Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Research ArticleRecognition of Kinematic Joints of 3D Assembly ModelsBased on Reciprocal Screw Theory
Tao Xiong1 Liping Chen1 Jianwan Ding1 Yizhong Wu1 and Wenjie Hou2
1National Engineering Research Center of CAD Software Information Technology School of Mechanical Science andEngineering Huazhong University of Science and Technology Wuhan 430074 China2Eaton (China) Investments Co Ltd Shanghai 200335 China
Correspondence should be addressed to Jianwan Ding dingjwhusteducn
Received 17 February 2016 Accepted 17 March 2016
Academic Editor Chaudry Masood Khalique
Copyright copy 2016 Tao Xiong et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited
Reciprocal screw theory is used to recognize the kinematic joints of assemblies restricted by arbitrary combinations of geometryconstraints Kinematic analysis is common for reaching a satisfactory design If a machine is large and the incidence of redesignfrequent is high then it becomes imperative to have fast analysis-redesign-reanalysis cycles This work addresses this problem byproviding recognition technology for converting a 3D assembly model into a kinematic joint model which is represented by agraph of parts with kinematic joints among themThe three basic components of the geometric constraints are described in termsof wrench and it is thus easy to model each common assembly constraint At the same time several different types of kinematicjoints in practice are presented in terms of twist For the reciprocal product of a twist and wrench which is equal to zero thegeometry constraints can be converted into the corresponding kinematic joints as a result To eliminate completely the redundantcomponents of different geometry constraints that act upon the same part the specific operation of a matrix space is applied Thisability is useful in supporting the kinematic design of properly constrained assemblies in CAD systems
1 Introduction
Currently we can design a machine in terms of a top-downapproach or a bottom-up approach in a CAD system Allparts of an assembly model are positioned using variousgeometry constraints so that they are in the correct relativepositions and orientations The 3D assembly models recordthe design parameters and some information that can beused for concept-level and detailed-level design Its benefitsare usually recognized only by those who design precisionmachines [1ndash3]
Any machine that consists of moving parts has to bedesigned to properly perform its kinematic functions Toanalyze the kinematic property of assemblies because of thelack of a kinematic description many researchers have usedscrew theory Applying screw theory to kinematics was firstperformed by Ball [4] Hunt [5] identified 22 cases of Ballrsquosreciprocal screw systems References [6ndash8] applied screwtheory to determine the degrees of freedom of any body
in a mechanism Eddie Baker [9] analyzed many types ofcomplex mechanisms based on [6] Mason and Salisbury[10] used screw theory to characterize the nature of contactsbetween robot gripper hands and objects To analyze roboticmultiple peg-in-hole Fei and Zhao [11] described the contactforces by using the screw theory in three dimensions Leeand Saitou [12] determined the dimensional integrity ofrobustness based on screw theory Kim et al [13] establisheda framework to estimate the reaction of constraints about aknee
Assembly features encode geometric and functional rela-tionships between parts that are widely used Thus manyresearchers perform kinematic analysis by using the assemblyfeatures Konkar and Cutkosky [14] created screw systemrepresentations of assembly mating features Adams et al[15ndash17] used Konkarrsquos algorithm and extended their workby defining extensible screw representations of some typesof assembly features Gerbino et al [18 19] proposed analgorithm to decompose the liaison diagram representing an
Hindawi Publishing CorporationMathematical Problems in EngineeringVolume 2016 Article ID 1761968 13 pageshttpdxdoiorg10115520161761968
2 Mathematical Problems in Engineering
assembly into simple paths to which the application of thetwist union or intersection algorithms may be easily appliedThey also highlighted the complexity behind the analysis ofoverconstraints with respect to the motion analysis Celen-tano [20] improved the algorithm presented in [18] anddeveloped a tool to analyze the motions and constraints ofCAD mechanisms Su et al [21] presented a theory-basedapproach for the analysis and synthesis of flexible jointsDixon and Shah [22] provided a mechanism for user-definedcustom assembly features by using an interactive systemAfter that twist and wrench matrices could be extractedfrom face pair relations and used in kinematic and structuralanalysis
Geometry constraints between two parts in an assemblymay be arbitrary combinations References [23ndash25] analyzedthe state of a constraint of an assembly mechanism or fixtureto see if it is properly constrained or whether it containsunwanted overconstraints Rusli et al [26 27] addressedattachment-level design in which decisions are made toestablish the types locations and orientations of assemblyfeatures Su et al [28] analyzed the mobility of overcon-strained linkages and compliant mechanisms To providea desired constrained motion for synthesizing a pattern offlexures a new screw theory that addresses ldquoline screwsrdquo andldquoline screw systemsrdquo was presented by Su and Tari [29]
According to [4ndash29] if the screw representations betweenany two parts of the assemblymodel are known we can easilyanalyze the modelrsquos kinematic functions References [4ndash1323ndash29] manually model the corresponding screw systemThus it is time consuming when the machine is large andcomplex On the other hand to determine the screw systemby using the methods proposed in [14ndash22] all of the screwrepresentations of different features that may be used firsthave to bemodeledThis is awkward and resembles geometricreasoning [30] If the incidence of redesign frequency is highthe screw system has to be rebuilt again and again which hasa great influence on the efficiency of the analysis procedure
We hope to automatically and easily obtain the screwsystem of any assembly model Thus when we modify theassembly model during analysis-redesign-reanalysis cyclesthe kinematic functions can be directly validated For con-venience we adopt a graph of parts with twist amongthem to express the screw system A screw system can beeasily used for kinematic analysis such as determining thedegrees of freedom estimating the reaction of constraintsand characterizing the nature of contacts between any twocomponents
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system based onreciprocal screw theoryThe two fundamental concepts in theconstraint-based approach that is freedom and constraintcan be represented mathematically by a twist and a wrenchBoth the twist and kinematic joint represent motion boththe wrench and geometry constraint represent constraintWhen a twist is given we can obtain the correspondingwrench as the reciprocal product of a twist and wrenchwhich is equal to zero Conversely if we know a wrench thenthe corresponding twist could be deduced as well To avoidgeometric reasoningwemodel the three basic components of
z
x
yo
A
Instantaneousscrew axis (ISA)
120596A
120596A
o
Figure 1 Twist
geometry constraints in terms of wrench Thus all commonassembly constraints can be automatically deduced For theredundant property of geometry constraints the specificoperation of a matrix space is applied Finally we achieve thegraph of parts with kinematic joints among themwith respectto an assembly model
11 Organization of the Paper Section 2 reviews the twist rep-resentations of common kinematic joints Section 3 analyzesthe basic components of geometry constraints and modelsthese components in terms of wrench Section 4 shows howto recognize the corresponding kinematic joints of assembliesrestricted by arbitrary combinations of geometry constraintsSection 5 contains several examples Section 6 presents ourconclusions
2 Screw Models of Kinematic Joints
To recognize the corresponding kinematic joints from geom-etry constraints a mathematical model must be developedthat can sufficiently and simultaneously describe the proper-ties of kinematic joints and geometry constraintsThis sectionshows how to define kinematic joints using twist
21 Twist A twist [32 33] takes the form
T = [120596119909120596119910120596119911 120592119909120592119910120592119911] (1)
A twist is a screw that describes the instantaneousmotionto the first order of a rigid body This means that the motiondescribed by a twist is allowed The first triplet representsthe angular velocity of the body with respect to a globalreference frameThe second triplet represents the velocity inthe global reference frame of that point on the body or itsextension which is instantaneously located at the origin ofthe global frame Therefore the unique line associated withthe first triplet is the axis of rotation and the unique pointassociated with the second triplet is the point on the body orits extension which is at that instant located at the origin ofthe global reference frame See Figure 1
The ISA in Figure 1 is the axis introduced in ChaslesThe-orem Both the twist and kinematic joint represent motion
Mathematical Problems in Engineering 3
ISA
Pv
Figure 2 Prismatic joint
Therefore we can use a twist to characterize a componentof the kinematic joint between two parts Afterwards thekinematic joint can be transformed into a correspondingtwist matrix
22 Twist Representations of Kinematic Joints There are6 types of kinematic joints in common prismatic jointrevolute joint helical joint cylindrical joint spherical jointand planar joint In this section we use s to characterizethe allowed motions of each kinematic joint and show theircorresponding twist matrices
221 Prismatic Joint The degrees of freedom (dof) of a pris-matic joint (119875) is one We use the unit vector 119875v = (V
119909 V119910V119911)
to characterize the allowed translation See Figure 2Thus 119875s has the form
119875s = 119875v (2)
The twist matrix representation of 119875 is given by
T119875 = [0 119875s] (3)
We use 0 instead of [0 0 0] unless otherwise stated
222 Revolute Joint The dof of a revolute joint (119877) is oneWeuse 119877L to characterize the axis of 119877 The unit vector of 119877Lis characterized by 119877v = (V
119909 V119910 V119911) 119877P = (119875
119909 119875119910 119875119911) is a
single point on the 119877L See Figure 3Thus 119877L has the form
119877L = 119877P + 119886 sdot 119877v (4)
where 119886 is a real number 119877s has the form
119877s = 119877P times 119877v (5)
Then the twist matrix representation of 119877 is given by
T119877 = [119877v 119877s] (6)
ISA
RP
Pv
RL
Figure 3 Revolute joint
ISA
HL
HP
Hv
Hp
Figure 4 Helical joint
223 Helical Joint The dof of a helical joint (H) is one Weuse119867L to characterize the axis of119867 The unit vector of119867Lis characterized by 119867v = (V
119909 V119910 V119911) 119867P = (119875
119909 119875119910 119875119911) is a
single point on the119867L See Figure 4Thus119867L has the form
119867L = 119867P + 119886 sdot 119867v (7)
where 119886 is a real number119867s has the form
119867s = 119867P times 119867v + 119867119901 sdot 119867v (8)
where119867119901 is the pitch of119867 Then the twist matrix represen-tation of119867 is given by
T119867 = [119867v 119867s] (9)
224 Cylindrical Joint The dof of a cylindrical joint (119862) istwo We use 119862L to characterize the axis of 119862 The unit vectorof119862L is characterized by119862v = (V
119909 V119910 V119911)119862P = (119875
119909 119875119910 119875119911)
is a single point on the 119862L See Figure 5Thus 119862L has the form
119862L = 119862P + 119886 sdot 119862v (10)
4 Mathematical Problems in Engineering
ISA
CL
CP
Cv
Figure 5 Cylindrical joint
SP
ISA
v1
v2
v3
Figure 6 Spherical joint
where 119886 is a real number 119862s has two components as shownin
119862s1= 119862P times 119862v
119862s2= 119862v
(11)
Therefore the twist matrix representation of119862 is given by
T119862 = [119862v 119862s
1
0 119862s2
] (12)
225 Spherical Joint The dof of a spherical joint (119878) is threeWe use 119878P = (119875
119909 119875119910 119875119911) to characterize the central point of
119878 and define v1= (1 0 0) v
2= (0 1 0) and v
3= (0 0 1) See
Figure 6Thus 119878s has three components as shown in
119878s1= 119878P times v
1= (0 119875
119911 minus119875119910)
119878s2= 119878P times v
2= (minus119875
119911 0 119875119909)
119878s3= 119878P times v
3= (119875119910 minus119875119909 0)
(13)
EP
ISA
En
Ev1
Ev2
Figure 7 Planar joint
Then the twist matrix representation of 119878 is given by
T119878 = [[[
v1 119878s1
v2 119878s2
v3 119878s3
]
]
]
(14)
226 Planar Joint The dof of a planar joint (119864) is three Thevectors 119864v
1= (v1199091 v1199101 v1199111) and 119864v
2= (v1199092 v1199102 v1199112) are
orthogonal in the planar surface and the unit vector 119864n isthe outer normal vector 119864P = (119875
119909 119875119910 119875119911) is a single point
on the surface See Figure 7Thus 119864s has three components as shown in
119864s1= 119864v1
119864s2= 119864v2
119864s3= 119864P times 119864n
(15)
Then the twist matrix representation of 119864 is given by
T119864 = [[[
0 119864s1
0 119864s2
119864n 119864s3
]
]
]
(16)
3 Reciprocal Screw Models ofGeometric Constraints
To transformarbitrary combinations of geometric constraintsbetween two parts into the corresponding kinematic jointwe should use the same mathematical model with kinematicjoints or a model that can be easily transformed into a twistThis section shows how to define geometry constraints usingwrench
31 Wrench A wrench [32 33] takes the form
W = [119891119909119891119910119891119911 119898119909119898119910119898119911] (17)
A wrench is also a screw and describes the resultant forceand moment of a force system acting on a rigid body Anyset of forces and couples applied to a body can be reduced toa set comprising a single force acting along a specific line inspace and a pure couple acting in a plane perpendicular tothat line This means that the motion described by a wrench
Mathematical Problems in Engineering 5
z
x
yo
M
Force system
Instantaneous screw axis (ISA)
Fi
F1 F2
F3
Fr
Mr
ri
Line of action of Fr
Fr
Figure 8 Wrench
is forbidden The first triplet describes the resultant force ina global reference frame The second triplet represents theresultant moment of the force system about the origin of theglobal frame See Figure 8
Both a wrench and geometry constraint are structuralconstraints Therefore we can use a wrench to characterizea component of the geometry constraints between two partsAfterwards the geometry constraints can be transformedinto a corresponding wrench matrix
32 Basic Components of Geometric Constraints When wedescribe a part in space the orientation and position shouldbe specified [34] The orientation cannot be described inabsolute terms An orientation is given by a rotation fromsome known reference orientationThe amount of rotation isknown as an angular displacement In other words describ-ing an orientation is mathematically equivalent to describingan angular displacement In the same way when we specifythe position of a part we cannot do so in absolute terms wemust always do so within the context of a specific referenceframeTherefore specifying a position is actually the same asspecifying an amount of translation from a given referencepoint (usually the origin of some coordinate system)
In practice there are many geometry constraints suchas align and point on line We can use one or arbitrarycombinations of geometric constraints to restrict the relativeorientation and position between two parts As we knowcommon geometry constraints such as coaxial constraintsand spherical constraints usually have a high degree ofconstraint (doc) However these geometry constraints are allcombined by the three different types of basic constraintsdot-1 constraint dot-2 constraint and the distance constraint[35] which can be divided into two categories the orientationpart and the position part
Consider a pair of rigid bodies denoted as bodies 119894 and119895 in Figure 9 Reference points P
119894and P
119895 reference frames
f119894-g119894-h119894and f
119895-g119895-h119895 and nonzero unit vectors a
119894and a
119895
are fixed in bodies 119894 and 119895 respectively Vector d119894119895connects
z
x
y
