - 1) - E (r 1 ) i=l
Since any kind of pair is permitted in a closed-n .
loop chain, E r 1 is always less than 5n. This yield i=l
the following inequality
1 ~ 6(n - 1) - 5n ~ n - 6
that is n ~ 7.
(2-4)
(2-5)
This means the number of links in a simple closed
loop chain for a constrained motion can not be more than
seven.
12
2. PRELIMINARY MATHEMATICS
Tensor calculus came to prominence with the development
of general theory of relativity by Einstein in 1916. It
provides the only suitable mathematical language for
general discussion of that theory. But actually the tensor
calculus is older than that. It was invented by the Italian
mathematicians Ricce and Levi-Vivita, showing its appli
cations in geometry and classical mathematical physics in
1900. Thus tensor calculus comes near to being a universal
language in mathematical physics. Not only does it enable
us to express general equations very compactly, but it
also guides us in the selection of physical variables,
by indicating, automatically, invariance with respect
to the transformation of coordinates.
a, Notational Conventions
A point in three-dimensional space located with
respect to a Cartesian coordinate system Xi , may also
be located with respect to another Cartesian coordinate • system, X. , by the equations ~
• t I
xl = Allxl + A2lx2 + A31x3 +
• I • x2 = Al2xl + AzzX2 + A32x3 +
• t 1
XJ = AlJxl + A23x2 + A33x3 +
• where the Aij s are constants of
Bl
Bz (2-6)
B3
rotation between the
axes of the two coordinate systems and the Bi's are
constants of translation between the origins of the
system.
Expression like Eq. (2-6) may be expressed more
concisely by adopting the following notational conven-
tions:
Range Convention:
When an index (su~script) occurs unrepeated in
a term of an expression, it is understood to
take, in turn, each value in the range of that
index. In this paper the values will always be
1, 2, J.
Summation Convention:
When an index is repeated in a term, summation
over the range of that index is implied. Using
these conventions, equation (2-6) is written
as, t
X. = A .• X. +B. ~ J~ J ~
(2-7)
and the inverse transformation is written as,
t I
X. = A .. X. + B. ~ ~J J ~
(2-8)
The range and summation conventions will be used
throughout this paper. No confusion should occur if
the reader remembers that they are implicitly present
in the notation henceforth.
13
With the orthogonal property of the coordinate
systems, the rotational coefficients have the relation
where
0 if i f j = A .A = 6 •. = [
kl kj lJ 1 if i = j
6 .. is the so-called Kronecker delta. lJ
(2-9)
In the paper the orthogonal transformation is always
positive; i.e., IAijl = 1.
b. Definition of a Tensor
14
If each number, Ti , of a set of quantities associated
with a Cartesian coordinate system, Xi , and with a point, • P, can be transformed to any other coordinate system Xi ,
according to the equation
• T. =A .• T. J Jl l
(2-10)
then is said to be a component of a tensor of the first
rank.
It will be seen that a Cartesian tensor of the first
rank is equivalent to an ordinary Cartesian vector.
Tensor of higher rank can also be defined. That is,
quantities Tab which transform according to
are called tensors of the second rank. A tensor of the
nth rank has n indices and transforms through its multi
plication by n coefficients:
' Tij •••• k = Ai~jb•••••••AkcTab •••• c
c. Some Tensor Properties
A tensor is said to be symmetric in two indices,
j and k, if the value of any component is not changed
by interchanging the positions of j and k. That is,
if
T. 'k 1e•••J •••• m = T. . 1 •••• kJ •••• m
then the tensor is symmetric in j and k. The tensor
is completely symmetric if its components retain the
same value when any two indices are interchanged.
Similarly, a tensor is said to be screw-symmetric in
tho indices j and k, if
s. 'k 1 •••• J •••• m = - s. . 1. • • .kJe ••• m
The tensor is completely screw-symmetric if its
components retain the same value, but are changed in
sign when any two indices are interchanged.
say
say
The product of any completely symmetric tensor,
T .. , and any completely screw-symmetric tensor, 1J
S. . , is 1J
T •. S •. = 0 1J 1J
This property will be used to great advantage in the
application of tensor operations discussed in this
paper.
