This action might not be possible to undo. Are you sure you want to continue?
th
National Conference on Mechanisms and Machines (NaCoMM07),
IISc, Bangalore, India, December 1213, 2007 NaCoMM2007130
221
Instantaneous Mobility and Constraint of a 3PPS Parallel
Mechanism
JingShan Zhao
*
, Fulei Chu, LiPing Wang and ZhiJing Feng
Department of Precision Instruments and Mechanology, Tsinghua University, Beijing 100084, P. R.
China.
*
Corresponding author (email: jingshanzhao@mail.tsinghua.edu.cn.)
Abstract
This paper investigates the mobility and constraint of a
spatial parallel mechanism whose endeffector has three
degrees of freedom. The novelty of the mechanism is that
its endeffector should output two rotations and one trans
lation with 3PPS kinematic chains. In this paper, the de
gree of freedom of an endeffector of a mechanism and the
number of actuations required to control the endeffector
are defined, individually. And then, the exponential prod
uct procedure to analyze the twists and wrenches of a ki
nematic chain is deduced in one fixed Cartesian coordinate
system. The mobility and constraint of the endeffector are
addressed from the viewpoint of instantaneous free mo
tions and the number of actuations needed to control it.
Actuation scheme analysis indicates that the degree of
freedom of an endeffector and the number of independent
actuations needed to uniquely control the endeffector
should not be ambiguously represented by one concept
such as the general mobility of a mechanism.
Keywords: Mobility, Constraint, Spatial Mechanism, End
effector, and Kinematic Chain
1 Introduction
The mobility and constraint of a mechanism are of primary
importance in the theory of mechanisms, and have been
extensively discussed either in analytical methods or in
quick calculation approaches based on algebra summations
of the number of the links, joints and the constraints in
duced by the joints in the previous literature during the
past 150 years [1]. This lengthy development process on
the one hand demonstrates the complexity and difficulty of
this problem, and on the other hand indicates that it is of
vital importance both in theory and in engineering applica
tions.
As to the development of the mobility of a mechanism,
Gogu [1] made a relatively detailed review. Therefore, this
paper does not intend to make a full development about the
history of mobility of a mechanism. Generally, the restric
tions of the method based on pure algebra summations of
the numbers of the links, joints and the constraints induced
by the joints have been discussed widely in [27]. Rico and
Ravani [6,7], Rico, Gallardo and Ravani [8], Rico and et al
[9] made contemporary investigations on the mobility
analysis of parallel mechanisms and kinematic chains with
Lie group and Lie algebra. In applications, it is very con
venient to investigate the constraint and mobility of a
mechanism with reciprocal screw theory. Phillips and Hunt
[10], Waldron [2], Hunt [3], Ohwovoriole and Roth [11],
Sugimoto and Duffy [12][13], Gibson and Hunt [14][15],
Phillips [4][16], McCarthy [17], Rico and Duffy [1820],
Tsai and Lee [21], Murray, Li and Sastry [22], Zhang and
Xu [23], Fang and Huang [24], Huang and Wang [25],
Bandyopadhyay and Ghosal [26] et al made great contribu
tions to the applications of screw theory in mechanisms
after Ball [27]. Traditionally, the mobility of a mechanism
is thought of as the number of independent parameters to
define the configuration of a mechanism.
However, with the advent of spatial parallel mecha
nisms, the primary considerations of the designers have
been focused on nothing but the mobility of the end
effector and its controllability. Consequently, it is urgently
necessary to discriminate between the degree of freedom
of an endeffector and the number of actuations to
uniquely control the endeffector under a configuration.
Therefore, this paper focuses on the two aspects of the
mobility of a mechanism. For convenience, it first intro
duces a pair of definitions:
Definition 1: The degree of freedom (DOF) of an end
effector totally characterizes the motions of the end
effector including the number, type and direction of the
independent motions.
Definition 2: The configuration degree of freedom
(CDOF) of a mechanism with a prescribed endeffector
indicates the number of actuations required to uniquely
control the endeffector under a configuration.
Obviously, the DOF of an endeffector in number is
not larger than 6 but the number of actuations required to
uniquely control the endeffector might be any nonnega
tive integer. Bearing the above two definitions in mind,
one can fall into two steps to investigate the mobility of a
mechanism—the DOF of the endeffector and the CDOF
of the mechanism with a specified endeffector. The for
mer definition indicates the full instantaneous mobility
properties of the endeffector while the later one tells us
the instantaneous controllability of the mechanism system.
13
th
National Conference on Mechanisms and Machines (NaCoMM07),
IISc, Bangalore, India, December 1213, 2007 NaCoMM2007130
222
2 Architecture of the Mechanism and
the Mobility and Constraint of Its Ki
nematic Chains
The mechanism proposed in this paper is shown in figure 1.
The endeffector E
1
E
2
E
3
of the mechanism has three iden
tical PPS kinematic chains P
i
A
i
E
i
(i=1,2,3) connected by
the fixed base B
1
B
2
B
3
. It has two rotational DOFs and one
translational DOF (2R1T). The subtended angles between
any two guides of the fixed base are 120º.
It is very convenient to express the kinematics of a
serial chain by exponential product of the revolute angles
of the joints in the chain [22]. Therefore, the free motions
of each PPS kinematic chain might be well depicted by an
exponential product formula. Consequently, this section
will investigate the mobility of a PPS kinematic chain with
reciprocal screw theory [35][16][27] based on exponential
product [22].
Figure 1: the Architecture of a Spatial Parallel Mechanism
with 2R1T EndEffector
A PPS kinematic chain of the endeffector E
1
E
2
E
3
,
P
1
A
1
E
1
for instance, and the fixed Cartesian coordinate
system are shown in figure 2. For the sake of building up
the mathematic model, a Cartesian coordinate system is set
here by letting xaxis be superimposed with the guide line
of the prismatic joint P
1
, yaxis locate in the base plane
B
1
B
2
B
3
, and zaxis be perpendicular to the base plane
B
1
B
2
B
3
and directing to the endeffector E
1
E
2
E
3
.
Figure 2: a Series PPS Kinematic Chain Connected with
the Base
Assume that ( ) ( ) { } 1 det , 3
3 3
= = ℜ ∈ =
×
R I RR R SO
T
denotes a 3 3× special orthogonal matrix group where I
denotes an identical matrix. Group ( ) 3 SO represents the
three dimensional rotations in a Cartesian coordinate space,
and therefore, it is also alternatively called configuration
space.
The exponential transformation for a point on a link ro
tated θ about a joint axis ω can be expressed with:

¹

\

=
∧
θ ω exp R (1)
where ∧ denotes a wedge operation such that the follow
ing equation holds for a vector ω:
(
(
(
¸
(
¸
−
−
−
=



¹

\

(
(
(
¸
(
¸
=
∧
∧
0
0
0
x y
x z
y z
z
y
x
ω ω
ω ω
ω ω
ω
ω
ω
ω (2)
and θ denotes the rotational angle about the axis ω.
Let ( ) { } S S S − = ℜ ∈ =
× T
so
3 3
3 , if ( ) 3 so ∈
∧
ω and
R ∈ θ , there must be ( ) 3 exp SO ∈

¹

\

∧
θ ω . It is not dif
ficult to prove that the exponential mapping from ( ) 3 so to
( ) 3 SO is surjective.
Define groups:
( ) ( )
)
`
¹
¹
´
¦
ℜ ∈ ∈

¹

\

=
∧ ∧
3
3 3 v , ω v , ω so se , and
( ) ( ) ( ) { } ( ) 3 3 3
3 3
SO SO SE × ℜ = ℜ ∈ ∈ = p , p , R R .
Obviously, the element ( ) ( ) 3 SE ∈ p , R can be used
to express the transformation from one coordinate system
to another.
Suppose the coordinates of a twist are denoted by
(
¸
(
¸
=
v
ω
ξ . Obviously, if a Vee operation, ∨ , and a wedge
operation, ∧ , are also defined such that
(
(
¸
(
¸
=
∧
∧
0 0
v ω
ξ
and ξ
v
ω
v ω
ξ =
(
¸
(
¸
=
(
(
¸
(
¸
=

¹

\

∨
∧
∨
∧
0 0
, then ( ) 3 SE e ∈
∧
θ ξ
.
If one assumes ( ) 0 g to represent the initial transforma
tion of a rigid body relative to the Cartesian coordinate
system, then the final configuration, still with respect to
the same Cartesian coordinate system, is given by:
( ) ( ) 0 exp g g