i
j
ai
Pihi
gi
f i
ri
sPidij
fj
hjgj
sPj
Pj aj
rj
Figure 9 Vectors fixed in and between bodies
P119894and P
119895between bodies The basic constraints are often
characterized by such elements
321 Dot-1 Constraint Dot-1 constraint restricts the relativeorientation of a pair of bodies that is
Φ
1198891
(a119894 a119895 1198621) equiv a119894
Ta119895minus 1198621= 0 119862
1isin (minus1 1) (18)
where doc(Φ1198891) = 1 and dot-1 constraint is symmetric withregard to bodies 119894 and 119895 Note that dot-1 constraint would notbe a basic constraint if119862
1= plusmn1Thismeans that the vectors a
119894
and a119895are parallel in the same or opposite directions In this
case doc(Φ1198891(a119894 a119895 plusmn1)) = 2
Writing the vectors a119894and a119895in terms of their respective
body reference transformationmatricesA119894andA
119895and body-
fixed constant vectors a1015840119894and a1015840
119895 where a
119894= A119894a1015840119894and a
119895=
A119895a1015840119895 (18) has the form
Φ
1198891
(a119894 a119895 1198621) equiv a1015840119894
TA119894
TA119895a1015840119895minus 1198621= 0
1198621isin (minus1 1)
(19)
where the transformation matrices A119894and A
119895in (19) depend
on the orientation generalized coordinates of bodies 119894 and 119895respectively
As a result the wrench representation for dot-1 constraintshould also freeze the relative orientation of the vectors a
119894and
a119895 Therefore the wrench matrix is given by
W1198891 = (0 a119894times a119895) (20)
If we use the body reference transformationmatrices (20)has the form
W1198891 = (0 (a1015840119894
TA119894
T) times (A
119895a1015840119895)) (21)
322 Dot-2 Constraint Dot-2 constraint restricts the relativeposition of a pair of bodies that is
Φ
1198892
(a119894 d119894119895 1198622) equiv a119894
Td119894119895minus 1198622= 0 d
119894119895= 0 (22)
6 Mathematical Problems in Engineering
where doc(Φ1198892) = 1 the same as that with dot-1 constraintalthough dot-2 constraint is not symmetric with regard tobodies 119894 and 119895 Analytically it is important to recall that dot-2constraint breaks down if d
119894119895= 0 Write the vector d
119894119895as
d119894119895= r119895+ A119895s1015840119895
Pminus r119894minus A119894s1015840119894
P (23)
Equation (22) becomes
Φ
1198892
(a119894 d119894119895 1198622) equiv a1015840119894
TA119894
T(r119895+ A119895s1015840119895
Pminus r119894minus A119894s1015840119894
P)
minus 1198622= 0
(24)
As a result the wrench representation for dot-2 constraintmust freeze the relative translation along the vector a
119894
Therefore the wrench matrix is given by
W1198892 = (a119894 0) (25)
Equation (25) also has the form
W1198892 = (A119894a1015840119894 0) (26)
323 Distance Constraint In practice we often require a pairof points on two bodies to coincide or maintain a determineddistance that is
Φ
119904
(119875119894 119875119895 1198623) equiv d119894119895
Td119894119895minus 1198623sdot 1198623= 0 (27)
where doc(Φ119904) would be different depending on whether 1198623
is zero Therefore we have
doc (Φ119904) = 1 if 1198623
= 0
doc (Φ119904) = 3 if 1198623= 0
(28)
As a result the wrench representation for the distanceconstraint should have two types of expressions as shown in(29) and (30)
W119904 = [d119894119895 0] if 119862
3= 0 (29)
where the wrench representation for the distance constraintshould freeze the relative translation along the vector d
119894119895if
1198623
= 0 However if 1198623= 0 the wrench representation has
the form
W119904 = [[[
1 0 0 00 1 0 00 0 1 0
]
]
]
if 1198623= 0 (30)
where the wrench which is illustrated in (30) freezes all therelative translations along the axes 119909 119910 and 119911
33 Wrench Representations of Geometry Constraints Tointroduce geometry constraints we give Figure 10
In Figure 10 reference points P1and P
2 reference frames
f1-g1-h1and f2-g2-h2 and central axes L
1and L
2are fixed in
cylinders 1198611and 119861
2 respectively Vector d
12connects P
1and
P2between the cylinders
L1
B1
h1
P1
f1
F1 g1
L2
B2
h2
P2
f2F2
g2
d12
Figure 10 Instance of geometry constraints
The reference point P1has the form
P1= (1198751199091 1198751199101 1198751199111) (31)
The central axis L1has the form
L1= P1+ 119886 sdot h
1 (32)
where 119886 is a real number The planar surface F1has the form
F1= P1+ 119887 sdot f
1+ 119888 sdot g
1 (33)
where 119887 and 119888 are both real numbers At the same timethe elements P
2 L2 and F
2have the same forms The
following geometry constraints are often characterized bysuch elements
We can divide all different types of geometry constraintswhich have more than one doc into six categories coaxialconstraint coplanar constraint spherical constraint align con-straint distance between line and surface and point on lineWe have modeled the three basic components of geometryconstraints in terms of wrench Thus the wrench models ofthese six types of geometry constraints can be automaticallydeduced
331 Coaxial Constraint Weuse119862119900119894119871119871(L1 L2) to denote the
coaxial constraint of axes L1and L
2 as shown in Figure 11
The doc of 119862119900119894119871119871(L1 L2) is four which restricts two
relative orientations and two relative translations In otherwords it comprises two dot-1 constraints and two dot-2constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(f1 d12 0)
Φ4
1198892
(g1 d12 0)
(34)
Then the wrench matrix of 119862119900119894119871119871(L1 L2) is given by
W119862119900119894119871119871(L
1L2)=
[
[
[
[
[
[
0 h1times f2
0 h1times g2
f1 0
g1 0
]
]
]
]
]
]
(35)
Mathematical Problems in Engineering 7
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
CoiLL(L1 L2)
Figure 11 Coaxial constraint
f2
P2
h2
F2
g2
L2
B2
f1P1
h1
F1
g1
L1
B1
CoiFF(F1 F2)
d12
Figure 12 Coplanar constraint
The corresponding kinematic joint of coaxial constraint iscylindrical joint
332 Coplanar Constraint We use 119862119900119894119865119865(F1 F2) to denote
the coplanar constraint of planar surfaces F1and F
2 as shown
in Figure 12The doc of 119862119900119894119865119865(F
1 F2) is three which restricts two
relative orientations and one relative translation In otherwords it comprises two dot-1 constraints and one dot-2constraint as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(h1 d12 0) (36)
Then the wrench matrix of 119862119900119894119865119865(F1 F2) is given by
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
(37)
The corresponding kinematic joint of coplanar constraintis planar joint
333 Spherical Constraint We use 119862119900119894119875119875(P1P2) to denote
the spherical constraint of two points P1and P
2 as shown in
Figure 13
f2
P2
h2
F2 g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPP(P1P2)
Figure 13 Spherical constraint
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
ParFF(F1 F2)
Figure 14 Align constraint
The doc of 119862119900119894119875119875(P1P2) is three which restricts three
relative translations in three unrelated directions In otherwords it comprises one distance constraint which has 119862
3= 0
in Section 323 as shown in
Φ
119904
(119875119894 119875119895 0) (38)
Then the wrench matrix of 119862119900119894119875119875(P1P2) is given by
W119862119900119894119875119875(P
1P2)=
[
[
[
1 0 0 00 1 0 00 0 1 0
]
]
]
(39)
The corresponding kinematic joint of spherical constraintis spherical joint
334 Align Constraint Align constraint (119875119886119903119865119865) is used tomake two planar surfaces F
1and F
2parallel as shown in
Figure 14The doc of 119875119886119903119865119865(F
1 F2) is two which restricts two
relative orientations In other words it comprises two dot-1constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) (40)
8 Mathematical Problems in Engineering
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
DisLF(L1 F2 D)
Figure 15 Distance between line and surface
Then the wrench matrix of 119875119886119903119865119865(F1 F2) is given by
W119875119886119903119865119865(F
1F2)= [
0 h1times f2
0 h1times g2
] (41)
We cannot find a corresponding kinematic joint for alignconstraint among the common kinematic joints
335 Distance between Line and Surface We use 119863119894119904119871119865(L1
F2 119863) to denote the axis L
1and the surface F
2that maintain
a certain distance119863 as shown in Figure 15The doc of 119863119894119904119871119865(L
1 F2 119863) is two which restricts one
relative orientation and one relative translation In otherwords it comprises one dot-1 constraint and one dot-2constraint as shown in
Φ1
1198891
(h1 h2 0) Φ
2
1198892
(h2 d12 119863) (42)
Then the wrench matrix of119863119894119904119871119865(L1 F2 119863) is given by
W119863119894119904119871119865(L
1F2119863)
= [
0 h1times h2
h2 0
] (43)
We also cannot find a corresponding kinematic jointfor distance between line and surface among the commonkinematic joints
336 Point on Line Weuse119862119900119894119875119871(P1 L2) to denote the point
P1on line L
2 as shown in Figure 16
The doc of 119862119900119894119875119871(P1 L2) is two which restricts two
relative translations In other words it comprises two dot-2constraints as shown in
Φ1
1198892
(f2 d12 0) Φ
2
1198892
(g2 d12 0) (44)
Then the wrench matrix of 119862119900119894119875119871(P1 L2) is given by
W119863119894119904119871119865(L
1F2119863)
= [
f2 0
g2 0
] (45)
We still cannot find a corresponding kinematic joint forpoint on line among the common kinematic joints
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPL(P1 L2)
Figure 16 Point on line
4 Recognition of Kinematic Joints
The key to recognizing kinematic joints in terms of geometryconstraints is the reciprocity relation We show how to definegeometry constraints using wrench The space of screws thatdefines the range of wrenches that can be exerted upon abody and the space of screws that defines the range of twiststhat the body can perform without breaking any contacts inthe joints of the mechanism are mutually reciprocal screwspaces When we know the wrenches acting upon a body wecan determine the motions it can perform Conversely if weknow the motions a body is capable of performing then theconstraints that are exerted upon it can also be deduced
41 Algorithm We assume that 119860 is a part in a 3D assemblymodel restricted by parts 119861
1 1198612 119861
119899 The twist space of 119860
is the space of screws reciprocal to the wrenches of contactacting upon it The space of wrenches acting upon 119860 is theunion of all the individual wrenches acting upon it fromthe parts 119861
119894 These individual wrenches are to be considered
as acting alone Algorithm 1 provides the twist space of 119860denoted by T119860
The 119877119890119888119894119901119903119900119888119886119897(P) in Algorithm 1 means to exchangethe first triplet and the second triplet of P For exampleif P = [119886 119887 119888 119889 119890 119891] then we have 119877119890119888119894119901119903119900119888119886119897(P) =
[119889 119890 119891 119886 119887 119888] The recursive nature of the procedurerequires the computation of the wrench space and twistspace of each component This proves to be computationallyexpensive because wrench space and twist space are definedin terms of the reciprocal of one another The operation ofderiving reciprocal screws involves computing the null spaceof matrices The speed of the algorithm is linearly related tothe number of 119861
119894 That is the algorithm is of the order 119874(119899)
42 Reciprocal Product According to reciprocal screw the-ory two screws (a twist and a wrench) are said to be mutuallyreciprocal if their reciprocal product is zero The reciprocal
Mathematical Problems in Engineering 9
BEGINFOR each part Bi which restrict AConvert all the geometry constraints between A and Bi into basic components 119862119895
119860 119895 = 1 119897
Determine the wrenchesW119896119860 119896= 1 119898 acting upon A through C119895
119860 We havem not less than l
Collect all the wrenchesW119896119860into a matrixW
END FORCompute the orthogonal matrixQ in terms ofW by Gram-Schmidt orthogonalization [31]IF Rank(Q) = 6 THENReturn ldquoIMMOBILErdquo
ELSECompute the orthogonal complement matrix P of Q
END IFTA =Reciprocal(P)
END
Algorithm 1 Function twist space
f2
A2
h2F2
g2
B2
f1A1
F1
g1B1
Figure 17 Redundant constraint
product of a twist T1and a wrenchW
1thatisequal to zero is
given by
T1∘W1= T11minus3
sdotW24minus6
+ T14minus6
sdotW21minus3
= [0 0] (46)
Mathematically because reciprocity imposes the relationT1∘W1 we can compute one of these quantities if the other is
known This allows us to compute the twists that a body mayexecute while in contact with other bodies without breakingthe contacts
43 Redundant Constraint When more than one geometryconstraint acts upon the same part because some compo-nents of these constraints may be redundant the resultantmotion is the logical intersection of the individual wrenchmatrix that defines each constraint for example when wemodel a revolute joint 119877
12with two links 119861
1and 119861
2 The
constraints used are 119862119900119894119871119871(A1A2) and 119862119900119894119865119865(F
1 F2) See
Figure 17Also dof(119877
12) = 1Thismeans that doc(119861
1 1198612) = 5 How-
ever we have doc(119862119900119894119871119871(A1A2)) = 4 and doc(119862119900119894119865119865(F
1
F2)) = 3 In other words the dimension of the orthog-
onal wrench space which comprises W119862119900119894119871119871(A
1A2)and
f2
h2
F2g2
B2
f1h1
g1
B1
F1
Figure 18 Recognition of planar joint
W119862119900119894119865119865(F
1F2) is 5 Therefore the motions that a body is
capable of performing while under the action of multiplewrenches are the logical intersection of the motions it canperform under the action of each wrench acting alone Thuswe can eliminate the redundant components of differentgeometry constraints that act upon the same part by usingGram-Schmidt orthogonalization in Algorithm 1
5 Examples
In each of the examples below different types of kinematicjoints are recognized
Example 1 (planar joint) Figure 18 shows a planar jointconsisting of two parts 119861
1and 119861
2 and a coplanar constraint
119862119900119894119865119865(F1 F2)
Following the process outlined in Section 332 we findbased on the definition of the coplanar constraint that thewrench matrix is
W119862119900119894119865119865(F
1F2)=
[
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
]
(47)
10 Mathematical Problems in Engineering
f2h2
F2
g2
B2
f1
h1 g1
B1
F1
f 9984002
h9984002g9984002
F9984002
f 9984001
h9984001
g9984001
F9984001
Figure 19 Recognition of prismatic joint
whereW119862119900119894119865119865(F
1F2)is already an orthogonalmatrixThen the
orthogonal complement matrix P119862119900119894119865119865(F
1F2)is
P119862119900119894119865119865(F
1F2)=
[
[
[
f1 0
g1 0
0 h1
]
]
]
(48)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119865119865(F1F2)) =
[
[
[
0 f1
0 g1
h1 0
]
]
]
(49)
where T12
represents the motion of a planar joint which isillustrated in Section 226 Moreover the recognition processof a cylindrical joint or spherical joint is approximately thesame as that of the planar joint
Example 2 (prismatic joint) Figure 19 shows a prismaticjoint consisting of two parts 119861
1and 119861
2 and two coplanar
constraints 119862119900119894119865119865(F1 F2) and 119862119900119894119865119865(F1015840
1 F10158402)
Following the process outlined in Example 1 we find thatthe wrench matrices between 119861
1and 119861
2are
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
W119862119900119894119865119865(F1015840
1F10158402)=
[
[
[
[
0 h10158401times f10158402
0 h10158401times g10158402
h10158401 0
]
]
]
]
(50)
B0
B1
B2
B3
B4
B5B6
B7
B8
B9
B10
B11
x
y
Figure 20 Hydraulic grab
where W12
= W119862119900119894119865119865(F
1F2)cup W119862119900119894119865119865(F1015840
1F10158402) The orthogonal
matrixQ12ofW12is
Q12=
[
[
[
[
[
[
[
[
[
[
[
0 h1times f2
0 h1times g2
0 h10158401times g10158402
h1 0
h10158401 0
]
]
]
]
]
]
]
]
]
]
]
(51)
The orthogonal complement matrix P12ofQ12is
P12= [g1 0] (52)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
12) = [0 g
1] (53)
where T12
represents the motion of a prismatic joint whichis shown in Section 221 Moreover the recognition processof a revolute joint is approximately the same as that of theprismatic joint
Example 3 (hydraulic grab) Here we present a projectexample A hydraulic grab with three degrees of freedom inCartesian coordinates is illustrated in Figure 20
The hydraulic grab comprises 12 parts 1198610 1198611 1198612 1198613
1198614 1198615 1198616 1198617 1198618 1198619 11986110 and 119861