15
(2-12)
(2-13)
(2-14)
(2-15)
The Kronecker delta defined in equation (2-9)
is an example of a symmetric tensor. It is also
called an isotropic tensor because its components
retain the same values in any coordinate system.
Another tensor which will be very useful in
the present application is the permutational symbol,
y .. k • This tensor is both isotropic and completley lJ . screw-symmetric; the values of its components are
obtained as follows;
0 if any two indices have the same value.
1 if the values of the indices ijk repre-
Yijk= sent an even permutation of the sequence
1,2,3.
-1 if the values of the indices ijk represent
an odd permutation of the sequence 1,2,3.
An important relation between o .. and y .. k is lJ lJ given by
(2-16)
16
17
CHAPTER III
POSITION ANALYSIS OF THE SPATIAL FOUR-BAR LINKAGE
A. General Description
In order to demonstrate the application of the
tensor method to the analysis of spatial mechanisms,
a closed-loop, four-bar spatial linkage has been studied.
Figure 1 represents the general diagram of the linkage
and relationship of the links to the coordinate frames.
It is convenient that each link can be determined with
respect to its local coordinate frame qy a set of spheri-
cal polar coordinates. The so-called ground link, Cc., 1.
has a length of C units and is directed along a vector
of unit length, c .• 1.
The components of ci with respect
to the ground coordinate frame are defined by
cl = sin% cos a
c2 = sin% sine (3-1)
c3 = cos%
Where }6 and e represent the polar and azimuthal angles,
respectively.
Likewise, the first link, Rri1 ' has magnitude R
and is directed along a unit vector ri1 which origi
nates at the first joint. The component of ri1 with
respect to its local coordinate frame are defined by
r 1I = sin%I cose 1
r 2I = sin%I sinei
r 3I = cos%I
(3-2)
where %I and e1 represent the polar and azimuthal
angles of the link in the first frame. In the same
manner, the components of the second and third link
vectors, s 1II and t 1II can be expressed in term of the
polar coordinates of the second and the third frames,
respectively.
s II 1 = sin,0II coseii
s II 2 = sin,0'II sine II (3-3)
s II 3 = cos,0II
t III 1 = sin,0III coseiii
t III 2 = sin,0III sine III (3-4)
t III 3 = cos_eiii
To derive relationships between the link vectors,
it is desirable to specify an arbitrary point to which
all the vectors may be referredo Through this paper
the ground joint will serve as the reference point.
The transformation matrix Ai~ can be written as the
19
20
direction cosines between the axes; that is
[All Al2 Al3 A •• = A21 A22 A23 1J
A31 A32 A33
0 0 0
= [
cos(X1 ,x1 ) cos(X1 ,x2 ) cos(X1 ,x3 )
cos(x2 ,xi) cos(x2 ,x;) cos(X2 ,xj) (3-5)
cos(x3 ,xi) cos(x3 ,x~) cos(x3 ,xj)
Bo Coordinate Relationship
Whenever one of the axes of the prime reference frame
is specified, the prime frame system is conveniently se-
lected in following way, as shown in Figure 2. xo
3
~~-----------+------~----~~x2 I //
-----4/
xo 2
Figure 2o Coordinate Relationship
If xj is the specified axis of the prime frame
system in xl ' x2 • x3 , reference system with given
polar and azimuthal angles ( ~o , 90 ) then it can be
expressed as
(3-6)
The other two prime reference axes are defined by
(X~)i = Yik3<x;)kx3 = Yik3wk
<x;)i = Yij~<x;)j(X~)k = Yijkwj Ykl3wl
= (oil oj3 - oi3 ojl) wjwl
= - 0i3 + w3wi
(3-7)
(3-8)
Equations (3-6) and (3-8) can be expanded and formulated
to be the unit tensor that
(3-9)
(3-10)
From (3-6), (3-9), and (3-10), the relation of
transformation can be readily got as
(3-11)
where
[ sine, -coseo 0
Amn = cose. cos%. sine. cos%0 -sin%0
sin%o cose. sin%. sine. cos% 0
( 3-12)
21
CHAPTER IV
METHOD OF APPROACH AND DERIVATION
A, Governing Equation of R-S-S-R Mechanism
Figure 3 is a diagram of the R-S-8-R four-bar
mechanism with revolute pairs on the ground and first
joints, and with spherical pairs on the second and
third joints.