¹

\

=
∧
θ θ ξ (3)
Thus, the exponential map for a twist gives the relative
motion of a rigid body. Therefore, the exponential of a
twist is a mapping from initial configuration to the final
13
th
National Conference on Mechanisms and Machines (NaCoMM07),
IISc, Bangalore, India, December 1213, 2007 NaCoMM2007130
223
one. It can be proved that given ( ) 3 SE g ∈ , there exists
( ) 3 se ∈
∧
ξ and R ∈ θ such that

¹

\

=
∧
θ ξ exp g . The
forward kinematics map for a series kinematic chain is
given by:
( ) ( ) 0 exp
1
g g
i i
n
i

¹

\

Π =
∧
=
θ θ ξ (4)
where ( ) θ g represents the rigid motion of the end
effector connected by the series kinematic chain,
i
ξ repre
sents the ith associated twist in the chain, and
i
θ repre
sents the ith relative angle.
Equation (4) can be expanded to
( )
( ) ( )
(
¸
(
¸
=
1 0
θ θ
θ
p R
g (5)
The velocity of the endeffector can be specified by de
fining wedge and Vee operations in the absolute coordi
nate system:
(
(
¸
(
¸
+ −
= =
• • •
• ∧
0 0
1 p p
v
T T
R R R R
g g
£
(6)
(
(
(
¸
(
¸
+ −

¹

\

=
(
¸
(
¸
=

¹

\

=
• •
∨
•
∨
∧
p p
v
ω
v v
T
T
R R
R R
(7)
Therefore, with equations (4), (6) and (7), one obtains
the velocity of the endeffector. From equation (6), one
might find that:
∑ ∑
=
•
=
• • ∧


¹

\

∂
∂
=


¹

\

∂
∂
= =
n
i
i
i
n
i
i
i
g
g
g
g
g g
1
1 1
1
1
θ
θ
θ
θ
£ £ £
v
(8)
In twist coordinates, equation (8) can be written as:
•
= θ
t
J v (9)
where
(
(
¸
(
¸