11 12 coplanar constraints
119862119900119894119865119865(1198610 1198612)119862119900119894119865119865(119861
0 1198611)119862119900119894119865119865(119861
1 1198613)119862119900119894119865119865(119861
1 1198615)
119862119900119894119865119865(1198611 1198614) 119862119900119894119865119865(119861
4 1198616) 119862119900119894119865119865(119861
4 11986111) 119862119900119894119865119865(119861
4
1198617) 119862119900119894119865119865(119861
4 1198618) 119862119900119894119865119865(119861
9 11986110) 119862119900119894119865119865(119861
7 11986110) and
119862119900119894119865119865(11986110 11986111) and 15 coaxial constraints 119862119900119894119871119871(119861
0 1198612)
119862119900119894119871119871(1198610 1198611) 119862119900119894119871119871(119861
1 1198613) 119862119900119894119871119871(119861
2 1198613) 119862119900119894119871119871(119861
1 1198615)
119862119900119894119871119871(1198611 1198614) 119862119900119894119871119871(119861
4 1198616) 119862119900119894119871119871(119861
5 1198616) 119862119900119894119871119871(119861
4
11986111) 119862119900119894119871119871(119861
4 1198617) 119862119900119894119871119871(119861
4 1198618) 119862119900119894119871119871(119861
8 1198619)
119862119900119894119871119871(1198619 11986110) 119862119900119894119871119871(119861
7 11986110) and 119862119900119894119871119871(119861
10 11986111) See
Figure 21Here we present the processes of recognition among 119861
1
1198612 and 119861
3 as shown in Figure 22 The geometry constraint
between 1198612and 119861
3is 119862119900119894119871119871(119861
2 1198613) Following the process
Mathematical Problems in Engineering 11
CoiLL
CoiFF
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
Figure 21 The constraint graph of a hydraulic grab
h2 g2
f 9984003
g9984003
B2
B1
B0
g3h3
g9984001
f 9984001
B3
x
y
Figure 22 Parts 1198611 1198612 and 119861
3in a hydraulic grab
outlined in Section 331 we find based on the definition ofthe coaxial constraint that the wrench matrix is
W119862119900119894119871119871(119861
21198613)=
[
[
[
[
[
[
0 h2times f3
0 h2times g3
f2 0
g2 0
]
]
]
]
]
]
(54)
The orthogonal complement matrix is
P119862119900119894119871119871(119861
21198613)= [
0 h2
h2 0
] (55)
Then the twist matrix T23
between 1198612and 119861
3has the
following form
T23= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119871119871(11986121198613)) = [
h2 0
0 h2
] (56)
The geometry constraints between 1198611
and 1198613
are119862119900119894119871119871(119861
1 1198613) and 119862119900119894119865119865(119861
1 1198613) Thus the wrench matrices
between 1198611and 119861
3are
W119862119900119894119871119871(119861
11198613)=
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
]
]
]
]
]
]
]
W119862119900119894119865119865(119861
11198613)=
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
h10158403 0
]
]
]
]
(57)
where W13
= W119862119900119894119871119871(119861
11198613)cup W119862119900119894119865119865(119861
11198613) The orthogonal
matrixQ13ofW13is given by
Q13=
[
[
[
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
h10158403 0
]
]
]
]
]
]
]
]
]
]
(58)
Then the orthogonal complement matrix P13ofQ13is
P13= [0 h1015840
3] (59)
Therefore the twist matrix T13between 119861
1and 119861
3is
T13= 119877119890119888119894119901119903119900119888119886119897 (P
13) = [h1015840
3 0] (60)
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system basedon reciprocal screw theory According to the geometryconstraints between any two parts in the hydraulic graband Algorithm 1 we obtain the screw system as shown inFigure 23
As mentioned in [4ndash29] we can perform some differenttypes of kinematic analyses using the screw system such asdetermining the degrees of freedom of 119861
1and 119861
4 that is
T1198611
= T10cap (T20cup T32cup T31)
T1198614
= (T1198611
cup T41) cap (T
1198611
cup T51cup T65cup T64)
(61)
where the resultant of the motions that the end-effector ofa purely serial chain may perform due to the action of eachjoint acting alone is the union of that set and the resultant ofthe set of twists allowed separately by multiple contacts upona body is the intersection of the individual twists Thus thedegrees of freedom of 119861
1are equal to the dimension of the
orthogonal twist space of1198611 denoted byT
1198611
In the samewaywe can obtain the degrees of freedomof119861
4and any part in the
hydraulic grab
12 Mathematical Problems in Engineering
T twist
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
T32
T31
T10
T20
T109
T1110
T114
T98
T74 T8
4
T41
T64
T65
T51
T107
Figure 23 The screw system of a hydraulic grab
R revolute jointC cylindrical joint
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
R31
R10
R20
1110
R114
C98
R74 R8
4
R41
R64 R5
1
R107
R
R109 C3
2
C65
Figure 24 The kinematic joint model of a hydraulic grab
According to Sections 222 and 224 the kinematic jointsbetween 119861
2and 119861
3and 119861
1and 119861
3are the cylindrical joint and
revolute joint As a result we can obtain the kinematic jointmodel See Figure 24 If the corresponding kinematic jointbetween any two parts cannot be recognized we can easilyanalyze the modelrsquos geometry constraints
6 Conclusions
This paper investigates a method for determining the kine-matic joints of assemblies restricted by arbitrary combina-tions of geometry constraints using reciprocal screw theoryWe achieve the process of recognition in terms of thebasic components of geometry constraints and kinematicjoints We describe the three basic components of geometricconstraints in terms of wrench At the same time the specificoperation of a matrix space is applied to eliminate the redun-dant components between different geometry constraintsthat act upon the same part A user of this method can auto-matically and easily convert any assembly model into a screwsystem in any CAD system Afterwards the screw system canbe easily used for validating kinematic functions analyzingover- or underconstrained assembly configurations and soon
Additional Points
(i) We provide a new method for converting any 3Dassembly model into a screw system based on recip-rocal screw theory
(ii) We describe the three basic components of geometricconstraints in terms of wrench
(iii) We eliminate the redundant components of differentgeometry constraints that act upon the same part bythe specific operation of a matrix space
(iv) We establish the maps between geometry constraintsand kinematic joints as the reciprocal product of atwist and wrench which is equal to zero
(v) We present several application examples
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
The authors would like to thank the Ministry of Scienceand Technology of the Peoplersquos Republic of China (nos2011ZX02403-005 2011CB706502 and 2013AA041301) Thiswork is also supported by the National Natural ScienceFoundation of China (nos 61370182 and 51175198)
References
[1] A H Slocum Precision Machine Design Prentice Hall NewYork NY USA 1991
[2] M Rajh S Glodez J Flasker K Gotlih and T KostanjevecldquoDesign and analysis of an fMRI compatible haptic robotrdquoRobotics and Computer-Integrated Manufacturing vol 27 no 2pp 267ndash275 2011
[3] S Awtar and G Parmar ldquoDesign of a large range XY nanoposi-tioning systemrdquo Journal of Mechanisms and Robotics vol 5 no2 Article ID 021008 2013
[4] R S Ball A Treatise on the Theory of Screws CambridgeUniversity Press Cambridge UK 1998
[5] K H Hunt Kinematic Geometry of Mechanisms ClarendonPress Oxford UK 1978
[6] K J Waldron ldquoThe constraint analysis of mechanismsrdquo Journalof Mechanisms vol 1 no 2 pp 101ndash114 1966
[7] R Konkar Incremental kinematic analysis and symbolic syn-thesis of mechanisms [PhD dissertation] Stanford UniversityStanford Calif USA 1993
[8] J-S Zhao Z-J Feng and J-X Dong ldquoComputation of theconfiguration degree of freedomof a spatial parallel mechanismby using reciprocal screw theoryrdquo Mechanism and MachineTheory vol 41 no 12 pp 1486ndash1504 2006
[9] J Eddie Baker ldquoScrew system algebra applied to special linkageconfigurationsrdquoMechanism and Machine Theory vol 15 no 4pp 255ndash265 1980
[10] M TMason and J K Salisbury Robot Hands and theMechanicsof Manipulation MIT Press Cambridge Mass USA 1985
[11] Y Fei and X Zhao ldquoAn assembly process modeling and analysisfor robotic multiple peg-in-holerdquo Journal of Intelligent andRobotic Systems vol 36 no 2 pp 175ndash189 2003
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
2 Mathematical Problems in Engineering
assembly into simple paths to which the application of thetwist union or intersection algorithms may be easily appliedThey also highlighted the complexity behind the analysis ofoverconstraints with respect to the motion analysis Celen-tano [20] improved the algorithm presented in [18] anddeveloped a tool to analyze the motions and constraints ofCAD mechanisms Su et al [21] presented a theory-basedapproach for the analysis and synthesis of flexible jointsDixon and Shah [22] provided a mechanism for user-definedcustom assembly features by using an interactive systemAfter that twist and wrench matrices could be extractedfrom face pair relations and used in kinematic and structuralanalysis
Geometry constraints between two parts in an assemblymay be arbitrary combinations References [23ndash25] analyzedthe state of a constraint of an assembly mechanism or fixtureto see if it is properly constrained or whether it containsunwanted overconstraints Rusli et al [26 27] addressedattachment-level design in which decisions are made toestablish the types locations and orientations of assemblyfeatures Su et al [28] analyzed the mobility of overcon-strained linkages and compliant mechanisms To providea desired constrained motion for synthesizing a pattern offlexures a new screw theory that addresses ldquoline screwsrdquo andldquoline screw systemsrdquo was presented by Su and Tari [29]
According to [4ndash29] if the screw representations betweenany two parts of the assemblymodel are known we can easilyanalyze the modelrsquos kinematic functions References [4ndash1323ndash29] manually model the corresponding screw systemThus it is time consuming when the machine is large andcomplex On the other hand to determine the screw systemby using the methods proposed in [14ndash22] all of the screwrepresentations of different features that may be used firsthave to bemodeledThis is awkward and resembles geometricreasoning [30] If the incidence of redesign frequency is highthe screw system has to be rebuilt again and again which hasa great influence on the efficiency of the analysis procedure
We hope to automatically and easily obtain the screwsystem of any assembly model Thus when we modify theassembly model during analysis-redesign-reanalysis cyclesthe kinematic functions can be directly validated For con-venience we adopt a graph of parts with twist amongthem to express the screw system A screw system can beeasily used for kinematic analysis such as determining thedegrees of freedom estimating the reaction of constraintsand characterizing the nature of contacts between any twocomponents
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system based onreciprocal screw theoryThe two fundamental concepts in theconstraint-based approach that is freedom and constraintcan be represented mathematically by a twist and a wrenchBoth the twist and kinematic joint represent motion boththe wrench and geometry constraint represent constraintWhen a twist is given we can obtain the correspondingwrench as the reciprocal product of a twist and wrenchwhich is equal to zero Conversely if we know a wrench thenthe corresponding twist could be deduced as well To avoidgeometric reasoningwemodel the three basic components of
z
x
yo
A
Instantaneousscrew axis (ISA)
120596A
120596A
o
Figure 1 Twist
geometry constraints in terms of wrench Thus all commonassembly constraints can be automatically deduced For theredundant property of geometry constraints the specificoperation of a matrix space is applied Finally we achieve thegraph of parts with kinematic joints among themwith respectto an assembly model
11 Organization of the Paper Section 2 reviews the twist rep-resentations of common kinematic joints Section 3 analyzesthe basic components of geometry constraints and modelsthese components in terms of wrench Section 4 shows howto recognize the corresponding kinematic joints of assembliesrestricted by arbitrary combinations of geometry constraintsSection 5 contains several examples Section 6 presents ourconclusions
2 Screw Models of Kinematic Joints
To recognize the corresponding kinematic joints from geom-etry constraints a mathematical model must be developedthat can sufficiently and simultaneously describe the proper-ties of kinematic joints and geometry constraintsThis sectionshows how to define kinematic joints using twist
21 Twist A twist [32 33] takes the form
T = [120596119909120596119910120596119911 120592119909120592119910120592119911] (1)
A twist is a screw that describes the instantaneousmotionto the first order of a rigid body This means that the motiondescribed by a twist is allowed The first triplet representsthe angular velocity of the body with respect to a globalreference frameThe second triplet represents the velocity inthe global reference frame of that point on the body or itsextension which is instantaneously located at the origin ofthe global frame Therefore the unique line associated withthe first triplet is the axis of rotation and the unique pointassociated with the second triplet is the point on the body orits extension which is at that instant located at the origin ofthe global reference frame See Figure 1
The ISA in Figure 1 is the axis introduced in ChaslesThe-orem Both the twist and kinematic joint represent motion
Mathematical Problems in Engineering 3
ISA
Pv
Figure 2 Prismatic joint
Therefore we can use a twist to characterize a componentof the kinematic joint between two parts Afterwards thekinematic joint can be transformed into a correspondingtwist matrix
22 Twist Representations of Kinematic Joints There are6 types of kinematic joints in common prismatic jointrevolute joint helical joint cylindrical joint spherical jointand planar joint In this section we use s to characterizethe allowed motions of each kinematic joint and show theircorresponding twist matrices
221 Prismatic Joint The degrees of freedom (dof) of a pris-matic joint (119875) is one We use the unit vector 119875v = (V
119909 V119910V119911)
to characterize the allowed translation See Figure 2Thus 119875s has the form
119875s = 119875v (2)
The twist matrix representation of 119875 is given by
T119875 = [0 119875s] (3)
We use 0 instead of [0 0 0] unless otherwise stated
222 Revolute Joint The dof of a revolute joint (119877) is oneWeuse 119877L to characterize the axis of 119877 The unit vector of 119877Lis characterized by 119877v = (V
119909 V119910 V119911) 119877P = (119875
119909 119875119910 119875119911) is a
single point on the 119877L See Figure 3Thus 119877L has the form
119877L = 119877P + 119886 sdot 119877v (4)
where 119886 is a real number 119877s has the form
119877s = 119877P times 119877v (5)
Then the twist matrix representation of 119877 is given by
T119877 = [119877v 119877s] (6)
ISA
RP
Pv
RL
Figure 3 Revolute joint
ISA
HL
HP
Hv
Hp
Figure 4 Helical joint
223 Helical Joint The dof of a helical joint (H) is one Weuse119867L to characterize the axis of119867 The unit vector of119867Lis characterized by 119867v = (V
119909 V119910 V119911) 119867P = (119875
119909 119875119910 119875119911) is a
single point on the119867L See Figure 4Thus119867L has the form
119867L = 119867P + 119886 sdot 119867v (7)
where 119886 is a real number119867s has the form
119867s = 119867P times 119867v + 119867119901 sdot 119867v (8)
where119867119901 is the pitch of119867 Then the twist matrix represen-tation of119867 is given by
T119867 = [119867v 119867s] (9)
224 Cylindrical Joint The dof of a cylindrical joint (119862) istwo We use 119862L to characterize the axis of 119862 The unit vectorof119862L is characterized by119862v = (V
119909 V119910 V119911)119862P = (119875
119909 119875119910 119875119911)
is a single point on the 119862L See Figure 5Thus 119862L has the form
119862L = 119862P + 119886 sdot 119862v (10)
4 Mathematical Problems in Engineering
ISA
CL
CP
Cv
Figure 5 Cylindrical joint
SP
ISA
v1
v2
v3
Figure 6 Spherical joint
where 119886 is a real number 119862s has two components as shownin
119862s1= 119862P times 119862v
119862s2= 119862v
(11)