In this mechanism. a single closed-loop vector
equation has been used:
Cc. + Rr. + Ss. + Tt. = 0 2 2 2 1 (4-1)
Components of unit vectors c. , r. , s. , and t. 1 2 1 1
are all taken with respect to the ground frame by the
transformations
r. = AI I 2 mi rm
s. = AI~ sii (4-2) 2 m2 m t. = 2
AI~Itiii m2 m
I II III where A., A ., A . , are the transformation matrices m1 m1 m2
relating the ground frame and the first, second, and
the third frames, respectively. Equation (4-1) actu
ally represents three individual equations (for i = 1,
2, 3).
B. Derivation of the output follower angle e11I, and the
polar, azimuthal angle of the second link e11 , %II.
22
/
xi 1
FOLLOWER
/
//
//
/ //
Rr. / Cc. l. / l.
/ /
,.., ... xy/
~
Figure 3, R-S-S-R four-bar mechanism
Xz
!\) \,.)
The ground link is Cc.; its magnitude and direction ~
are always given. In usual, the magnitude and direction
of the ground frame are chosen to designate the relative
positions of the other links, input angle ei, the polar
angle %I, and the magnitude R. AI. and r~ are known, m~ ~
magnitude S, polar angle %III between the ground frame
and ti' and the magnitude T are given.
In this paper, the auther wants to use the given
informations to find and analyze the components of the
unit vector si with respect to the ground frame; and
24
the polar and azimuthal angles of the link Ssi' %II and
eii with respect to the ground frame; output azimuthal
angle eiii, anglar velocity aiii and angular acceleration
eiii with respect to the ground frame.
From Eqs. (4-1) and (4-2), the unknown quantities
can be determined by solving
where
t. = ~
Letting K. = Cc. + RAI~rmi ~ ~ m ....
any given input crank angle ei,
expanded as:
(4-3)
a constant vector for
equation (4-3) can be
25
Kl + s sin,0'11cose 11 + T sin}?JIIIcoseiii = 0 (4-4a)
Kz + s sin,011sine 11 + T sin,0IIIsineiii = 0 (4-4b)
K3 + s cos%II + T cos%III = 0 (4-4c)
Equations (4-4) are the general scheme for analyzing
the R-S-S-R mechanism. In the R-S-S-R mechanism shown in
Figure 3, the axis x3 of the ground frame is the axis of
rotation of the follower. The angle, %III, between the
x3 axis and the follower T is a right angle.
For ,0III= 1T/2 , equations (4-4) now become
Kl + s sin,0I1 cosa 1I + T cose 111 = 0 (4-5a)
K2 + s sin,0I1sine 11 + T sinaiii = 0 (4-5b)
K3 + s cos,0II = 0 (4-5c)
The solution of equation (4-5c) is
,0 = arc cos( -K 3 /S )
= arc sin ( J S 2 - K ~ /S ) (4-6)
substituting Eq. (4-6) into Eqs, (4-5a) and (4-5b)
Kl + Jsz_ K2 + )sz_
(4-?a)
( 4-Tb)
From Eqs. (4-?a), and (4-?b), it becomes clear
that the o·utput follower angle e111 has two possible
solutionso This confirms the statement by Harrisberger
(8) that the R-S-S-R mechanism has two degrees of
freedom.
26
Rewrite equations (4-?a) and (4-7b) as
Js2 K2 - 3 cose 11 = - T cosa 111 - Kl (4-8a)
js2 K2 - 3 sine 11 = - T sina 111 - K2 (4-8b)
square (4-8a), and (4-8b) and add them together
(s2- K~)(cos2 a 11+ sin2e11 ) = T2cos2e111+ T2sin2e111
2 2 + K1 + K2 + 2K1T cosa 111+ 2K T sine111 (4-9) 2
s2- K~ = T2 + Ki + K~ + 2T(K cose 111+ K sine111 ) 1 2
K1cose 111+ K2sina 111= l/2T (S 2- T2- Ki- K~- K~) 2 2 2 2 2 letting B = l/2T (S - T - K1- K2- K3 )
(4-10)
(4-11)
(4-12)
where B is another constant vector for any given crank
angle e1 , equation (4-11) can be changed in form as
(4-13)
From equation (4-13), sina111 and cose 111 can be
solved easily as follow:
sine 111 = ----~--~--~---=-------BK2 ± K1JKi + K~ - B2
K2 + K2 1 2
(4-14)
BKl + K2JKi + K~ - B2
K2 + K2 1 2
cosa 111 = ----~--~--~~-=------- (4-15)
cose 11 and sine11 can be obtained by substituting the
value of sine 111 , cose 111 , and equation (4-6) into
equations (4-5a), (4-5b).