¹

\

∂
∂


¹

\

∂
∂


¹

\

∂
∂
=
∨
−
∨
−
∨
− 1 1
2
1
1
g
g
g
g
g
g
n
t
θ θ θ
Λ J
and
t
J represents the Jacobian matrix of the kinematic
chain in the absolute Cartesian coordinate system, and
T
n
(
¸
(
¸
=
• • • •
θ θ θ Λ
2 1
θ and
•
θ indicates the magni
tude vector of general rotational velocities of the chain.
Each column of the Jacobian matrix
t
J represents the
twist coordinates of the corresponding joint in the Carte
sian coordinate system. Therefore, the twist coordinates of
the point attached to the rigid endeffector are given with
equation (8) or (9). Given a fixed Cartesian coordinate
system, one can directly write out the Jacobian matrix
t
J
with the above conclusion. For the mechanism shown in
figure 1, the Jacobian matrix of the kinematic chain P
1
A
1
E
1
,
denoted by
1
J , can be obtained in the fixed Cartesian co
ordinate system shown in figure 2 in accordance with the
following steps.
Step 1: write out the twist of the kinematic pair in the
chain, individually.
The twist of the prismatic pair
1
P can be directly ob
tained:
( )
T
0 0 1 0 0 0
1
= $ (10)
The twist of the prismatic pair
1
A can be directly ob
tained:
( )
T
1 0 0 0 0 0
2
= $ (11)
For the spherical kinematic pair
1
E , if its Cartesian
coordinates are denoted by ( )
T
E E E E
z y x
1 1 1
= r , the
twists for the three orthogonal axes can be expressed as:
(
¸
(
¸
×
=
i E
i
E
i
e r
e
$
1
(12)
where ) 3 , 2 , 1 ( = i
i
e indicates the three orthogonal direc
tions of the spherical joint
1
E .
Consequently, according to equation (12), the three
twists of the spherical joint can be expressed as:
( )
( )
( )
¦
¦
¹
¦
¦
´
¦
− =
− =
− =
T
E E E
T
E E E
T
E E E
x y
x z
y z
0 0 1 0
0 0 1 0
0 0 0 1
1 1 13
1 1 12
1 1 11
$
$
$
(13)
Step 2: write out the Jacobian matrix of the kinematic
chain by letting the twists of the chain be the columns one
by one.
For example, the Jacobian matrix of the kinematic
chain P
1
A
1
E
1
can be obtained from equations (10), (11)
and (13):
(
(
(
(
(
(
(
(
¸
(
¸
−
−
−
=
0 1 0
0 0 0
0 0 1
1 0 0 0 0
0 1 0 0 0
0 0 1 0 0
1 1
1 1
1 1
1
E E
E E
E E
x y
x z
y z
J (14)
The twist of the terminal point, E
1
, of the kinematic
chain P
1
A
1
E
1
is:
1 1
1
c v J =
E
(15)
where
1
c represents a vector whose elements are the rela
tive rotational or translational speeds of the corresponding
joints and
T
E E E A P
) ( c
13 12 11 1 1
1
ω ω ω ω ω = here.
For the sake of length of the paper, the exponential
product theory of screw is not fully developed here. Please
refer to [22] for the details. The reciprocal screw theory
indicates that the constraints exerted to the point attached
13
th
National Conference on Mechanisms and Machines (NaCoMM07),
IISc, Bangalore, India, December 1213, 2007 NaCoMM2007130
224
to the rigid endeffector are reciprocal to the twist of the
point. Therefore,
( ) 0 = =
•
θ τ v τ
t
T T
E E J (16)
where τ indicate the wrench coordinates in the absolute
coordinate system, the superscript T indicates the trans
pose of a vector or matrix, and
(
¸
(
¸
=
0
0
3
3
I
I
E and
(
(
(
¸
(
¸
=
1 0 0
0 1 0
0 0 1
3
I .
Because of the independency and arbitrariness of
•
θ ,
the necessary and sufficient criteria for equation (16) can
be simplified as:
( ) 0 = τ E
T
t
J (17)
Vice versa, the twist(s) of a rigid body can be obtained
with the following formula if the wrenches exerted to it are
known:
0 =
t
T
EJ τ (18)
Consequently, it is very convenient to investigate the in
stantaneous mobility and constraint of a series kinematic
chain with reciprocal screw theory.
With equation (17), one immediately obtains:
( )
T
E E
x z F
1 1
0 0 1 0
1 1
− = τ (19)
where
1
F indicates the magnitude of the wrench.
The physical meaning of
1
τ is that the terminal point,
E
1
, of kinematic chain P
1
A
1
E
1
is subjected to a constraint
force passing through it and with a direction of
( )
T
0 1 0 . In the like manner, one can get the twists
and wrenches of the points E
2
and E
3
in the same Cartesian
coordinate system shown in figure 2, respectively:
(
(
(
(
(
(
(
(
(
¸
(
¸
−
−
− −
=
0 1 0
0 0
2
3
0 0
2
1
1 0 0 0 0
0 1 0 0 0
0 0 1 0 0
2 2
2 2
2 2
2
E E
E E
E E
x y
x z
y z
J (20)
where ( )
2 2 2
E E E
z y x represent the Cartesian coordi
nates of the spherical joint
2
E , and
2 2
2
c v J =
E
(21)
where ( )
T
E E E A P
23 22 21 2 2
2
ω ω ω ω ω = c .
T
E E E E
y x z z F


¹

\

− − =
2 2 2 2
2
3
2
1
2
3
2
1
0
2
1
2
3
2 2
τ
(22)
where
2
F indicates the magnitude of the wrench.
(
(
(
(
(
(
(
(
(
¸
(
¸
−
− −
− −
=
0 1 0
0 0
2
3
0 0
2
1
1 0 0 0 0
0 1 0 0 0
0 0 1 0 0
3 3
3 3
3 3
3
E E
E E
E E
x y
x z
y z
J (23)
where ( )
3 3 3
E E E
z y x represent the Cartesian coordi
nates of the spherical joint
3
E , and
3 3
3
c v J =
E
(24)
where ( )
T
E E E A P
33 32 31 3 3
3
ω ω ω ω ω = c .
T
E E E E
y x z z F