Therefore the twist matrix representation of119862 is given by
T119862 = [119862v 119862s
1
0 119862s2
] (12)
225 Spherical Joint The dof of a spherical joint (119878) is threeWe use 119878P = (119875
119909 119875119910 119875119911) to characterize the central point of
119878 and define v1= (1 0 0) v
2= (0 1 0) and v
3= (0 0 1) See
Figure 6Thus 119878s has three components as shown in
119878s1= 119878P times v
1= (0 119875
119911 minus119875119910)
119878s2= 119878P times v
2= (minus119875
119911 0 119875119909)
119878s3= 119878P times v
3= (119875119910 minus119875119909 0)
(13)
EP
ISA
En
Ev1
Ev2
Figure 7 Planar joint
Then the twist matrix representation of 119878 is given by
T119878 = [[[
v1 119878s1
v2 119878s2
v3 119878s3
]
]
]
(14)
226 Planar Joint The dof of a planar joint (119864) is three Thevectors 119864v
1= (v1199091 v1199101 v1199111) and 119864v
2= (v1199092 v1199102 v1199112) are
orthogonal in the planar surface and the unit vector 119864n isthe outer normal vector 119864P = (119875
119909 119875119910 119875119911) is a single point
on the surface See Figure 7Thus 119864s has three components as shown in
119864s1= 119864v1
119864s2= 119864v2
119864s3= 119864P times 119864n
(15)
Then the twist matrix representation of 119864 is given by
T119864 = [[[
0 119864s1
0 119864s2
119864n 119864s3
]
]
]
(16)
3 Reciprocal Screw Models ofGeometric Constraints
To transformarbitrary combinations of geometric constraintsbetween two parts into the corresponding kinematic jointwe should use the same mathematical model with kinematicjoints or a model that can be easily transformed into a twistThis section shows how to define geometry constraints usingwrench
31 Wrench A wrench [32 33] takes the form
W = [119891119909119891119910119891119911 119898119909119898119910119898119911] (17)
A wrench is also a screw and describes the resultant forceand moment of a force system acting on a rigid body Anyset of forces and couples applied to a body can be reduced toa set comprising a single force acting along a specific line inspace and a pure couple acting in a plane perpendicular tothat line This means that the motion described by a wrench
Mathematical Problems in Engineering 5
z
x
yo
M
Force system
Instantaneous screw axis (ISA)
Fi
F1 F2
F3
Fr
Mr
ri
Line of action of Fr
Fr
Figure 8 Wrench
is forbidden The first triplet describes the resultant force ina global reference frame The second triplet represents theresultant moment of the force system about the origin of theglobal frame See Figure 8
Both a wrench and geometry constraint are structuralconstraints Therefore we can use a wrench to characterizea component of the geometry constraints between two partsAfterwards the geometry constraints can be transformedinto a corresponding wrench matrix
32 Basic Components of Geometric Constraints When wedescribe a part in space the orientation and position shouldbe specified [34] The orientation cannot be described inabsolute terms An orientation is given by a rotation fromsome known reference orientationThe amount of rotation isknown as an angular displacement In other words describ-ing an orientation is mathematically equivalent to describingan angular displacement In the same way when we specifythe position of a part we cannot do so in absolute terms wemust always do so within the context of a specific referenceframeTherefore specifying a position is actually the same asspecifying an amount of translation from a given referencepoint (usually the origin of some coordinate system)
In practice there are many geometry constraints suchas align and point on line We can use one or arbitrarycombinations of geometric constraints to restrict the relativeorientation and position between two parts As we knowcommon geometry constraints such as coaxial constraintsand spherical constraints usually have a high degree ofconstraint (doc) However these geometry constraints are allcombined by the three different types of basic constraintsdot-1 constraint dot-2 constraint and the distance constraint[35] which can be divided into two categories the orientationpart and the position part
Consider a pair of rigid bodies denoted as bodies 119894 and119895 in Figure 9 Reference points P
119894and P
119895 reference frames
f119894-g119894-h119894and f
119895-g119895-h119895 and nonzero unit vectors a
119894and a
119895
are fixed in bodies 119894 and 119895 respectively Vector d119894119895connects
z
x
y
i
j
ai
Pihi
gi
f i
ri
sPidij
fj
hjgj
sPj
Pj aj
rj
Figure 9 Vectors fixed in and between bodies
P119894and P
119895between bodies The basic constraints are often
characterized by such elements
321 Dot-1 Constraint Dot-1 constraint restricts the relativeorientation of a pair of bodies that is
Φ
1198891
(a119894 a119895 1198621) equiv a119894
Ta119895minus 1198621= 0 119862
1isin (minus1 1) (18)
where doc(Φ1198891) = 1 and dot-1 constraint is symmetric withregard to bodies 119894 and 119895 Note that dot-1 constraint would notbe a basic constraint if119862
1= plusmn1Thismeans that the vectors a
119894
and a119895are parallel in the same or opposite directions In this
case doc(Φ1198891(a119894 a119895 plusmn1)) = 2
Writing the vectors a119894and a119895in terms of their respective
body reference transformationmatricesA119894andA
119895and body-
fixed constant vectors a1015840119894and a1015840
119895 where a
119894= A119894a1015840119894and a
119895=
A119895a1015840119895 (18) has the form
Φ
1198891
(a119894 a119895 1198621) equiv a1015840119894
TA119894
TA119895a1015840119895minus 1198621= 0
1198621isin (minus1 1)
(19)
where the transformation matrices A119894and A
119895in (19) depend
on the orientation generalized coordinates of bodies 119894 and 119895respectively
As a result the wrench representation for dot-1 constraintshould also freeze the relative orientation of the vectors a
119894and
a119895 Therefore the wrench matrix is given by
W1198891 = (0 a119894times a119895) (20)
If we use the body reference transformationmatrices (20)has the form
W1198891 = (0 (a1015840119894
TA119894
T) times (A
119895a1015840119895)) (21)
322 Dot-2 Constraint Dot-2 constraint restricts the relativeposition of a pair of bodies that is
Φ
1198892
(a119894 d119894119895 1198622) equiv a119894
Td119894119895minus 1198622= 0 d
119894119895= 0 (22)
6 Mathematical Problems in Engineering
where doc(Φ1198892) = 1 the same as that with dot-1 constraintalthough dot-2 constraint is not symmetric with regard tobodies 119894 and 119895 Analytically it is important to recall that dot-2constraint breaks down if d
119894119895= 0 Write the vector d
119894119895as
d119894119895= r119895+ A119895s1015840119895
Pminus r119894minus A119894s1015840119894
P (23)
Equation (22) becomes
Φ
1198892
(a119894 d119894119895 1198622) equiv a1015840119894
TA119894
T(r119895+ A119895s1015840119895
Pminus r119894minus A119894s1015840119894
P)
minus 1198622= 0
(24)
As a result the wrench representation for dot-2 constraintmust freeze the relative translation along the vector a
119894
Therefore the wrench matrix is given by
W1198892 = (a119894 0) (25)
Equation (25) also has the form
W1198892 = (A119894a1015840119894 0) (26)
323 Distance Constraint In practice we often require a pairof points on two bodies to coincide or maintain a determineddistance that is
Φ
119904
(119875119894 119875119895 1198623) equiv d119894119895
Td119894119895minus 1198623sdot 1198623= 0 (27)
where doc(Φ119904) would be different depending on whether 1198623
is zero Therefore we have
doc (Φ119904) = 1 if 1198623
= 0
doc (Φ119904) = 3 if 1198623= 0
(28)
As a result the wrench representation for the distanceconstraint should have two types of expressions as shown in(29) and (30)
W119904 = [d119894119895 0] if 119862
3= 0 (29)
where the wrench representation for the distance constraintshould freeze the relative translation along the vector d
119894119895if
1198623
= 0 However if 1198623= 0 the wrench representation has
the form
W119904 = [[[
1 0 0 00 1 0 00 0 1 0
]
]
]
if 1198623= 0 (30)
where the wrench which is illustrated in (30) freezes all therelative translations along the axes 119909 119910 and 119911
33 Wrench Representations of Geometry Constraints Tointroduce geometry constraints we give Figure 10
In Figure 10 reference points P1and P
2 reference frames
f1-g1-h1and f2-g2-h2 and central axes L
1and L
2are fixed in
cylinders 1198611and 119861
2 respectively Vector d
12connects P
1and
P2between the cylinders
L1
B1
h1
P1
f1
F1 g1
L2
B2
h2
P2
f2F2
g2
d12
Figure 10 Instance of geometry constraints
The reference point P1has the form
P1= (1198751199091 1198751199101 1198751199111) (31)
The central axis L1has the form
L1= P1+ 119886 sdot h
1 (32)
where 119886 is a real number The planar surface F1has the form
F1= P1+ 119887 sdot f
1+ 119888 sdot g
1 (33)
where 119887 and 119888 are both real numbers At the same timethe elements P
2 L2 and F
2have the same forms The
following geometry constraints are often characterized bysuch elements
We can divide all different types of geometry constraintswhich have more than one doc into six categories coaxialconstraint coplanar constraint spherical constraint align con-straint distance between line and surface and point on lineWe have modeled the three basic components of geometryconstraints in terms of wrench Thus the wrench models ofthese six types of geometry constraints can be automaticallydeduced
331 Coaxial Constraint Weuse119862119900119894119871119871(L1 L2) to denote the
coaxial constraint of axes L1and L
2 as shown in Figure 11
The doc of 119862119900119894119871119871(L1 L2) is four which restricts two
relative orientations and two relative translations In otherwords it comprises two dot-1 constraints and two dot-2constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(f1 d12 0)
Φ4
1198892
(g1 d12 0)
(34)
Then the wrench matrix of 119862119900119894119871119871(L1 L2) is given by
W119862119900119894119871119871(L
1L2)=
[
[
[
[
[
[
0 h1times f2
0 h1times g2
f1 0
g1 0
]
]
]
]
]
]
(35)
Mathematical Problems in Engineering 7
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
CoiLL(L1 L2)
Figure 11 Coaxial constraint
f2
P2
h2
F2
g2
L2
B2
f1P1
h1
F1
g1
L1
B1
CoiFF(F1 F2)
d12
Figure 12 Coplanar constraint
The corresponding kinematic joint of coaxial constraint iscylindrical joint
332 Coplanar Constraint We use 119862119900119894119865119865(F1 F2) to denote
the coplanar constraint of planar surfaces F1and F
2 as shown
in Figure 12The doc of 119862119900119894119865119865(F
1 F2) is three which restricts two
relative orientations and one relative translation In otherwords it comprises two dot-1 constraints and one dot-2constraint as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(h1 d12 0) (36)
Then the wrench matrix of 119862119900119894119865119865(F1 F2) is given by
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
(37)
The corresponding kinematic joint of coplanar constraintis planar joint
333 Spherical Constraint We use 119862119900119894119875119875(P1P2) to denote
the spherical constraint of two points P1and P
2 as shown in
Figure 13
f2
P2
h2
F2 g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPP(P1P2)
Figure 13 Spherical constraint
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
ParFF(F1 F2)
Figure 14 Align constraint
The doc of 119862119900119894119875119875(P1P2) is three which restricts three
relative translations in three unrelated directions In otherwords it comprises one distance constraint which has 119862
3= 0
in Section 323 as shown in
Φ
119904
(119875119894 119875119895 0) (38)
Then the wrench matrix of 119862119900119894119875119875(P1P2) is given by
W119862119900119894119875119875(P
1P2)=
[
[
[
1 0 0 00 1 0 00 0 1 0
]
]
]
(39)
The corresponding kinematic joint of spherical constraintis spherical joint
334 Align Constraint Align constraint (119875119886119903119865119865) is used tomake two planar surfaces F
1and F
2parallel as shown in
Figure 14The doc of 119875119886119903119865119865(F
1 F2) is two which restricts two
relative orientations In other words it comprises two dot-1constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) (40)
8 Mathematical Problems in Engineering
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
DisLF(L1 F2 D)
Figure 15 Distance between line and surface
Then the wrench matrix of 119875119886119903119865119865(F1 F2) is given by
W119875119886119903119865119865(F
1F2)= [
0 h1times f2
0 h1times g2
] (41)
We cannot find a corresponding kinematic joint for alignconstraint among the common kinematic joints
335 Distance between Line and Surface We use 119863119894119904119871119865(L1
F2 119863) to denote the axis L
1and the surface F
2that maintain
a certain distance119863 as shown in Figure 15The doc of 119863119894119904119871119865(L
1 F2 119863) is two which restricts one
relative orientation and one relative translation In otherwords it comprises one dot-1 constraint and one dot-2constraint as shown in
Φ1
1198891
(h1 h2 0) Φ
2
1198892
(h2 d12 119863) (42)
Then the wrench matrix of119863119894119904119871119865(L1 F2 119863) is given by
W119863119894119904119871119865(L
1F2119863)
= [
0 h1times h2
h2 0
] (43)
We also cannot find a corresponding kinematic jointfor distance between line and surface among the commonkinematic joints
336 Point on Line Weuse119862119900119894119875119871(P1 L2) to denote the point
P1on line L
2 as shown in Figure 16
The doc of 119862119900119894119875119871(P1 L2) is two which restricts two
relative translations In other words it comprises two dot-2constraints as shown in
Φ1
1198892
(f2 d12 0) Φ
2
1198892
(g2 d12 0) (44)
Then the wrench matrix of 119862119900119894119875119871(P1 L2) is given by
W119863119894119904119871119865(L
1F2119863)
= [
f2 0
g2 0
] (45)
We still cannot find a corresponding kinematic joint forpoint on line among the common kinematic joints
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPL(P1 L2)
Figure 16 Point on line
4 Recognition of Kinematic Joints
The key to recognizing kinematic joints in terms of geometryconstraints is the reciprocity relation We show how to definegeometry constraints using wrench The space of screws thatdefines the range of wrenches that can be exerted upon abody and the space of screws that defines the range of twiststhat the body can perform without breaking any contacts inthe joints of the mechanism are mutually reciprocal screwspaces When we know the wrenches acting upon a body wecan determine the motions it can perform Conversely if weknow the motions a body is capable of performing then theconstraints that are exerted upon it can also be deduced
41 Algorithm We assume that 119860 is a part in a 3D assemblymodel restricted by parts 119861
1 1198612 119861
119899 The twist space of 119860
is the space of screws reciprocal to the wrenches of contactacting upon it The space of wrenches acting upon 119860 is theunion of all the individual wrenches acting upon it fromthe parts 119861
119894 These individual wrenches are to be considered
as acting alone Algorithm 1 provides the twist space of 119860denoted by T119860
The 119877119890119888119894119901119903119900119888119886119897(P) in Algorithm 1 means to exchangethe first triplet and the second triplet of P For exampleif P = [119886 119887 119888 119889 119890 119891] then we have 119877119890119888119894119901119903119900119888119886119897(P) =
[119889 119890 119891 119886 119887 119888] The recursive nature of the procedurerequires the computation of the wrench space and twistspace of each component This proves to be computationallyexpensive because wrench space and twist space are definedin terms of the reciprocal of one another The operation ofderiving reciprocal screws involves computing the null spaceof matrices The speed of the algorithm is linearly related tothe number of 119861
119894 That is the algorithm is of the order 119874(119899)
42 Reciprocal Product According to reciprocal screw the-ory two screws (a twist and a wrench) are said to be mutuallyreciprocal if their reciprocal product is zero The reciprocal
Mathematical Problems in Engineering 9
BEGINFOR each part Bi which restrict AConvert all the geometry constraints between A and Bi into basic components 119862119895
119860 119895 = 1 119897
Determine the wrenchesW119896119860 119896= 1 119898 acting upon A through C119895
119860 We havem not less than l
Collect all the wrenchesW119896119860into a matrixW
END FORCompute the orthogonal matrixQ in terms ofW by Gram-Schmidt orthogonalization [31]IF Rank(Q) = 6 THENReturn ldquoIMMOBILErdquo
ELSECompute the orthogonal complement matrix P of Q