27
) (4-16)
) ( 4-17)
c. Output Follower Angular Velocity,
For K. = Cc. + Rr. ~ ~ ~
where r. = A. r , therefore ~ ~m m
sine. cose. cos%o cose. sin.f4 cosei
r. = -cos eo sineo cos%o sin,0. sine sine 1 ~ 0 0
0 -sin%. cos%. 0
and the derivative of r 1 , r 2 , and r 3 with respect to
time are
• ( . . I cos90 cos%0 cose I) •I rl = -s~neo s~ne + e
• ( cose. sinei sineo cosfd.cosei) er (4-18) r2 = + 0
( -s in%o cose I) •I r3 = e
and Kl = Cc1 + Rr1
K2 = cc2 + Rr2 (4-19)
K3 = cc3 + Rr3
R, C, cl' c2 • and c3 are all constants, so the time
derivative of K. ~
(Eq. 4-19) is
28
• • K.= Rr. J. J.
• • K1 = Rr1 • • K? = Rr2 (4-20)
"--
• • K3 = Rr3
From equation (4-11)
K1 coseiii+ K2 sineiii= 1/2T (S2- T2- Ki - K~ - K~)
In this equation, T and S are constants, so the
time derivative is:
(4-21)
• • • • III • III e I I I = _-_l_/_T_(K......;l;;;;...K...;;l;;;..+_K_2_K....;.2_+K~JK::-:'!J::-)--_K..-1.._c_o_s-:e ==---K..;..;2_s_i_n_e --so.
where
-K1sin9III+ K2coseiii
[ sineiii == (BK2 ± K1 jKi + K~- B2 )/(Ki + K~)
coseiii (BK1 + K2 )Ki + K~ - B2 )/(Ki + K~)
(4-22)
D. ••III Output Angular Acceleration, 9 •
In order to find the output angular acceleration, •• •• we have to calculate the K. and r. by taking an addi-~ ~
• • tional derivative from Ki and ri. Therefore,
•• R"" K. = r. l ~
[ K1
•• = Rr1 •• .. K2 = Rr2 •• •• KJ = Rr 3
(4-23)
•• I I ••II I r 1=(-sine0 sine +cos90 cos.¢ocose )e +(-sin9.cose-
cose., cos%" sinei)(ei) 2
•• I d. I .. I I r 2=(cos90 sine +sin90 cos;u0 cose )e +(cose., cose -
sineo cos%o sinai) (ei )2
r 3=( -sin%., sine I )ei+( sin,0'0 sine I)( a I )2 (4-24)
The 9III can be obtained the same way as before by
taking another derivative to eii (Eq. 4-22) •
where
.. III •2 "2 •2 •• .. .,. III 9 = -1/T (K1+K2+KJ+K1K1+K2K2+K3K3 )+(K1cos9 +
K2sine 11I)(eiii) 2+(2K1sine 111-2K2cose 111 )
(eiii)-K1cose 111-R2sineiii /(-K1sineiii+
K2cose 11I) (4-25)
[ Kl = cc1 + Rr1
Kz = Cc2 + Rr2
KJ = cc3 + Rr3
29
• I I • I
[
~1= R(-sine. sine + cosa. cos,0'0 cosa ) a
K2:::: R(cose. sinai + sine. cos,0'. cosai) ai
KJ= R(-sin,ef. cosai) ai
K1 = R (-sine. sine I+ cos a. cos%. cos a I )a I+ R ( -s ina. cos a I
cos90 cos,0'. sine I)( e I ) 2
K2== R(cosa. sinai+ sine. cos%. cosa 1 )ei+ R(cose. cosa 1-
s ina. cos,0'. sinaI)( a I) 2
K3== R(-sin,ef. sine 1 )ai+ (sin9f. sine 1 )(ai) 2
[ sine 111
cose 111 =
A specific example of this analysis has been
demonstrated in Appendix B.