¹

\

− − − =
3 3 3 3
2
3
2
1
2
3
2
1
0
2
1
2
3
3 3
τ
(25)
where
3
F indicates the magnitude of the wrench.
3 The Mobility of the EndEffector
and the Actuators Required to Control
It
The terminal constraints of the three kinematic chains have
been obtained in section 2, which are also the constraints
exerted to the endeffector E
1
E
2
E
3
by the three chains.
Therefore, the constraints applied to E
1
E
2
E
3
are directly
expressed by equations (19), (22) and (25):
(
(
(
¸
(
¸
=
3
2
1
F
F
F
J τ (26)
where
T
E E E E
E E E E
E E
y x z z
y x z z
x z
(
(
(
(
(
(
¸
(
¸
− − −
− −
−
=
3 3 3 3
2 2 2 2
1 1
2
3
2
1
2
3
2
1
0
2
1
2
3
2
3
2
1
2
3
2
1
0
2
1
2
3
0 0 1 0
J
.
With equation (16), one can also deduce that
( ) 0 = v E
T
J (27)
where v denote the twists of the endeffector.
Immediately, one obtains
K v
v
J = (28)
where
v
J denotes the kinematic Jacobian matrix of the
endeffector and
13
th
National Conference on Mechanisms and Machines (NaCoMM07),
IISc, Bangalore, India, December 1213, 2007 NaCoMM2007130
225
( )
( )
( ) [ ]
( )
( )
( )
( ) ( )
( )
( )
T
y y x x x
z z x
y y x x x
z y z x z x z y z x z x
y y x x x
z z
y y x x x
z x z x z x z x
z y z y
y y x x x
z x z x z x z x z x z x
z y z y z y z y
y y x x x
z z z
(
(
(
(
(
(
(
(
¸
(
¸
− + − −
− −
− + − −
− + − + + −
− + − −
−
− + − −
(
(
¸
(
¸
− − + +
−
− + − −
(
(
¸
(
¸
− − − + + +
− − +
− + − −
− −
=
0
3 2
3
3 2
3 3
3 2
3
1 0
0
3 2
3
3 2 3
3
3
3 2
2
0 1
1 0 0 0 0 0
3 2 3 2 1
3 2 1
3 2 3 2 1
3 2 3 2 3 1 2 3 2 3 2 1
3 2 3 2 1
3 2
3 2 3 2 1
1 3 1 2 2 1 3 1
1 3 1 2
3 2 3 2 1
1 2 2 3 3 1 1 3 3 2 2 1
3 2 2 3 1 2 1 3
3 2 3 2 1
3 2 1
v
J
(29)
[ ]
T
t t t
3 2 1
= K (30)
and ) 3 , 2 , 1 ( = i t
i
denote any real values.
Therefore, the instantaneous mobility of the end
effector is expressed by equation (28). For any prescribed
K , the endeffector has one determinate instantaneous
free motion represented with equation (28).
The rank of the twist matrix (29) represents the
number of degrees of freedom of the endeffector. What
must be noted is that equation (28) dynamically expresses
the instant motion of the endeffector and consequently,
the rank of the twist matrix (29) also dynamically ex
presses the instant DOF of the endeffector. In general
configuration, there is
( ) 3 =
v
J Rank (31)
It is not difficult to find from equation (28) that the
free motion of the endeffector E
1
E
2
E
3
is a proper screw
whose pitch is a nonzero value in general, which is shown
in figure 3.
Figure 3: the General Free Motion of the EndEffector
The direction of the instantaneous twist of the end
effector might be any vector but zaxis provided that
( ) 0 3 2
3 2 3 2 1
≠ − + − − y y x x x
. When
E
z z z z = = =
3 2 1
, the