END IFTA =Reciprocal(P)
END
Algorithm 1 Function twist space
f2
A2
h2F2
g2
B2
f1A1
F1
g1B1
Figure 17 Redundant constraint
product of a twist T1and a wrenchW
1thatisequal to zero is
given by
T1∘W1= T11minus3
sdotW24minus6
+ T14minus6
sdotW21minus3
= [0 0] (46)
Mathematically because reciprocity imposes the relationT1∘W1 we can compute one of these quantities if the other is
known This allows us to compute the twists that a body mayexecute while in contact with other bodies without breakingthe contacts
43 Redundant Constraint When more than one geometryconstraint acts upon the same part because some compo-nents of these constraints may be redundant the resultantmotion is the logical intersection of the individual wrenchmatrix that defines each constraint for example when wemodel a revolute joint 119877
12with two links 119861
1and 119861
2 The
constraints used are 119862119900119894119871119871(A1A2) and 119862119900119894119865119865(F
1 F2) See
Figure 17Also dof(119877
12) = 1Thismeans that doc(119861
1 1198612) = 5 How-
ever we have doc(119862119900119894119871119871(A1A2)) = 4 and doc(119862119900119894119865119865(F
1
F2)) = 3 In other words the dimension of the orthog-
onal wrench space which comprises W119862119900119894119871119871(A
1A2)and
f2
h2
F2g2
B2
f1h1
g1
B1
F1
Figure 18 Recognition of planar joint
W119862119900119894119865119865(F
1F2) is 5 Therefore the motions that a body is
capable of performing while under the action of multiplewrenches are the logical intersection of the motions it canperform under the action of each wrench acting alone Thuswe can eliminate the redundant components of differentgeometry constraints that act upon the same part by usingGram-Schmidt orthogonalization in Algorithm 1
5 Examples
In each of the examples below different types of kinematicjoints are recognized
Example 1 (planar joint) Figure 18 shows a planar jointconsisting of two parts 119861
1and 119861
2 and a coplanar constraint
119862119900119894119865119865(F1 F2)
Following the process outlined in Section 332 we findbased on the definition of the coplanar constraint that thewrench matrix is
W119862119900119894119865119865(F
1F2)=
[
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
]
(47)
10 Mathematical Problems in Engineering
f2h2
F2
g2
B2
f1
h1 g1
B1
F1
f 9984002
h9984002g9984002
F9984002
f 9984001
h9984001
g9984001
F9984001
Figure 19 Recognition of prismatic joint
whereW119862119900119894119865119865(F
1F2)is already an orthogonalmatrixThen the
orthogonal complement matrix P119862119900119894119865119865(F
1F2)is
P119862119900119894119865119865(F
1F2)=
[
[
[
f1 0
g1 0
0 h1
]
]
]
(48)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119865119865(F1F2)) =
[
[
[
0 f1
0 g1
h1 0
]
]
]
(49)
where T12
represents the motion of a planar joint which isillustrated in Section 226 Moreover the recognition processof a cylindrical joint or spherical joint is approximately thesame as that of the planar joint
Example 2 (prismatic joint) Figure 19 shows a prismaticjoint consisting of two parts 119861
1and 119861
2 and two coplanar
constraints 119862119900119894119865119865(F1 F2) and 119862119900119894119865119865(F1015840
1 F10158402)
Following the process outlined in Example 1 we find thatthe wrench matrices between 119861
1and 119861
2are
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
W119862119900119894119865119865(F1015840
1F10158402)=
[
[
[
[
0 h10158401times f10158402
0 h10158401times g10158402
h10158401 0
]
]
]
]
(50)
B0
B1
B2
B3
B4
B5B6
B7
B8
B9
B10
B11
x
y
Figure 20 Hydraulic grab
where W12
= W119862119900119894119865119865(F
1F2)cup W119862119900119894119865119865(F1015840
1F10158402) The orthogonal
matrixQ12ofW12is
Q12=
[
[
[
[
[
[
[
[
[
[
[
0 h1times f2
0 h1times g2
0 h10158401times g10158402
h1 0
h10158401 0
]
]
]
]
]
]
]
]
]
]
]
(51)
The orthogonal complement matrix P12ofQ12is
P12= [g1 0] (52)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
12) = [0 g
1] (53)
where T12
represents the motion of a prismatic joint whichis shown in Section 221 Moreover the recognition processof a revolute joint is approximately the same as that of theprismatic joint
Example 3 (hydraulic grab) Here we present a projectexample A hydraulic grab with three degrees of freedom inCartesian coordinates is illustrated in Figure 20
The hydraulic grab comprises 12 parts 1198610 1198611 1198612 1198613
1198614 1198615 1198616 1198617 1198618 1198619 11986110 and 119861
11 12 coplanar constraints
119862119900119894119865119865(1198610 1198612)119862119900119894119865119865(119861
0 1198611)119862119900119894119865119865(119861
1 1198613)119862119900119894119865119865(119861
1 1198615)
119862119900119894119865119865(1198611 1198614) 119862119900119894119865119865(119861
4 1198616) 119862119900119894119865119865(119861
4 11986111) 119862119900119894119865119865(119861
4
1198617) 119862119900119894119865119865(119861
4 1198618) 119862119900119894119865119865(119861
9 11986110) 119862119900119894119865119865(119861
7 11986110) and
119862119900119894119865119865(11986110 11986111) and 15 coaxial constraints 119862119900119894119871119871(119861
0 1198612)
119862119900119894119871119871(1198610 1198611) 119862119900119894119871119871(119861
1 1198613) 119862119900119894119871119871(119861
2 1198613) 119862119900119894119871119871(119861
1 1198615)
119862119900119894119871119871(1198611 1198614) 119862119900119894119871119871(119861
4 1198616) 119862119900119894119871119871(119861
5 1198616) 119862119900119894119871119871(119861
4
11986111) 119862119900119894119871119871(119861
4 1198617) 119862119900119894119871119871(119861
4 1198618) 119862119900119894119871119871(119861
8 1198619)
119862119900119894119871119871(1198619 11986110) 119862119900119894119871119871(119861
7 11986110) and 119862119900119894119871119871(119861
10 11986111) See
Figure 21Here we present the processes of recognition among 119861
1
1198612 and 119861
3 as shown in Figure 22 The geometry constraint
between 1198612and 119861
3is 119862119900119894119871119871(119861
2 1198613) Following the process
Mathematical Problems in Engineering 11
CoiLL
CoiFF
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
Figure 21 The constraint graph of a hydraulic grab
h2 g2
f 9984003
g9984003
B2
B1
B0
g3h3
g9984001
f 9984001
B3
x
y
Figure 22 Parts 1198611 1198612 and 119861
3in a hydraulic grab
outlined in Section 331 we find based on the definition ofthe coaxial constraint that the wrench matrix is
W119862119900119894119871119871(119861
21198613)=
[
[
[
[
[
[
0 h2times f3
0 h2times g3
f2 0
g2 0
]
]
]
]
]
]
(54)
The orthogonal complement matrix is
P119862119900119894119871119871(119861
21198613)= [
0 h2
h2 0
] (55)
Then the twist matrix T23
between 1198612and 119861
3has the
following form
T23= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119871119871(11986121198613)) = [
h2 0
0 h2
] (56)
The geometry constraints between 1198611
and 1198613
are119862119900119894119871119871(119861
1 1198613) and 119862119900119894119865119865(119861
1 1198613) Thus the wrench matrices
between 1198611and 119861
3are
W119862119900119894119871119871(119861
11198613)=
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
]
]
]
]
]
]
]
W119862119900119894119865119865(119861
11198613)=
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
h10158403 0
]
]
]
]
(57)
where W13
= W119862119900119894119871119871(119861
11198613)cup W119862119900119894119865119865(119861
11198613) The orthogonal
matrixQ13ofW13is given by
Q13=
[
[
[
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
h10158403 0
]
]
]
]
]
]
]
]
]
]
(58)
Then the orthogonal complement matrix P13ofQ13is
P13= [0 h1015840
3] (59)
Therefore the twist matrix T13between 119861
1and 119861
3is
T13= 119877119890119888119894119901119903119900119888119886119897 (P
13) = [h1015840
3 0] (60)
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system basedon reciprocal screw theory According to the geometryconstraints between any two parts in the hydraulic graband Algorithm 1 we obtain the screw system as shown inFigure 23
As mentioned in [4ndash29] we can perform some differenttypes of kinematic analyses using the screw system such asdetermining the degrees of freedom of 119861
1and 119861
4 that is
T1198611
= T10cap (T20cup T32cup T31)
T1198614
= (T1198611
cup T41) cap (T
1198611
cup T51cup T65cup T64)
(61)
where the resultant of the motions that the end-effector ofa purely serial chain may perform due to the action of eachjoint acting alone is the union of that set and the resultant ofthe set of twists allowed separately by multiple contacts upona body is the intersection of the individual twists Thus thedegrees of freedom of 119861
1are equal to the dimension of the
orthogonal twist space of1198611 denoted byT
1198611
In the samewaywe can obtain the degrees of freedomof119861
4and any part in the
hydraulic grab
12 Mathematical Problems in Engineering
T twist
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
T32
T31
T10
T20
T109
T1110
T114
T98
T74 T8
4
T41
T64
T65
T51
T107
Figure 23 The screw system of a hydraulic grab
R revolute jointC cylindrical joint
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
R31
R10
R20
1110
R114
C98
R74 R8
4
R41
R64 R5
1
R107
R
R109 C3
2
C65
Figure 24 The kinematic joint model of a hydraulic grab
According to Sections 222 and 224 the kinematic jointsbetween 119861
2and 119861
3and 119861
1and 119861
3are the cylindrical joint and
revolute joint As a result we can obtain the kinematic jointmodel See Figure 24 If the corresponding kinematic jointbetween any two parts cannot be recognized we can easilyanalyze the modelrsquos geometry constraints
6 Conclusions
This paper investigates a method for determining the kine-matic joints of assemblies restricted by arbitrary combina-tions of geometry constraints using reciprocal screw theoryWe achieve the process of recognition in terms of thebasic components of geometry constraints and kinematicjoints We describe the three basic components of geometricconstraints in terms of wrench At the same time the specificoperation of a matrix space is applied to eliminate the redun-dant components between different geometry constraintsthat act upon the same part A user of this method can auto-matically and easily convert any assembly model into a screwsystem in any CAD system Afterwards the screw system canbe easily used for validating kinematic functions analyzingover- or underconstrained assembly configurations and soon
Additional Points
(i) We provide a new method for converting any 3Dassembly model into a screw system based on recip-rocal screw theory
(ii) We describe the three basic components of geometricconstraints in terms of wrench
(iii) We eliminate the redundant components of differentgeometry constraints that act upon the same part bythe specific operation of a matrix space
(iv) We establish the maps between geometry constraintsand kinematic joints as the reciprocal product of atwist and wrench which is equal to zero
(v) We present several application examples
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
The authors would like to thank the Ministry of Scienceand Technology of the Peoplersquos Republic of China (nos2011ZX02403-005 2011CB706502 and 2013AA041301) Thiswork is also supported by the National Natural ScienceFoundation of China (nos 61370182 and 51175198)
References
[1] A H Slocum Precision Machine Design Prentice Hall NewYork NY USA 1991
[2] M Rajh S Glodez J Flasker K Gotlih and T KostanjevecldquoDesign and analysis of an fMRI compatible haptic robotrdquoRobotics and Computer-Integrated Manufacturing vol 27 no 2pp 267ndash275 2011
[3] S Awtar and G Parmar ldquoDesign of a large range XY nanoposi-tioning systemrdquo Journal of Mechanisms and Robotics vol 5 no2 Article ID 021008 2013
[4] R S Ball A Treatise on the Theory of Screws CambridgeUniversity Press Cambridge UK 1998
[5] K H Hunt Kinematic Geometry of Mechanisms ClarendonPress Oxford UK 1978
[6] K J Waldron ldquoThe constraint analysis of mechanismsrdquo Journalof Mechanisms vol 1 no 2 pp 101ndash114 1966
[7] R Konkar Incremental kinematic analysis and symbolic syn-thesis of mechanisms [PhD dissertation] Stanford UniversityStanford Calif USA 1993
[8] J-S Zhao Z-J Feng and J-X Dong ldquoComputation of theconfiguration degree of freedomof a spatial parallel mechanismby using reciprocal screw theoryrdquo Mechanism and MachineTheory vol 41 no 12 pp 1486ndash1504 2006
[9] J Eddie Baker ldquoScrew system algebra applied to special linkageconfigurationsrdquoMechanism and Machine Theory vol 15 no 4pp 255ndash265 1980
[10] M TMason and J K Salisbury Robot Hands and theMechanicsof Manipulation MIT Press Cambridge Mass USA 1985
[11] Y Fei and X Zhao ldquoAn assembly process modeling and analysisfor robotic multiple peg-in-holerdquo Journal of Intelligent andRobotic Systems vol 36 no 2 pp 175ndash189 2003
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
Mathematical Problems in Engineering 3
ISA
Pv
Figure 2 Prismatic joint
Therefore we can use a twist to characterize a componentof the kinematic joint between two parts Afterwards thekinematic joint can be transformed into a correspondingtwist matrix
22 Twist Representations of Kinematic Joints There are6 types of kinematic joints in common prismatic jointrevolute joint helical joint cylindrical joint spherical jointand planar joint In this section we use s to characterizethe allowed motions of each kinematic joint and show theircorresponding twist matrices
221 Prismatic Joint The degrees of freedom (dof) of a pris-matic joint (119875) is one We use the unit vector 119875v = (V
119909 V119910V119911)
to characterize the allowed translation See Figure 2Thus 119875s has the form
119875s = 119875v (2)
The twist matrix representation of 119875 is given by
T119875 = [0 119875s] (3)
We use 0 instead of [0 0 0] unless otherwise stated
222 Revolute Joint The dof of a revolute joint (119877) is oneWeuse 119877L to characterize the axis of 119877 The unit vector of 119877Lis characterized by 119877v = (V
119909 V119910 V119911) 119877P = (119875
119909 119875119910 119875119911) is a
single point on the 119877L See Figure 3Thus 119877L has the form
119877L = 119877P + 119886 sdot 119877v (4)
where 119886 is a real number 119877s has the form
119877s = 119877P times 119877v (5)
Then the twist matrix representation of 119877 is given by
T119877 = [119877v 119877s] (6)
ISA
RP
Pv
RL
Figure 3 Revolute joint
ISA
HL
HP
Hv
Hp
Figure 4 Helical joint
223 Helical Joint The dof of a helical joint (H) is one Weuse119867L to characterize the axis of119867 The unit vector of119867Lis characterized by 119867v = (V
119909 V119910 V119911) 119867P = (119875
119909 119875119910 119875119911) is a
single point on the119867L See Figure 4Thus119867L has the form
119867L = 119867P + 119886 sdot 119867v (7)
where 119886 is a real number119867s has the form
119867s = 119867P times 119867v + 119867119901 sdot 119867v (8)
where119867119901 is the pitch of119867 Then the twist matrix represen-tation of119867 is given by
T119867 = [119867v 119867s] (9)
224 Cylindrical Joint The dof of a cylindrical joint (119862) istwo We use 119862L to characterize the axis of 119862 The unit vectorof119862L is characterized by119862v = (V
119909 V119910 V119911)119862P = (119875
119909 119875119910 119875119911)
is a single point on the 119862L See Figure 5Thus 119862L has the form
119862L = 119862P + 119886 sdot 119862v (10)
4 Mathematical Problems in Engineering
ISA
CL
CP
Cv
Figure 5 Cylindrical joint
SP
ISA
v1
v2
v3
Figure 6 Spherical joint
where 119886 is a real number 119862s has two components as shownin
119862s1= 119862P times 119862v
119862s2= 119862v
(11)
Therefore the twist matrix representation of119862 is given by
T119862 = [119862v 119862s
1
0 119862s2
] (12)
225 Spherical Joint The dof of a spherical joint (119878) is threeWe use 119878P = (119875
119909 119875119910 119875119911) to characterize the central point of
119878 and define v1= (1 0 0) v
2= (0 1 0) and v
3= (0 0 1) See
Figure 6Thus 119878s has three components as shown in
119878s1= 119878P times v
1= (0 119875
119911 minus119875119910)
119878s2= 119878P times v
2= (minus119875
119911 0 119875119909)
119878s3= 119878P times v
3= (119875119910 minus119875119909 0)
(13)
EP
ISA
En
Ev1
Ev2
Figure 7 Planar joint
Then the twist matrix representation of 119878 is given by
T119878 = [[[
v1 119878s1
v2 119878s2
v3 119878s3
]
]
]
(14)
226 Planar Joint The dof of a planar joint (119864) is three Thevectors 119864v
1= (v1199091 v1199101 v1199111) and 119864v
2= (v1199092 v1199102 v1199112) are