30
CHAPTER V
CONCLUSION
The work in this paper is basically an extensional
analysis of the work of Ho (10) who developed the tensor
analysis of spatial mechanism.
Tensor notation offers a convenient and compact
means for expressing relationship in spatial mechanismo
but other mathematical methods such as matrices, vectors,
etc, still have merit in the analysis of spatial mechan
ism problems. Futher, the tensor transformation used
in spatial mechanism have great advantages in relieving
the burdens of the tedious and confusing reference of
the coordinate frame.
This paper has shown that tensor notation combined
with matrix and vector methods provide a suitable
addition· to the existing methods of exploring the spatial
domain of linkage, and also lends itself well to pro
gramming for computer solution of problems in spatial
kinematics.
In practical use, the R-S-S-R mechanism offers
merit over Hook's joints and bevel gears for the purpose
of indirect transmission, because the transmission
torque is not dependent on the rotating angle as Hook's
joints and is evenly distributed on the connecting
links instead of overloading a single gear tooth pair
of the bevel gears.
31
BIBLIOGRAPHY
(1) Altman, F. G.
a. "Uber Rambliche Sechsgliedrige Koppelgetriebe"
z. VDI., Vol. 96, March 11, 1954, pp. 245-249.
b. "Raumgetriebe", Feinwerktechnik, Vol. 60, No. 3,
1956, pp. 83-92.
( 2) Beyer, R.
a. "Kinematische Getriebesynthese .. , Springer-Verlag,
Berlin, Germany, 1953. pp. 189-191.
b. "Space Mechanisms", Transaction of the Fifth
Conference on Mechanisms, Purdue University,
w. Lafayette, Ind., 1958, pp. 141-163.
(3) Hain, K.
"Das Spektrum der Gelenkvierechs bei Veraender
licher Gestel1-Laenge", Forch Ing.-Wes., Vol 30,
No. 2, 1964, pp. 33-42.
(4) Keler, M. L.
"Analysis and Synthesis of Spatial Crank Linkages
32
by means of Spatial Line Geometry and Dual Values",
Forsching auf dem Gebiete des Ingenieurwegens,
Vol. 25. 1959, PP• 26-32, 55-63.
(5) Kislisyn, s. G.
"Use of the Method of Tensors in the Theory of
Spatial Mechanisms", Trudii Sem. Theorii Mash.
Mekh., Vol. 14, 1954, pp. 51.
(6) Hartenberg, R. s., and Denavit, J.
a. "Approximate Synthesis of Spatial Linkage",
ASME paper A-24, 1959, No. 59. Kinematic Syn
thesis of Linkages. McGraw-Hill Book Co,,
New York, 1964, pp. 343.
b. "A Kinematic Notation for Lower-Pair Mechanisms
Based on Matrices", ASME Trans., Journ, Appl.
Mech. 22, 215 (1955).
(7) Chase, M. A.
"Solutions to the Vector.Tetrahedron Equation",
ASME Trans., Journ. Eng. for Ind., 87, 228
(1963).
(8) Harrisberg, L.
"Mobility Analysis of Three-Dimensional Four
Link Mechanisms", ASME Conference Paper 64-
WA/MD-16 (1965).
(9) Yang, A. T., and Feudenstein, F.
"Application of Dual-Number Quaternion Algebra
to the Analysis of Spatial Mechanisms", Journ.
of Applied Mechanics, Vol. 31, ASME Trans.,
Vol. 86, Series E, June, 1964, pp. 300-308.
33
(10) Ho, C. Y.
"Tensor Analysis of Spatial Mechanisms", IBM
Journ. of Research and Development, Vol. 10,
No. 3, May, 1966.
(11) Raven, F. H.
"Position, Velocity, and Acceleration Analysis
Synthesis of Plane and Spatial Mechanism",
Ph. D. Thesis, Connell University, pp. 59,
1956.
(12) Uicker, J. J., Jr., Denavit, J., and Hartenberg,
R. s.
"An Iterative Method for the Displacement Analysis
of Spatial Mechanism", Journ. of Applied Mechanics,
Vol. 31, Trans. ASME, Vol, 86, Series E. June,
1964, PP• 309-314.