kinematic Jacobian of the endeffector will be simplified
as:
T
E
E
z
z
(
(
(
¸
(
¸
−
=
0 0 0 1 0
0 0 0 0 1
1 0 0 0 0 0
v
J
(32)
The instantaneous free motion of the endeffector
E
1
E
2
E
3
might be represented by a twist:
( )
T
E E
t t z t z t t
1 2 3 3 2
0 − = v (33)
Equation (33) indicates that the endeffector E
1
E
2
E
3
might rotate about any axis within the plane
E
z z = pro
vided that
E
z z z z = = =
3 2 1
. Therefore, equation (28)
can be utilized to control the endeffector to output two
decoupled rotations perpendicular to zaxis and one trans
lation along zaxis.
After obtaining the mobility of the endeffector,
another question might arise——Can three actuators com
pletely control the endeffector E
1
E
2
E
3
? This is quite an
other problem.
As is proposed in [28], the controllability of the
endeffector should be discussed separately after obtaining
the instantaneous mobility of the endeffector. According
to the above analysis, three actuations must be given to the
mechanism. In the mechanism shown in figure 1, the fea
sible actuators might be assigned to any three of the 6
prismatic joints. Therefore, there are 20
3
6
= C possible
actuation schemes for the mechanism in total, which can
be grouped into 6 different schemes in essence. If the ac
tuation scheme is named after the joints assigned as actua
tors, these six different scheme can be individually de
noted by
3 2 1
P P P ,
3 2 1
A A A ,
) (
A P P
2 1 2 1
,
3 2 1
A P P ,
3 2 1
A A P ,
3 2 ) 3 ( 2
A A P actuation schemes. In the following section,
these six different actuation schemes will be investigated,
individually.
Scheme 1:
3 2 1
P P P Actuation Scheme
Assign 3 actuators to the prismatic joints P
1
, P
2
, P
3
.
If a set of 3 actuations are applied to these three
sliders, the twist matrixes of the three kinematic chains,
1 1 1
E A P ,
2 2 2
E A P and
3 3 3
E A P can be obtained from
equations (14), (20) and (23) by expelling the twist corre
sponding to the prismatic joints P
1
, P
2
, P
3
:
(
(
(
(
(
(
(
(
¸
(
¸
−
−
−
=
0 1
0 0
0 0
1 0 0 0
0 1 0 0
0 0 1 0
1 1
1 1
1 1
1
1
E E
E E
E E
S
x y
x z
y z
J (34)
(
(
(
(
(
(
(
(
¸
(
¸
−
−
−
=
0 1
0 0
0 0
1 0 0 0
0 1 0 0
0 0 1 0
2 2
2 2
2 2
1
2
E E
E E
E E
S
x y
x z
y z
J (35)
13
th
National Conference on Mechanisms and Machines (NaCoMM07),
IISc, Bangalore, India, December 1213, 2007 NaCoMM2007130
226
(
(
(
(
(
(
(
(
¸
(
¸
−
−
−
=
0 1
0 0
0 0
1 0 0 0
0 1 0 0
0 0 1 0
3 3
3 3
3 3
1
3
E E
E E
E E
S
x y
x z
y z
J (36)
The terminal constraints of these three actuated kine
matic chains can be similarly obtained according to equa
tion (17), respectively:
( )
( )
¦
¹
¦
´
¦
− =
− =
T
E E
T
E E
x z F
y z F
1 1
1 1
0 0 1 0
0 0 0 1
12 12
11 11
τ
τ
(37)
¦
¦
¦
¹
¦
¦
¦
´
¦