orthogonal in the planar surface and the unit vector 119864n isthe outer normal vector 119864P = (119875
119909 119875119910 119875119911) is a single point
on the surface See Figure 7Thus 119864s has three components as shown in
119864s1= 119864v1
119864s2= 119864v2
119864s3= 119864P times 119864n
(15)
Then the twist matrix representation of 119864 is given by
T119864 = [[[
0 119864s1
0 119864s2
119864n 119864s3
]
]
]
(16)
3 Reciprocal Screw Models ofGeometric Constraints
To transformarbitrary combinations of geometric constraintsbetween two parts into the corresponding kinematic jointwe should use the same mathematical model with kinematicjoints or a model that can be easily transformed into a twistThis section shows how to define geometry constraints usingwrench
31 Wrench A wrench [32 33] takes the form
W = [119891119909119891119910119891119911 119898119909119898119910119898119911] (17)
A wrench is also a screw and describes the resultant forceand moment of a force system acting on a rigid body Anyset of forces and couples applied to a body can be reduced toa set comprising a single force acting along a specific line inspace and a pure couple acting in a plane perpendicular tothat line This means that the motion described by a wrench
Mathematical Problems in Engineering 5
z
x
yo
M
Force system
Instantaneous screw axis (ISA)
Fi
F1 F2
F3
Fr
Mr
ri
Line of action of Fr
Fr
Figure 8 Wrench
is forbidden The first triplet describes the resultant force ina global reference frame The second triplet represents theresultant moment of the force system about the origin of theglobal frame See Figure 8
Both a wrench and geometry constraint are structuralconstraints Therefore we can use a wrench to characterizea component of the geometry constraints between two partsAfterwards the geometry constraints can be transformedinto a corresponding wrench matrix
32 Basic Components of Geometric Constraints When wedescribe a part in space the orientation and position shouldbe specified [34] The orientation cannot be described inabsolute terms An orientation is given by a rotation fromsome known reference orientationThe amount of rotation isknown as an angular displacement In other words describ-ing an orientation is mathematically equivalent to describingan angular displacement In the same way when we specifythe position of a part we cannot do so in absolute terms wemust always do so within the context of a specific referenceframeTherefore specifying a position is actually the same asspecifying an amount of translation from a given referencepoint (usually the origin of some coordinate system)
In practice there are many geometry constraints suchas align and point on line We can use one or arbitrarycombinations of geometric constraints to restrict the relativeorientation and position between two parts As we knowcommon geometry constraints such as coaxial constraintsand spherical constraints usually have a high degree ofconstraint (doc) However these geometry constraints are allcombined by the three different types of basic constraintsdot-1 constraint dot-2 constraint and the distance constraint[35] which can be divided into two categories the orientationpart and the position part
Consider a pair of rigid bodies denoted as bodies 119894 and119895 in Figure 9 Reference points P
119894and P
119895 reference frames
f119894-g119894-h119894and f
119895-g119895-h119895 and nonzero unit vectors a
119894and a
119895
are fixed in bodies 119894 and 119895 respectively Vector d119894119895connects
z
x
y
i
j
ai
Pihi
gi
f i
ri
sPidij
fj
hjgj
sPj
Pj aj
rj
Figure 9 Vectors fixed in and between bodies
P119894and P
119895between bodies The basic constraints are often
characterized by such elements
321 Dot-1 Constraint Dot-1 constraint restricts the relativeorientation of a pair of bodies that is
Φ
1198891
(a119894 a119895 1198621) equiv a119894
Ta119895minus 1198621= 0 119862
1isin (minus1 1) (18)
where doc(Φ1198891) = 1 and dot-1 constraint is symmetric withregard to bodies 119894 and 119895 Note that dot-1 constraint would notbe a basic constraint if119862
1= plusmn1Thismeans that the vectors a
119894
and a119895are parallel in the same or opposite directions In this
case doc(Φ1198891(a119894 a119895 plusmn1)) = 2
Writing the vectors a119894and a119895in terms of their respective
body reference transformationmatricesA119894andA
119895and body-
fixed constant vectors a1015840119894and a1015840
119895 where a
119894= A119894a1015840119894and a
119895=
A119895a1015840119895 (18) has the form
Φ
1198891
(a119894 a119895 1198621) equiv a1015840119894
TA119894
TA119895a1015840119895minus 1198621= 0
1198621isin (minus1 1)
(19)
where the transformation matrices A119894and A
119895in (19) depend
on the orientation generalized coordinates of bodies 119894 and 119895respectively
As a result the wrench representation for dot-1 constraintshould also freeze the relative orientation of the vectors a
119894and
a119895 Therefore the wrench matrix is given by
W1198891 = (0 a119894times a119895) (20)
If we use the body reference transformationmatrices (20)has the form
W1198891 = (0 (a1015840119894
TA119894
T) times (A
119895a1015840119895)) (21)
322 Dot-2 Constraint Dot-2 constraint restricts the relativeposition of a pair of bodies that is
Φ
1198892
(a119894 d119894119895 1198622) equiv a119894
Td119894119895minus 1198622= 0 d
119894119895= 0 (22)
6 Mathematical Problems in Engineering
where doc(Φ1198892) = 1 the same as that with dot-1 constraintalthough dot-2 constraint is not symmetric with regard tobodies 119894 and 119895 Analytically it is important to recall that dot-2constraint breaks down if d
119894119895= 0 Write the vector d
119894119895as
d119894119895= r119895+ A119895s1015840119895
Pminus r119894minus A119894s1015840119894
P (23)
Equation (22) becomes
Φ
1198892
(a119894 d119894119895 1198622) equiv a1015840119894
TA119894
T(r119895+ A119895s1015840119895
Pminus r119894minus A119894s1015840119894
P)
minus 1198622= 0
(24)
As a result the wrench representation for dot-2 constraintmust freeze the relative translation along the vector a
119894
Therefore the wrench matrix is given by
W1198892 = (a119894 0) (25)
Equation (25) also has the form
W1198892 = (A119894a1015840119894 0) (26)
323 Distance Constraint In practice we often require a pairof points on two bodies to coincide or maintain a determineddistance that is
Φ
119904
(119875119894 119875119895 1198623) equiv d119894119895
Td119894119895minus 1198623sdot 1198623= 0 (27)
where doc(Φ119904) would be different depending on whether 1198623
is zero Therefore we have
doc (Φ119904) = 1 if 1198623
= 0
doc (Φ119904) = 3 if 1198623= 0
(28)
As a result the wrench representation for the distanceconstraint should have two types of expressions as shown in(29) and (30)
W119904 = [d119894119895 0] if 119862
3= 0 (29)
where the wrench representation for the distance constraintshould freeze the relative translation along the vector d
119894119895if
1198623
= 0 However if 1198623= 0 the wrench representation has
the form
W119904 = [[[
1 0 0 00 1 0 00 0 1 0
]
]
]
if 1198623= 0 (30)
where the wrench which is illustrated in (30) freezes all therelative translations along the axes 119909 119910 and 119911
33 Wrench Representations of Geometry Constraints Tointroduce geometry constraints we give Figure 10
In Figure 10 reference points P1and P
2 reference frames
f1-g1-h1and f2-g2-h2 and central axes L
1and L
2are fixed in
cylinders 1198611and 119861
2 respectively Vector d
12connects P
1and
P2between the cylinders
L1
B1
h1
P1
f1
F1 g1
L2
B2
h2
P2
f2F2
g2
d12
Figure 10 Instance of geometry constraints
The reference point P1has the form
P1= (1198751199091 1198751199101 1198751199111) (31)
The central axis L1has the form
L1= P1+ 119886 sdot h
1 (32)
where 119886 is a real number The planar surface F1has the form
F1= P1+ 119887 sdot f
1+ 119888 sdot g
1 (33)
where 119887 and 119888 are both real numbers At the same timethe elements P
2 L2 and F
2have the same forms The
following geometry constraints are often characterized bysuch elements
We can divide all different types of geometry constraintswhich have more than one doc into six categories coaxialconstraint coplanar constraint spherical constraint align con-straint distance between line and surface and point on lineWe have modeled the three basic components of geometryconstraints in terms of wrench Thus the wrench models ofthese six types of geometry constraints can be automaticallydeduced
331 Coaxial Constraint Weuse119862119900119894119871119871(L1 L2) to denote the
coaxial constraint of axes L1and L
2 as shown in Figure 11
The doc of 119862119900119894119871119871(L1 L2) is four which restricts two
relative orientations and two relative translations In otherwords it comprises two dot-1 constraints and two dot-2constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(f1 d12 0)
Φ4
1198892
(g1 d12 0)
(34)
Then the wrench matrix of 119862119900119894119871119871(L1 L2) is given by
W119862119900119894119871119871(L
1L2)=
[
[
[
[
[
[
0 h1times f2
0 h1times g2
f1 0
g1 0
]
]
]
]
]
]
(35)
Mathematical Problems in Engineering 7
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
CoiLL(L1 L2)
Figure 11 Coaxial constraint
f2
P2
h2
F2
g2
L2
B2
f1P1
h1
F1
g1
L1
B1
CoiFF(F1 F2)
d12
Figure 12 Coplanar constraint
The corresponding kinematic joint of coaxial constraint iscylindrical joint
332 Coplanar Constraint We use 119862119900119894119865119865(F1 F2) to denote
the coplanar constraint of planar surfaces F1and F
2 as shown
in Figure 12The doc of 119862119900119894119865119865(F
1 F2) is three which restricts two
relative orientations and one relative translation In otherwords it comprises two dot-1 constraints and one dot-2constraint as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(h1 d12 0) (36)
Then the wrench matrix of 119862119900119894119865119865(F1 F2) is given by
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
(37)
The corresponding kinematic joint of coplanar constraintis planar joint
333 Spherical Constraint We use 119862119900119894119875119875(P1P2) to denote
the spherical constraint of two points P1and P
2 as shown in
Figure 13
f2
P2
h2
F2 g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPP(P1P2)
Figure 13 Spherical constraint
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
ParFF(F1 F2)
Figure 14 Align constraint
The doc of 119862119900119894119875119875(P1P2) is three which restricts three
relative translations in three unrelated directions In otherwords it comprises one distance constraint which has 119862
3= 0
in Section 323 as shown in
Φ
119904
(119875119894 119875119895 0) (38)
Then the wrench matrix of 119862119900119894119875119875(P1P2) is given by
W119862119900119894119875119875(P
1P2)=
[
[
[
1 0 0 00 1 0 00 0 1 0
]
]
]
(39)
The corresponding kinematic joint of spherical constraintis spherical joint
334 Align Constraint Align constraint (119875119886119903119865119865) is used tomake two planar surfaces F
1and F
2parallel as shown in
Figure 14The doc of 119875119886119903119865119865(F
1 F2) is two which restricts two
relative orientations In other words it comprises two dot-1constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) (40)
8 Mathematical Problems in Engineering
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
DisLF(L1 F2 D)
Figure 15 Distance between line and surface
Then the wrench matrix of 119875119886119903119865119865(F1 F2) is given by
W119875119886119903119865119865(F
1F2)= [
0 h1times f2
0 h1times g2
] (41)
We cannot find a corresponding kinematic joint for alignconstraint among the common kinematic joints
335 Distance between Line and Surface We use 119863119894119904119871119865(L1
F2 119863) to denote the axis L
1and the surface F
2that maintain
a certain distance119863 as shown in Figure 15The doc of 119863119894119904119871119865(L
1 F2 119863) is two which restricts one
relative orientation and one relative translation In otherwords it comprises one dot-1 constraint and one dot-2constraint as shown in
Φ1
1198891
(h1 h2 0) Φ
2
1198892
(h2 d12 119863) (42)
Then the wrench matrix of119863119894119904119871119865(L1 F2 119863) is given by
W119863119894119904119871119865(L
1F2119863)
= [
0 h1times h2
h2 0
] (43)
We also cannot find a corresponding kinematic jointfor distance between line and surface among the commonkinematic joints
336 Point on Line Weuse119862119900119894119875119871(P1 L2) to denote the point
P1on line L
2 as shown in Figure 16
The doc of 119862119900119894119875119871(P1 L2) is two which restricts two
relative translations In other words it comprises two dot-2constraints as shown in
Φ1
1198892
(f2 d12 0) Φ
2
1198892
(g2 d12 0) (44)
Then the wrench matrix of 119862119900119894119875119871(P1 L2) is given by
W119863119894119904119871119865(L
1F2119863)
= [
f2 0
g2 0
] (45)
We still cannot find a corresponding kinematic joint forpoint on line among the common kinematic joints
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPL(P1 L2)
Figure 16 Point on line
4 Recognition of Kinematic Joints
The key to recognizing kinematic joints in terms of geometryconstraints is the reciprocity relation We show how to definegeometry constraints using wrench The space of screws thatdefines the range of wrenches that can be exerted upon abody and the space of screws that defines the range of twiststhat the body can perform without breaking any contacts inthe joints of the mechanism are mutually reciprocal screwspaces When we know the wrenches acting upon a body wecan determine the motions it can perform Conversely if weknow the motions a body is capable of performing then theconstraints that are exerted upon it can also be deduced
41 Algorithm We assume that 119860 is a part in a 3D assemblymodel restricted by parts 119861
1 1198612 119861
119899 The twist space of 119860
is the space of screws reciprocal to the wrenches of contactacting upon it The space of wrenches acting upon 119860 is theunion of all the individual wrenches acting upon it fromthe parts 119861
119894 These individual wrenches are to be considered
as acting alone Algorithm 1 provides the twist space of 119860denoted by T119860
The 119877119890119888119894119901119903119900119888119886119897(P) in Algorithm 1 means to exchangethe first triplet and the second triplet of P For exampleif P = [119886 119887 119888 119889 119890 119891] then we have 119877119890119888119894119901119903119900119888119886119897(P) =
[119889 119890 119891 119886 119887 119888] The recursive nature of the procedurerequires the computation of the wrench space and twistspace of each component This proves to be computationallyexpensive because wrench space and twist space are definedin terms of the reciprocal of one another The operation ofderiving reciprocal screws involves computing the null spaceof matrices The speed of the algorithm is linearly related tothe number of 119861
119894 That is the algorithm is of the order 119874(119899)
42 Reciprocal Product According to reciprocal screw the-ory two screws (a twist and a wrench) are said to be mutuallyreciprocal if their reciprocal product is zero The reciprocal
Mathematical Problems in Engineering 9
BEGINFOR each part Bi which restrict AConvert all the geometry constraints between A and Bi into basic components 119862119895
119860 119895 = 1 119897
Determine the wrenchesW119896119860 119896= 1 119898 acting upon A through C119895
119860 We havem not less than l
Collect all the wrenchesW119896119860into a matrixW
END FORCompute the orthogonal matrixQ in terms ofW by Gram-Schmidt orthogonalization [31]IF Rank(Q) = 6 THENReturn ldquoIMMOBILErdquo
ELSECompute the orthogonal complement matrix P of Q
END IFTA =Reciprocal(P)
END
Algorithm 1 Function twist space
f2
A2
h2F2
g2
B2
f1A1
F1
g1B1
Figure 17 Redundant constraint
product of a twist T1and a wrenchW
1thatisequal to zero is
given by
T1∘W1= T11minus3
sdotW24minus6
+ T14minus6
sdotW21minus3
= [0 0] (46)
Mathematically because reciprocity imposes the relationT1∘W1 we can compute one of these quantities if the other is
known This allows us to compute the twists that a body mayexecute while in contact with other bodies without breakingthe contacts
43 Redundant Constraint When more than one geometryconstraint acts upon the same part because some compo-nents of these constraints may be redundant the resultantmotion is the logical intersection of the individual wrenchmatrix that defines each constraint for example when wemodel a revolute joint 119877
12with two links 119861
1and 119861
2 The