(13) Skreiner, M.
"Acceleration Analysis of Spatial Linkage Using
Axodes and the Instantaneous Screw Axis", Journ,
of Eng. for Ind., ASME Trans., Vol, 89, Series
B. 1967, PP• 97-101.
(14) Greubler, M.
"Civilingenieur", Vol. 29, 1883, PP• 187 and
"Encyklopadie der Mathematischen Wissenschaften",
Vol. IV, 1901, PP• 127.
NAMES OF FORTRAN VARIABLES
The following is a list of the important variables
of the program and what they represent.
1) AA's
2) BB's
3) CA's
4) CITO
5) CIT
6) CI2
7) CI)
8) CI4
9) DCIT
10) DCIJ
11) DDCIJ
components of the direction vector of
the driving link with respect to the
input coordinate system.
matrix of transformation for the input
coordinate system.
components of the direction vector of
the ground link.
azimuthal angle of the unit vector of
the input shaft, eo. I input angle, e , or azimuthal angle of
the driving link with respect to the
input coordinate system.
azimuthal angle of the second link,
9II.
azimuthal angle of the direction vector
of the driven link, aiii.
azimuthal angle of the ground link, e1v. angular velocity of the driving link, ·I a •
angular velocity of the output link, •III a • angular acceleration of the output link,
••III e •
37
12) DDCIT
13) GA's
14) PAO
15) PA2
16) PA4
17) c 18) R
19) s 20) T
21) DGA(I)
22) DDGA(I)
23) RA's
angular acceleration of the driving link ••I or the input link, a •
components of the direction vector of
K. • ~
polar angle of the input shaft, ~o•
polar angle of the second link, ~rr.
polar angle of the ground link, ~rv.
length of the ground link.
length of the input crank.
length of the second link.
length of the output link, or the
follower. • the first time derivative of GA, or Ki.
the second time derivative of GA, or •• Ki •
components of the direction vector of
the driving link with respect to the
ground coordinate system.
38
FORTRAN IV G LEVEL 1, MOD 2 MAIN DATE = 69092 19/15/55
0001
0002 0003 0004 0005 0006 0007 0008 0009 0010 0011 0012 0013 0014 0015 0016 0017 0018 0019 0020 0021 0022 0023 0024 0025 0026 0027 0028 0029 0030 0031 0032 0033 00.34 0035
c C THE ANALYSIS OF R-S-S-R SPATIAL FOUR-BAR MECHANISM c
DIMENSION AA(3),BB(3,3),CA(3),GA(361), RA(36l),X(36l),CI3(36l), 1CI2(361),PA2(36l),DCI3(361),DDCI3(.361) C=l. R=0.2 S=1,6 T=1.23 DEG=3.141593/l80. PA0=60.*DEG CI0=60.*DEG PA4=30. *DEG CI4=75.*DEG CIT0=60.*DEG DCIT=1. DDCIT=O. SICO=SIN ( CIO) COCO=COS(CIO) SIPAO=SIN (PAO) CO PAO=COS ( PAO) BB(1,1)=SIN(CITO) BB(1,2)=-COS(CITO) BB(1,3)=0. BB(2,1)=COS(CITO)*COS(PAO) BB(2,2)=SIN(CITO)*COS(PAO) BB(2g3)=-SIN(PAO) BB(3,1)=COS(CITO)*SIN(PAO) BB(3,2)=SIN(CITO)*SIN(PAO) BB(3,3)=COS(PAO) CA(1)=SIN(PA4)*COS(CI4) CA(2)=SIN(PA4)*SIN(CI4) CA ( 3) =COS ('PA4) DO 3 1=1,361 X(I)=1.*(I-1) CIT=(I-1)*1.*DEG AA(1)=COS(CIT) AA ( 2 ) =SIN (CIT) AA(.3)=0o