¹

\

− − =


¹

\

− − − =
T
E E E E
T
E E E E
y x z z F
y x z z F
2 2 2 2
2 2 2 2
2
3
2
1
2
3
2
1
0
2
1
2
3
2
1
2
3
2
1
2
3
0
2
3
2
1
22 22
21 21
τ
τ
(38)
¦
¦
¦
¹
¦
¦
¦
´
¦


¹

\

− − − =


¹

\

− − =
T
E E E E
T
E E E E
y x z z F
y x z z F
3 3 3 3
3 3 3 3
2
3
2
1
2
3
2
1
0
2
1
2
3
2
1
2
3
2
1
2
3
0
2
3
2
1
32 32
31 31
τ
τ
(39)
Therefore, the terminal constraints exerted to the end
effector can be obtained:
[ ]
32 31 22 21 12 11
τ τ τ τ τ τ τ = (40)
Substituting equation (40) into equation (18) yields:
( )
T
k 1 0 0 0 0 0 = v (41)
where k represents the magnitude of the twist and k
might be any real value.
Equation (41) indicates that the endeffector will
always have one free translation along zdirection under
the three actuations in scheme 1 under any configurations.
Therefore, the actuation scheme 1 can not safeguard the
controllability of the endeffector even the same number of
actuations as the DOFs of the endeffector are exerted to
the mechanism. So, for this actuation scheme, one more
actuation has to be applied to the mechanism. Suppose the
additional actuation is exerted to the prismatic joint A
1
, of
course, A
2
or A
3
is also available for this additional actua
tor. It is not difficult to find that under these 4 actuations
the mechanism shown in figure 1 is stable.
Alternatively, however, one can also select another
actuation scheme, namely,
Scheme 2:
3 2 1
A A A Actuation Scheme
Exert three actuations to the prismatic joints A
1
, A
2
,
and A
3
. For this actuation scheme, it is not difficult to find
that the endeffector does be stably controlled. And there
fore, the number of actuators needed to control the end
effector is 3 for this actuation scheme.
Scheme 3:
) 2 ( 1 2 1
A P P Actuation Scheme
Exert 2 actuations to the prismatic joints P
1
and P
2
,
and the third actuation is assigned to prismatic joint A
1
or
A
2
. For this actuation scheme, it is not difficult to find that
the endeffector does be stably controlled. And therefore,
the number of actuators needed to control the endeffector
is 3 for this actuation scheme.
With a similar analytical process, one can find that
among the rest three actuation schemes,
3 2 1
A P P ,
3 2 1
A A P
and
3 2 ) 3 ( 2
A A P , only
3 2 1
A A P and
3 2 ) 3 ( 2
A A P can control
the endeffector.
All in all, the mechanism proposed here demon
strates that the number of actuations required to control the
endeffector of a same specified mechanism might be dif
ferent if the actuation schemes are allowed to be different;
on the other hand, it also indicates that the mobility of the
endeffector and the actuations needed to control it should
be treated differently.
4 Conclusions
The mobility of a mechanism is very important in the
mechanism and machine theory. The DOF of an end
effector and the number of actuations required to control
the motion of the endeffector are different and should not
be depicted by one concept. Therefore, by proposing a
novel spatial parallel mechanism whose endeffector could
produce three free motions, including two decoupled rota
tions and one perpendicular translation, this paper investi
gates the analytical model to validly investigate the DOF
of an endeffector and the number of actuations required to
control the movement of the endeffector. With the actua
tion scheme analysis, it reveals that the concept of general
mobility cannot effectively represent the endeffector’s
free motions and controllability, and therefore the DOF of
an endeffector and the number of actuations required to
control it should be disposed separately.
Acknowledgment
This research was supported by FANEDD under Grant
200741 and the National Natural Science Foundation of
China under Grant 50425516. The authors gratefully ac
knowledge these support agencies.
References
[1] G. Gogu, “Mobility of Mechanisms: a Critical Re
view,” Mechanism and Machine Theory, Vol. 40, No. 9,
2005, pp. 10681097.
[2] K. J. Waldron, “The Constraint Analysis of Mecha
nisms,” Journal of Mechanisms, Vol. 1, 1966, pp. 101114.
[3] K. H. Hunt, Kinematic Geometry of Mechanisms, Ox