constraints used are 119862119900119894119871119871(A1A2) and 119862119900119894119865119865(F
1 F2) See
Figure 17Also dof(119877
12) = 1Thismeans that doc(119861
1 1198612) = 5 How-
ever we have doc(119862119900119894119871119871(A1A2)) = 4 and doc(119862119900119894119865119865(F
1
F2)) = 3 In other words the dimension of the orthog-
onal wrench space which comprises W119862119900119894119871119871(A
1A2)and
f2
h2
F2g2
B2
f1h1
g1
B1
F1
Figure 18 Recognition of planar joint
W119862119900119894119865119865(F
1F2) is 5 Therefore the motions that a body is
capable of performing while under the action of multiplewrenches are the logical intersection of the motions it canperform under the action of each wrench acting alone Thuswe can eliminate the redundant components of differentgeometry constraints that act upon the same part by usingGram-Schmidt orthogonalization in Algorithm 1
5 Examples
In each of the examples below different types of kinematicjoints are recognized
Example 1 (planar joint) Figure 18 shows a planar jointconsisting of two parts 119861
1and 119861
2 and a coplanar constraint
119862119900119894119865119865(F1 F2)
Following the process outlined in Section 332 we findbased on the definition of the coplanar constraint that thewrench matrix is
W119862119900119894119865119865(F
1F2)=
[
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
]
(47)
10 Mathematical Problems in Engineering
f2h2
F2
g2
B2
f1
h1 g1
B1
F1
f 9984002
h9984002g9984002
F9984002
f 9984001
h9984001
g9984001
F9984001
Figure 19 Recognition of prismatic joint
whereW119862119900119894119865119865(F
1F2)is already an orthogonalmatrixThen the
orthogonal complement matrix P119862119900119894119865119865(F
1F2)is
P119862119900119894119865119865(F
1F2)=
[
[
[
f1 0
g1 0
0 h1
]
]
]
(48)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119865119865(F1F2)) =
[
[
[
0 f1
0 g1
h1 0
]
]
]
(49)
where T12
represents the motion of a planar joint which isillustrated in Section 226 Moreover the recognition processof a cylindrical joint or spherical joint is approximately thesame as that of the planar joint
Example 2 (prismatic joint) Figure 19 shows a prismaticjoint consisting of two parts 119861
1and 119861
2 and two coplanar
constraints 119862119900119894119865119865(F1 F2) and 119862119900119894119865119865(F1015840
1 F10158402)
Following the process outlined in Example 1 we find thatthe wrench matrices between 119861
1and 119861
2are
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
W119862119900119894119865119865(F1015840
1F10158402)=
[
[
[
[
0 h10158401times f10158402
0 h10158401times g10158402
h10158401 0
]
]
]
]
(50)
B0
B1
B2
B3
B4
B5B6
B7
B8
B9
B10
B11
x
y
Figure 20 Hydraulic grab
where W12
= W119862119900119894119865119865(F
1F2)cup W119862119900119894119865119865(F1015840
1F10158402) The orthogonal
matrixQ12ofW12is
Q12=
[
[
[
[
[
[
[
[
[
[
[
0 h1times f2
0 h1times g2
0 h10158401times g10158402
h1 0
h10158401 0
]
]
]
]
]
]
]
]
]
]
]
(51)
The orthogonal complement matrix P12ofQ12is
P12= [g1 0] (52)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
12) = [0 g
1] (53)
where T12
represents the motion of a prismatic joint whichis shown in Section 221 Moreover the recognition processof a revolute joint is approximately the same as that of theprismatic joint
Example 3 (hydraulic grab) Here we present a projectexample A hydraulic grab with three degrees of freedom inCartesian coordinates is illustrated in Figure 20
The hydraulic grab comprises 12 parts 1198610 1198611 1198612 1198613
1198614 1198615 1198616 1198617 1198618 1198619 11986110 and 119861
11 12 coplanar constraints
119862119900119894119865119865(1198610 1198612)119862119900119894119865119865(119861
0 1198611)119862119900119894119865119865(119861
1 1198613)119862119900119894119865119865(119861
1 1198615)
119862119900119894119865119865(1198611 1198614) 119862119900119894119865119865(119861
4 1198616) 119862119900119894119865119865(119861
4 11986111) 119862119900119894119865119865(119861
4
1198617) 119862119900119894119865119865(119861
4 1198618) 119862119900119894119865119865(119861
9 11986110) 119862119900119894119865119865(119861
7 11986110) and
119862119900119894119865119865(11986110 11986111) and 15 coaxial constraints 119862119900119894119871119871(119861
0 1198612)
119862119900119894119871119871(1198610 1198611) 119862119900119894119871119871(119861
1 1198613) 119862119900119894119871119871(119861
2 1198613) 119862119900119894119871119871(119861
1 1198615)
119862119900119894119871119871(1198611 1198614) 119862119900119894119871119871(119861
4 1198616) 119862119900119894119871119871(119861
5 1198616) 119862119900119894119871119871(119861
4
11986111) 119862119900119894119871119871(119861
4 1198617) 119862119900119894119871119871(119861
4 1198618) 119862119900119894119871119871(119861
8 1198619)
119862119900119894119871119871(1198619 11986110) 119862119900119894119871119871(119861
7 11986110) and 119862119900119894119871119871(119861
10 11986111) See
Figure 21Here we present the processes of recognition among 119861
1
1198612 and 119861
3 as shown in Figure 22 The geometry constraint
between 1198612and 119861
3is 119862119900119894119871119871(119861
2 1198613) Following the process
Mathematical Problems in Engineering 11
CoiLL
CoiFF
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
Figure 21 The constraint graph of a hydraulic grab
h2 g2
f 9984003
g9984003
B2
B1
B0
g3h3
g9984001
f 9984001
B3
x
y
Figure 22 Parts 1198611 1198612 and 119861
3in a hydraulic grab
outlined in Section 331 we find based on the definition ofthe coaxial constraint that the wrench matrix is
W119862119900119894119871119871(119861
21198613)=
[
[
[
[
[
[
0 h2times f3
0 h2times g3
f2 0
g2 0
]
]
]
]
]
]
(54)
The orthogonal complement matrix is
P119862119900119894119871119871(119861
21198613)= [
0 h2
h2 0
] (55)
Then the twist matrix T23
between 1198612and 119861
3has the
following form
T23= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119871119871(11986121198613)) = [
h2 0
0 h2
] (56)
The geometry constraints between 1198611
and 1198613
are119862119900119894119871119871(119861
1 1198613) and 119862119900119894119865119865(119861
1 1198613) Thus the wrench matrices
between 1198611and 119861
3are
W119862119900119894119871119871(119861
11198613)=
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
]
]
]
]
]
]
]
W119862119900119894119865119865(119861
11198613)=
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
h10158403 0
]
]
]
]
(57)
where W13
= W119862119900119894119871119871(119861
11198613)cup W119862119900119894119865119865(119861
11198613) The orthogonal
matrixQ13ofW13is given by
Q13=
[
[
[
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
h10158403 0
]
]
]
]
]
]
]
]
]
]
(58)
Then the orthogonal complement matrix P13ofQ13is
P13= [0 h1015840
3] (59)
Therefore the twist matrix T13between 119861
1and 119861
3is
T13= 119877119890119888119894119901119903119900119888119886119897 (P
13) = [h1015840
3 0] (60)
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system basedon reciprocal screw theory According to the geometryconstraints between any two parts in the hydraulic graband Algorithm 1 we obtain the screw system as shown inFigure 23
As mentioned in [4ndash29] we can perform some differenttypes of kinematic analyses using the screw system such asdetermining the degrees of freedom of 119861
1and 119861
4 that is
T1198611
= T10cap (T20cup T32cup T31)
T1198614
= (T1198611
cup T41) cap (T
1198611
cup T51cup T65cup T64)
(61)
where the resultant of the motions that the end-effector ofa purely serial chain may perform due to the action of eachjoint acting alone is the union of that set and the resultant ofthe set of twists allowed separately by multiple contacts upona body is the intersection of the individual twists Thus thedegrees of freedom of 119861
1are equal to the dimension of the
orthogonal twist space of1198611 denoted byT
1198611
In the samewaywe can obtain the degrees of freedomof119861
4and any part in the
hydraulic grab
12 Mathematical Problems in Engineering
T twist
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
T32
T31
T10
T20
T109
T1110
T114
T98
T74 T8
4
T41
T64
T65
T51
T107
Figure 23 The screw system of a hydraulic grab
R revolute jointC cylindrical joint
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
R31
R10
R20
1110
R114
C98
R74 R8
4
R41
R64 R5
1
R107
R
R109 C3
2
C65
Figure 24 The kinematic joint model of a hydraulic grab
According to Sections 222 and 224 the kinematic jointsbetween 119861
2and 119861
3and 119861
1and 119861
3are the cylindrical joint and
revolute joint As a result we can obtain the kinematic jointmodel See Figure 24 If the corresponding kinematic jointbetween any two parts cannot be recognized we can easilyanalyze the modelrsquos geometry constraints
6 Conclusions
This paper investigates a method for determining the kine-matic joints of assemblies restricted by arbitrary combina-tions of geometry constraints using reciprocal screw theoryWe achieve the process of recognition in terms of thebasic components of geometry constraints and kinematicjoints We describe the three basic components of geometricconstraints in terms of wrench At the same time the specificoperation of a matrix space is applied to eliminate the redun-dant components between different geometry constraintsthat act upon the same part A user of this method can auto-matically and easily convert any assembly model into a screwsystem in any CAD system Afterwards the screw system canbe easily used for validating kinematic functions analyzingover- or underconstrained assembly configurations and soon
Additional Points
(i) We provide a new method for converting any 3Dassembly model into a screw system based on recip-rocal screw theory
(ii) We describe the three basic components of geometricconstraints in terms of wrench
(iii) We eliminate the redundant components of differentgeometry constraints that act upon the same part bythe specific operation of a matrix space
(iv) We establish the maps between geometry constraintsand kinematic joints as the reciprocal product of atwist and wrench which is equal to zero
(v) We present several application examples
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
The authors would like to thank the Ministry of Scienceand Technology of the Peoplersquos Republic of China (nos2011ZX02403-005 2011CB706502 and 2013AA041301) Thiswork is also supported by the National Natural ScienceFoundation of China (nos 61370182 and 51175198)
References
[1] A H Slocum Precision Machine Design Prentice Hall NewYork NY USA 1991
[2] M Rajh S Glodez J Flasker K Gotlih and T KostanjevecldquoDesign and analysis of an fMRI compatible haptic robotrdquoRobotics and Computer-Integrated Manufacturing vol 27 no 2pp 267ndash275 2011
[3] S Awtar and G Parmar ldquoDesign of a large range XY nanoposi-tioning systemrdquo Journal of Mechanisms and Robotics vol 5 no2 Article ID 021008 2013
[4] R S Ball A Treatise on the Theory of Screws CambridgeUniversity Press Cambridge UK 1998
[5] K H Hunt Kinematic Geometry of Mechanisms ClarendonPress Oxford UK 1978
[6] K J Waldron ldquoThe constraint analysis of mechanismsrdquo Journalof Mechanisms vol 1 no 2 pp 101ndash114 1966
[7] R Konkar Incremental kinematic analysis and symbolic syn-thesis of mechanisms [PhD dissertation] Stanford UniversityStanford Calif USA 1993
[8] J-S Zhao Z-J Feng and J-X Dong ldquoComputation of theconfiguration degree of freedomof a spatial parallel mechanismby using reciprocal screw theoryrdquo Mechanism and MachineTheory vol 41 no 12 pp 1486ndash1504 2006
[9] J Eddie Baker ldquoScrew system algebra applied to special linkageconfigurationsrdquoMechanism and Machine Theory vol 15 no 4pp 255ndash265 1980
[10] M TMason and J K Salisbury Robot Hands and theMechanicsof Manipulation MIT Press Cambridge Mass USA 1985
[11] Y Fei and X Zhao ldquoAn assembly process modeling and analysisfor robotic multiple peg-in-holerdquo Journal of Intelligent andRobotic Systems vol 36 no 2 pp 175ndash189 2003
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
4 Mathematical Problems in Engineering
ISA
CL
CP
Cv
Figure 5 Cylindrical joint
SP
ISA
v1
v2
v3
Figure 6 Spherical joint
where 119886 is a real number 119862s has two components as shownin
119862s1= 119862P times 119862v
119862s2= 119862v
(11)
Therefore the twist matrix representation of119862 is given by
T119862 = [119862v 119862s
1
0 119862s2
] (12)
225 Spherical Joint The dof of a spherical joint (119878) is threeWe use 119878P = (119875
119909 119875119910 119875119911) to characterize the central point of
119878 and define v1= (1 0 0) v
2= (0 1 0) and v
3= (0 0 1) See
Figure 6Thus 119878s has three components as shown in
119878s1= 119878P times v
1= (0 119875
119911 minus119875119910)
119878s2= 119878P times v
2= (minus119875
119911 0 119875119909)
119878s3= 119878P times v
3= (119875119910 minus119875119909 0)
(13)
EP
ISA
En
Ev1
Ev2
Figure 7 Planar joint
Then the twist matrix representation of 119878 is given by
T119878 = [[[
v1 119878s1
v2 119878s2
v3 119878s3
]
]
]
(14)
226 Planar Joint The dof of a planar joint (119864) is three Thevectors 119864v
1= (v1199091 v1199101 v1199111) and 119864v
2= (v1199092 v1199102 v1199112) are
orthogonal in the planar surface and the unit vector 119864n isthe outer normal vector 119864P = (119875
119909 119875119910 119875119911) is a single point
on the surface See Figure 7Thus 119864s has three components as shown in
119864s1= 119864v1
119864s2= 119864v2
119864s3= 119864P times 119864n
(15)
Then the twist matrix representation of 119864 is given by
T119864 = [[[
0 119864s1
0 119864s2
119864n 119864s3
]
]
]
(16)
3 Reciprocal Screw Models ofGeometric Constraints
To transformarbitrary combinations of geometric constraintsbetween two parts into the corresponding kinematic jointwe should use the same mathematical model with kinematicjoints or a model that can be easily transformed into a twistThis section shows how to define geometry constraints usingwrench
31 Wrench A wrench [32 33] takes the form
W = [119891119909119891119910119891119911 119898119909119898119910119898119911] (17)
A wrench is also a screw and describes the resultant forceand moment of a force system acting on a rigid body Anyset of forces and couples applied to a body can be reduced toa set comprising a single force acting along a specific line inspace and a pure couple acting in a plane perpendicular tothat line This means that the motion described by a wrench
Mathematical Problems in Engineering 5
z
x
yo
M
Force system
Instantaneous screw axis (ISA)
Fi
F1 F2
F3
Fr
Mr
ri
Line of action of Fr
Fr
Figure 8 Wrench
is forbidden The first triplet describes the resultant force ina global reference frame The second triplet represents theresultant moment of the force system about the origin of theglobal frame See Figure 8
Both a wrench and geometry constraint are structuralconstraints Therefore we can use a wrench to characterizea component of the geometry constraints between two partsAfterwards the geometry constraints can be transformedinto a corresponding wrench matrix
32 Basic Components of Geometric Constraints When wedescribe a part in space the orientation and position shouldbe specified [34] The orientation cannot be described inabsolute terms An orientation is given by a rotation fromsome known reference orientationThe amount of rotation isknown as an angular displacement In other words describ-ing an orientation is mathematically equivalent to describingan angular displacement In the same way when we specifythe position of a part we cannot do so in absolute terms wemust always do so within the context of a specific referenceframeTherefore specifying a position is actually the same asspecifying an amount of translation from a given referencepoint (usually the origin of some coordinate system)
In practice there are many geometry constraints suchas align and point on line We can use one or arbitrarycombinations of geometric constraints to restrict the relativeorientation and position between two parts As we knowcommon geometry constraints such as coaxial constraintsand spherical constraints usually have a high degree ofconstraint (doc) However these geometry constraints are allcombined by the three different types of basic constraintsdot-1 constraint dot-2 constraint and the distance constraint[35] which can be divided into two categories the orientationpart and the position part
Consider a pair of rigid bodies denoted as bodies 119894 and119895 in Figure 9 Reference points P
119894and P
119895 reference frames
f119894-g119894-h119894and f