w '-.()
0036 0037 0038 0039 0040 0041 0042 0043 0044 0045 0046 0047 0048 0049 00.50 0051 00.52 0053 00.54 005.5 0056 0057 0058 00.59 oo6o 0061 0062 0063 0064 006.5 0066 0067 0068 0069 0070 0071 0072 0073
DO 2 J=1,3 SM=O. DO 1 K=1,3
1 SM=SM+AA(K)*BB(K,J) RA(J)=SM
2 GA(J)=C*CA(J)+R*RA(J) B=(S**2-T**2-GA(1)**2-GA(2)**2-GA(3)**2)/(2.*T) D=GA(1)**2+GA(2)**2 E=SQRT(D-B**2) SIC3=(B*GA(2)+GA(1)*E)/D CIC3=(B*GA(1)-GA(2)*E)/D F=SQRT(S**2-GA(3)**2) SIC2=-(GA(2)+T*SIC3)/F CIC2=-(GA(1)+T*CIC3)/F PA2(I)=ARCOS(-GA(3)/S)/DEG DGA1=R*(-SICO*AA(2)+COCO*COPAO*AA(l))*DCIT DGA2=R*(COCO*AA(2)+SICO*COPAO*AA(1))*DCIT DGA3=-R*(SIPAO*AA(1))*DCIT TN2=SIC2/CIC2 CI2(I)=ATAN(ABS(TN2))/DEG IF(CIC2) 11, 16~ 14
11 IF(SIC2) 13, 16, 12 12 CI2(!)=180,-CI2(I)
GO TO 16 13 CI2(I)=180,+CI2(I}
GO TO 16 . 14 IF(SIC2) 15, 16, 16 15 CI2(I)=360.-CI2(I)
RUTH·=ABS ( (B*GA ( 1 )-GA (2 )*E)/D) 16 CI3(I)=(ARCOS(RUTH))/DEG
IF(CIC3) 21, 26, 24 21 IF(SIC3) 23, 26, 22 22 CI3(I)=180.-CI3(I)
GO TO 26 23 CI3(I)=180,+CI3(I)
GO TO 26 24 IF(SIC3) 2.5, 26, 26 25 CI3(I)=360,-CI3(I) .{:::"
0
0074
0075 0076 0077 0078
0079 0080
0081 0082 0083 0084 0085 0086 0087 0088
26 DCI3(I)=(DGA1*CIC3+DGA2*SIC3+(GA(1)*DGA1+GA(2)*DGA2+GA(3)*DGAJ) 1/T)/(GA(1)*SICJ-GA(2)*CICJ)
DDGA1=DGA1*DDCIT/DCIT-R*(SICO*AA(1)+COCO*COPAO*AA(2))*DCIT**2 DDGA2=DGA2*DDCIT/DCIT+R*(COCO*AA(1)-SICO*COPAO*AA(2))*DCIT**2 DDGA3=DGA3DDCIT/DCIT+R*(SIPAO*AA(2))*DCIT**2 DDCI3(I)=(-(DGA1**2+DGA2**2+DGA3**2+GA(l)*DDGAl+GA(2)*DDGA2+GA(3)
1*DDGA3)/T+(GA(l)*CIC3+GA(2)*SICJ)*DCI3(I)**2+2.*(DGA1*SIC3-DGA2* 1CICJ)*DCI3(I)-DDGA1*CICJ-DDGA2*SICJ)/(-GA(l)*SIC3+GA(2)*CIC3)
3 WRITE(3,102) X(I),CI2(I),CI3(I),DCIJ(I),DDCI3(I),PA2(I) 101 FORMAT(5X,'INPUT ANGLE',4X,'CIT2',5X,'OUTPUT ANGLE',JX,'OUTPUT AN
lGVEL',4X,'OUTPUT ANGACC',4X,'PA2',////) 102 FORMAT (3X, F5,1,5(5X,F8.4))
CALL PPLT(X,CI2,361) CALL PPLT(X,PA2,36l) CALL PPLT(X,CI3,361) CALL PPLT(X,DCI3,36l) CALL PPLT(X,DDCI3,361) SWP END
.{::" 1-'
APPENDIX B
FIGURES OF SOME VARIABLES OF R-S-S-R
SPATIAL MECHANISM
For a specific R-S-S-R spatial four-bar mechanism
with a constant input angular velocity, ai=l.O rad/sec,
and the known properties are chosen as,
a) length of the ground link, C=l.o~.
b) length of the crank, R=0.2".
c) length of the second link, R=l.6".
d) length of the follower, T=l.2}".
e) azimuthal angle of the ground link, eiV=75°,
) 1 1 t . aiV o f po ar ang e of he ground l1nk, ? =30 • ..
g) azimuthal angle of the reference frame, 9 .. =60 •
h) polar angle of the reference frame, % .. =60".