ford University Press, Oxford, 1978.
[4] J. Phillips, Freedom in Machinery, Vol. 1: Introducing
13
th
National Conference on Mechanisms and Machines (NaCoMM07),
IISc, Bangalore, India, December 1213, 2007 NaCoMM2007130
227
Screw Theory, Cambridge University Press, Cambridge,
1984.
[5] J. M. McCarthy, Geometric Design of Linkages,
SpringerVerlag, New York, 2000.
[6] J. M., Rico and B. Ravani, “On Mobility Analysis of
Linkage Using Group Theory,” ASME Journal of Me
chanical Design, Vol. 125, No.1, 2003, pp. 7080.
[7] J. M., Rico and B. Ravani, “On Calculating the Degrees
of Freedom or Mobility of Overconstrained Linkages: Sin
gleLoop Exceptional Linkages,” ASME Journal of Me
chanical Design, Vol. 129, No.3, 2007, pp. 301311.
[8] J. M. Rico, J. Gallardo, and B. Ravani, “Lie Algebra
and the Mobility of Kinematic Chains,” Journal of Robotic
System, Vol. 20, No. 8, 2003, pp. 477499.
[9] J. M. Rico, L. D. Aguilera, and et al, “A More General
Mobility Criterion for Parallel Platforms,” ASME Journal
of Mechanical Design, Vol. 128, No. 1, 2006, pp. 207219.
[10] J. R. Phillips and K. H. Hunt, “On the Theorem of
Three Axes in the Spatial Motion of Three Bodies”, Aus
tralian Journal of Exerted Science, Vol. 15, 1964, pp. 267
287.
[11] M. S. Ohwovoriole, and B. Roth, “An Extension of
Screw Theory,” ASME Journal of Mechanical Design, Vol.
103, No. 4, 1981, pp. 725735.
[12] K. Sugimoto, and J. Duffy, “Application of Linear
Algebra to Screw Systems,” Mechanism and Machine
Theory, Vol. 17, No. 1, 1982, pp. 7383.
[13] K. Sugimoto, and J. Duffy, “On Ball’s Mapping of the
Twosystem of Screws,” Mechanism and Machine Theory,
Vol. 21, No. 6, 1986, pp. 499507.
[14] C. G. Gibson and K. H. Hunt, “Geometry of Screw
Systems—1 Screws: Genesis and Geometry,” Mechanism
and Machine Theory, Vol. 25, No. 1, 1990, pp. 110.
[15] C. G. Gibson and K. H. Hunt, “Geometry of Screw
Systems—2 Classification of Screw Systems,” Mechanism
and Machine Theory, Vol. 25, No. 1, 1990, pp. 1127.
[16] J. Phillips, Freedom in Machinery, Vol. 2: Screw
Theory Exemplified, Cambridge University Press, Cam
bridge, 1990.
[17] J. M. McCarthy, An Introduction to Theoretical
Kinematics, MIT Press, Cambridge, 1990.
[18] J. M. Rico Martínez and J. Duffy, “Orthogonal Spaces
and Screw Systems,” Mechanism and Machine Theory,
Vol. 27, No. 4, 1992, pp. 451458.
[19] J. M. Rico Martínez and J. Duffy, “Classification of
Screw Systems—. One and TwoSystems,” Mechanism
and Machine Theory, Vol. 27, No. 4, 1992, pp. 459470.
[20] J. M. Rico Martínez and Joseph Duffy, “Classification
of Screw Systems—. ThreeSystems,” Mechanism and
Machine Theory, Vol. 27, No. 4, 1992, pp. 471490.
[21] M. J. Tsai and H. W. Lee, “On the Special Bases of
Two and ThreeScrew Systems,” ASME Journal of Me
chanical Design, Vol. 115, No. 3, 1993, pp. 540546.
[22] R. M. Murray, Z. Li and S. S. Sastry, A Mathematical
Introduction to Robotic Manipulation, CRC Press, Boca
Raton, 1994.
[23] W. X. Zhang and Z. C. Xu, “Algebraic Construction
of the ThreeSystem of Screws,” Mechanism and Machine
Theory, Vol. 33, No.7, 1998, pp. 925930.
[24] Y. Fang and Z. Huang, “Analytical Identification of
the Principal Screws of the Third Order Screw System,”
Mechanism and Machine Theory, Vol. 33, No. 7, 1998, pp.
987992.
[25] Z. Huang and J. Wang, “Identification of principal
screws of 3DOF parallel manipulators by quadric degen
eration,” Mechanism and Machine Theory, Vol. 36, No. 8,
2001, pp. 893911.
[26] S. Bandyopadhyay, and A. Ghosal, “Analytical De
termination of Principal Twists in Serial, Parallel and Hy
brid Manipulators Using Dual Vectors and Matrices,”
Mechanism and Machine Theory, Vol. 39, No. 12, 2004,
pp. 12891305.
[27] R. S. Ball, A Treatise on The Theory of Screws, Cam
bridge University Press, Cambridge, Initially Published in
1900 and Republished in 1998.
[28] J.S. Zhao, Z.J. Feng and J.X. Dong, “Computation
of the Configuration Degree of Freedom of a Spatial Paral
lel Mechanism by Using Reciprocal Screw Theory,”
Mechanism and Machine Theory, Vol. 41, No. 12, 2006,
pp. 14861504.
228
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.