The unknown quantities,
a) azimuthal angle of the second link, aii.
b) polar angle of the second link, ¢1I.
c) azimuthal angle of the follower, a11I.
d) azimuthal angular velocity of the follower, alii.
42
e) azimuthal angular acceleration of the follower, aiii,
can be determined by Eqs. (4-14), (4-15), (4-16), (4-22),
and (4-25).
These unknown quantities are plotted versus the
input angle, ai (from o" to 360" for each cycle), as
follows:
349 CONSTANT INPUT ANGULAR VELOCITY = 1,0 (rad/sec)
-~ ~
ffi ~ Q
JJ8 ...........
H H
(I)
-~ z ~ 328 Q z 0 0 ~ t:ll
~
~ 317 lit 0
f:il
~ ~ 307 ...:;~ <(
g:; :::> :;E H lSI <(
296 108 144 . INPUT ANGLE (DEGREE)
FIGURE 4. THE AZIMUTHAL ANGLE OF THE SECOND LINK (a~~) VERSUS INPUT ANGLE l9-J ~ \......) -
-~ ~
ffi ~ Q ..._,
H
~ • ~ z H ~
Q z 0 u ll:l Cll
~
~ ll:t 0
~ z <
~ &'!
CONSTANT INPUT ANGULAR VELOCITY = 1.0 (rad/sec) 131 ~
128
119
INPUT ANGLE (DEGREE) FIGURE 5. _THE POLAR ANGLE OF THE SECOND LINK (~II) VERSUS INPUT ANGLE (ei) ~
-«="
CONSTANT INPUT ANGULAR VELOCITY = 1,0 (rad/sec) 192r-----------------------------~~~------------------
......... ~ 181 t§ f:il Q ........
H H H
(t) 170 • ~ eJ
~ 8 ~ 159 8 ::> 0
f:il
~ 149
144 180 INPUT ANGLE (DEGREE)
FIGURE 6. THE OUTPUT ANGLE (alii) VERSUS INPUT ANGLE (e 1 ) .{:::"" \J\
0.395
.......... 0
~ 0.193 .......,_ 'd ro f...l -
H H ':!:n -0.00923
~ H 0
s ~-0.311
ll::
~ 5 ~ 8 -0.413 ~ E-1 :::> 0
CONSTANT INPUT ANGULAR VELOCITY = 1.0 (rad/sec)
-0.616 I I I I J I I 0 16 72 108 144 180 21
INPUT ANGLE (DEGREE) •• FIGURE 7. THE OUTPUT ANGULAR VELOCITY ( eiii) VERSUS INPUT ANGLE (ei) -{::"
0\
CONSTANT INPUT ANGULAR VELOCITY = 1.0 (rad/sec) o. 5441 ________________ __:._:._~-----::; ........ ;.:--"l
......... N
('.) Q)
~ "d 0.30 ro S... .._.
H H H :<t> .. z 0.07) 0 H 8 <I! ~ r:q H r:q 0 ~ -0.162
~ 8 ~ E-1 -0.398 fi! E-1 ::> 0
-0.633~----~----~----~~--~~--~~----~----~~--~----~----~ 0 16 72 108 144 180 216 252 288 324 360
INPUT ANGLE (DEGREE) FIGURE 8. THE-OUTPUT ANGULAR ACCELERATION (alii) VERSUS INPUT ANGLE (a 1 )
-t::--..J
VITA
David Perng Chyi was born in Shenyang, China, on
July 9, 1944o His high school education was received
at Taiwan, Republic of China. Later, he attended
National Taiwan University, and received the Bachelor
of Science in Mechanical Engineering in June 1966. He
served in the Chinese Army as a 2nd Lt. for one year
immediately after graduation from college.
In September, 1967, he enrolled in the graduate
school of the University of Missouri - Rolla, Rolla,
Missouri, to pursue the degree of Master of Science
in Mechanical Engineeringo
48
1S3~94