119895-g119895-h119895 and nonzero unit vectors a
119894and a
119895
are fixed in bodies 119894 and 119895 respectively Vector d119894119895connects
z
x
y
i
j
ai
Pihi
gi
f i
ri
sPidij
fj
hjgj
sPj
Pj aj
rj
Figure 9 Vectors fixed in and between bodies
P119894and P
119895between bodies The basic constraints are often
characterized by such elements
321 Dot-1 Constraint Dot-1 constraint restricts the relativeorientation of a pair of bodies that is
Φ
1198891
(a119894 a119895 1198621) equiv a119894
Ta119895minus 1198621= 0 119862
1isin (minus1 1) (18)
where doc(Φ1198891) = 1 and dot-1 constraint is symmetric withregard to bodies 119894 and 119895 Note that dot-1 constraint would notbe a basic constraint if119862
1= plusmn1Thismeans that the vectors a
119894
and a119895are parallel in the same or opposite directions In this
case doc(Φ1198891(a119894 a119895 plusmn1)) = 2
Writing the vectors a119894and a119895in terms of their respective
body reference transformationmatricesA119894andA
119895and body-
fixed constant vectors a1015840119894and a1015840
119895 where a
119894= A119894a1015840119894and a
119895=
A119895a1015840119895 (18) has the form
Φ
1198891
(a119894 a119895 1198621) equiv a1015840119894
TA119894
TA119895a1015840119895minus 1198621= 0
1198621isin (minus1 1)
(19)
where the transformation matrices A119894and A
119895in (19) depend
on the orientation generalized coordinates of bodies 119894 and 119895respectively
As a result the wrench representation for dot-1 constraintshould also freeze the relative orientation of the vectors a
119894and
a119895 Therefore the wrench matrix is given by
W1198891 = (0 a119894times a119895) (20)
If we use the body reference transformationmatrices (20)has the form
W1198891 = (0 (a1015840119894
TA119894
T) times (A
119895a1015840119895)) (21)
322 Dot-2 Constraint Dot-2 constraint restricts the relativeposition of a pair of bodies that is
Φ
1198892
(a119894 d119894119895 1198622) equiv a119894
Td119894119895minus 1198622= 0 d
119894119895= 0 (22)
6 Mathematical Problems in Engineering
where doc(Φ1198892) = 1 the same as that with dot-1 constraintalthough dot-2 constraint is not symmetric with regard tobodies 119894 and 119895 Analytically it is important to recall that dot-2constraint breaks down if d
119894119895= 0 Write the vector d
119894119895as
d119894119895= r119895+ A119895s1015840119895
Pminus r119894minus A119894s1015840119894
P (23)
Equation (22) becomes
Φ
1198892
(a119894 d119894119895 1198622) equiv a1015840119894
TA119894
T(r119895+ A119895s1015840119895
Pminus r119894minus A119894s1015840119894
P)
minus 1198622= 0
(24)
As a result the wrench representation for dot-2 constraintmust freeze the relative translation along the vector a
119894
Therefore the wrench matrix is given by
W1198892 = (a119894 0) (25)
Equation (25) also has the form
W1198892 = (A119894a1015840119894 0) (26)
323 Distance Constraint In practice we often require a pairof points on two bodies to coincide or maintain a determineddistance that is
Φ
119904
(119875119894 119875119895 1198623) equiv d119894119895
Td119894119895minus 1198623sdot 1198623= 0 (27)
where doc(Φ119904) would be different depending on whether 1198623
is zero Therefore we have
doc (Φ119904) = 1 if 1198623
= 0
doc (Φ119904) = 3 if 1198623= 0
(28)
As a result the wrench representation for the distanceconstraint should have two types of expressions as shown in(29) and (30)
W119904 = [d119894119895 0] if 119862
3= 0 (29)
where the wrench representation for the distance constraintshould freeze the relative translation along the vector d
119894119895if
1198623
= 0 However if 1198623= 0 the wrench representation has
the form
W119904 = [[[
1 0 0 00 1 0 00 0 1 0
]
]
]
if 1198623= 0 (30)
where the wrench which is illustrated in (30) freezes all therelative translations along the axes 119909 119910 and 119911
33 Wrench Representations of Geometry Constraints Tointroduce geometry constraints we give Figure 10
In Figure 10 reference points P1and P
2 reference frames
f1-g1-h1and f2-g2-h2 and central axes L
1and L
2are fixed in
cylinders 1198611and 119861
2 respectively Vector d
12connects P
1and
P2between the cylinders
L1
B1
h1
P1
f1
F1 g1
L2
B2
h2
P2
f2F2
g2
d12
Figure 10 Instance of geometry constraints
The reference point P1has the form
P1= (1198751199091 1198751199101 1198751199111) (31)
The central axis L1has the form
L1= P1+ 119886 sdot h
1 (32)
where 119886 is a real number The planar surface F1has the form
F1= P1+ 119887 sdot f
1+ 119888 sdot g
1 (33)
where 119887 and 119888 are both real numbers At the same timethe elements P
2 L2 and F
2have the same forms The
following geometry constraints are often characterized bysuch elements
We can divide all different types of geometry constraintswhich have more than one doc into six categories coaxialconstraint coplanar constraint spherical constraint align con-straint distance between line and surface and point on lineWe have modeled the three basic components of geometryconstraints in terms of wrench Thus the wrench models ofthese six types of geometry constraints can be automaticallydeduced
331 Coaxial Constraint Weuse119862119900119894119871119871(L1 L2) to denote the
coaxial constraint of axes L1and L
2 as shown in Figure 11
The doc of 119862119900119894119871119871(L1 L2) is four which restricts two
relative orientations and two relative translations In otherwords it comprises two dot-1 constraints and two dot-2constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(f1 d12 0)
Φ4
1198892
(g1 d12 0)
(34)
Then the wrench matrix of 119862119900119894119871119871(L1 L2) is given by
W119862119900119894119871119871(L
1L2)=
[
[
[
[
[
[
0 h1times f2
0 h1times g2
f1 0
g1 0
]
]
]
]
]
]
(35)
Mathematical Problems in Engineering 7
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
CoiLL(L1 L2)
Figure 11 Coaxial constraint
f2
P2
h2
F2
g2
L2
B2
f1P1
h1
F1
g1
L1
B1
CoiFF(F1 F2)
d12
Figure 12 Coplanar constraint
The corresponding kinematic joint of coaxial constraint iscylindrical joint
332 Coplanar Constraint We use 119862119900119894119865119865(F1 F2) to denote
the coplanar constraint of planar surfaces F1and F
2 as shown
in Figure 12The doc of 119862119900119894119865119865(F
1 F2) is three which restricts two
relative orientations and one relative translation In otherwords it comprises two dot-1 constraints and one dot-2constraint as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(h1 d12 0) (36)
Then the wrench matrix of 119862119900119894119865119865(F1 F2) is given by
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
(37)
The corresponding kinematic joint of coplanar constraintis planar joint
333 Spherical Constraint We use 119862119900119894119875119875(P1P2) to denote
the spherical constraint of two points P1and P
2 as shown in
Figure 13
f2
P2
h2
F2 g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPP(P1P2)
Figure 13 Spherical constraint
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
ParFF(F1 F2)
Figure 14 Align constraint
The doc of 119862119900119894119875119875(P1P2) is three which restricts three
relative translations in three unrelated directions In otherwords it comprises one distance constraint which has 119862
3= 0
in Section 323 as shown in
Φ
119904
(119875119894 119875119895 0) (38)
Then the wrench matrix of 119862119900119894119875119875(P1P2) is given by
W119862119900119894119875119875(P
1P2)=
[
[
[
1 0 0 00 1 0 00 0 1 0
]
]
]
(39)
The corresponding kinematic joint of spherical constraintis spherical joint
334 Align Constraint Align constraint (119875119886119903119865119865) is used tomake two planar surfaces F
1and F
2parallel as shown in
Figure 14The doc of 119875119886119903119865119865(F
1 F2) is two which restricts two
relative orientations In other words it comprises two dot-1constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) (40)
8 Mathematical Problems in Engineering
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
DisLF(L1 F2 D)
Figure 15 Distance between line and surface
Then the wrench matrix of 119875119886119903119865119865(F1 F2) is given by
W119875119886119903119865119865(F
1F2)= [
0 h1times f2
0 h1times g2
] (41)
We cannot find a corresponding kinematic joint for alignconstraint among the common kinematic joints
335 Distance between Line and Surface We use 119863119894119904119871119865(L1
F2 119863) to denote the axis L
1and the surface F
2that maintain
a certain distance119863 as shown in Figure 15The doc of 119863119894119904119871119865(L
1 F2 119863) is two which restricts one
relative orientation and one relative translation In otherwords it comprises one dot-1 constraint and one dot-2constraint as shown in
Φ1
1198891
(h1 h2 0) Φ
2
1198892
(h2 d12 119863) (42)
Then the wrench matrix of119863119894119904119871119865(L1 F2 119863) is given by
W119863119894119904119871119865(L
1F2119863)
= [
0 h1times h2
h2 0
] (43)
We also cannot find a corresponding kinematic jointfor distance between line and surface among the commonkinematic joints
336 Point on Line Weuse119862119900119894119875119871(P1 L2) to denote the point
P1on line L
2 as shown in Figure 16
The doc of 119862119900119894119875119871(P1 L2) is two which restricts two
relative translations In other words it comprises two dot-2constraints as shown in
Φ1
1198892
(f2 d12 0) Φ
2
1198892
(g2 d12 0) (44)
Then the wrench matrix of 119862119900119894119875119871(P1 L2) is given by
W119863119894119904119871119865(L
1F2119863)
= [
f2 0
g2 0
] (45)
We still cannot find a corresponding kinematic joint forpoint on line among the common kinematic joints
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPL(P1 L2)
Figure 16 Point on line
4 Recognition of Kinematic Joints
The key to recognizing kinematic joints in terms of geometryconstraints is the reciprocity relation We show how to definegeometry constraints using wrench The space of screws thatdefines the range of wrenches that can be exerted upon abody and the space of screws that defines the range of twiststhat the body can perform without breaking any contacts inthe joints of the mechanism are mutually reciprocal screwspaces When we know the wrenches acting upon a body wecan determine the motions it can perform Conversely if weknow the motions a body is capable of performing then theconstraints that are exerted upon it can also be deduced
41 Algorithm We assume that 119860 is a part in a 3D assemblymodel restricted by parts 119861
1 1198612 119861
119899 The twist space of 119860
is the space of screws reciprocal to the wrenches of contactacting upon it The space of wrenches acting upon 119860 is theunion of all the individual wrenches acting upon it fromthe parts 119861
119894 These individual wrenches are to be considered
as acting alone Algorithm 1 provides the twist space of 119860denoted by T119860
The 119877119890119888119894119901119903119900119888119886119897(P) in Algorithm 1 means to exchangethe first triplet and the second triplet of P For exampleif P = [119886 119887 119888 119889 119890 119891] then we have 119877119890119888119894119901119903119900119888119886119897(P) =
[119889 119890 119891 119886 119887 119888] The recursive nature of the procedurerequires the computation of the wrench space and twistspace of each component This proves to be computationallyexpensive because wrench space and twist space are definedin terms of the reciprocal of one another The operation ofderiving reciprocal screws involves computing the null spaceof matrices The speed of the algorithm is linearly related tothe number of 119861
119894 That is the algorithm is of the order 119874(119899)
42 Reciprocal Product According to reciprocal screw the-ory two screws (a twist and a wrench) are said to be mutuallyreciprocal if their reciprocal product is zero The reciprocal
Mathematical Problems in Engineering 9
BEGINFOR each part Bi which restrict AConvert all the geometry constraints between A and Bi into basic components 119862119895
119860 119895 = 1 119897
Determine the wrenchesW119896119860 119896= 1 119898 acting upon A through C119895
119860 We havem not less than l
Collect all the wrenchesW119896119860into a matrixW
END FORCompute the orthogonal matrixQ in terms ofW by Gram-Schmidt orthogonalization [31]IF Rank(Q) = 6 THENReturn ldquoIMMOBILErdquo
ELSECompute the orthogonal complement matrix P of Q
END IFTA =Reciprocal(P)
END
Algorithm 1 Function twist space
f2
A2
h2F2
g2
B2
f1A1
F1
g1B1
Figure 17 Redundant constraint
product of a twist T1and a wrenchW
1thatisequal to zero is
given by
T1∘W1= T11minus3
sdotW24minus6
+ T14minus6
sdotW21minus3
= [0 0] (46)
Mathematically because reciprocity imposes the relationT1∘W1 we can compute one of these quantities if the other is
known This allows us to compute the twists that a body mayexecute while in contact with other bodies without breakingthe contacts
43 Redundant Constraint When more than one geometryconstraint acts upon the same part because some compo-nents of these constraints may be redundant the resultantmotion is the logical intersection of the individual wrenchmatrix that defines each constraint for example when wemodel a revolute joint 119877
12with two links 119861
1and 119861
2 The
constraints used are 119862119900119894119871119871(A1A2) and 119862119900119894119865119865(F
1 F2) See
Figure 17Also dof(119877
12) = 1Thismeans that doc(119861
1 1198612) = 5 How-
ever we have doc(119862119900119894119871119871(A1A2)) = 4 and doc(119862119900119894119865119865(F
1
F2)) = 3 In other words the dimension of the orthog-
onal wrench space which comprises W119862119900119894119871119871(A
1A2)and
f2
h2
F2g2
B2
f1h1
g1
B1
F1
Figure 18 Recognition of planar joint
W119862119900119894119865119865(F
1F2) is 5 Therefore the motions that a body is
capable of performing while under the action of multiplewrenches are the logical intersection of the motions it canperform under the action of each wrench acting alone Thuswe can eliminate the redundant components of differentgeometry constraints that act upon the same part by usingGram-Schmidt orthogonalization in Algorithm 1
5 Examples
In each of the examples below different types of kinematicjoints are recognized
Example 1 (planar joint) Figure 18 shows a planar jointconsisting of two parts 119861
1and 119861
2 and a coplanar constraint
119862119900119894119865119865(F1 F2)
Following the process outlined in Section 332 we findbased on the definition of the coplanar constraint that thewrench matrix is
W119862119900119894119865119865(F
1F2)=
[
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
]
(47)
10 Mathematical Problems in Engineering
f2h2
F2
g2
B2
f1
h1 g1
B1
F1
f 9984002
h9984002g9984002
F9984002
f 9984001
h9984001
g9984001
F9984001
Figure 19 Recognition of prismatic joint
whereW119862119900119894119865119865(F
1F2)is already an orthogonalmatrixThen the
orthogonal complement matrix P119862119900119894119865119865(F
1F2)is
P119862119900119894119865119865(F
1F2)=
[
[
[
f1 0
g1 0
0 h1
]
]
]
(48)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119865119865(F1F2)) =
[
[
[
0 f1
0 g1
h1 0
]
]
]
(49)
where T12
represents the motion of a planar joint which isillustrated in Section 226 Moreover the recognition processof a cylindrical joint or spherical joint is approximately thesame as that of the planar joint
Example 2 (prismatic joint) Figure 19 shows a prismaticjoint consisting of two parts 119861
1and 119861
2 and two coplanar
constraints 119862119900119894119865119865(F1 F2) and 119862119900119894119865119865(F1015840
1 F10158402)
Following the process outlined in Example 1 we find thatthe wrench matrices between 119861
1and 119861
2are
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
W119862119900119894119865119865(F1015840
1F10158402)=
[
[
[
[
0 h10158401times f10158402
0 h10158401times g10158402
h10158401 0
]
]
]
]
(50)
B0
B1
B2
B3
B4
B5B6
B7
B8
B9
B10
B11
x
y
Figure 20 Hydraulic grab
where W12
= W119862119900119894119865119865(F
1F2)cup W119862119900119894119865119865(F1015840
1F10158402) The orthogonal
matrixQ12ofW12is
Q12=
[
[
[
[
[
[
[
[
[
[
[
0 h1times f2
0 h1times g2
0 h10158401times g10158402
h1 0
h10158401 0
]
]
]
]
]
]
]
]
]
]
]
(51)
The orthogonal complement matrix P12ofQ12is
P12= [g1 0] (52)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
12) = [0 g
1] (53)
where T12
represents the motion of a prismatic joint whichis shown in Section 221 Moreover the recognition processof a revolute joint is approximately the same as that of theprismatic joint
Example 3 (hydraulic grab) Here we present a projectexample A hydraulic grab with three degrees of freedom inCartesian coordinates is illustrated in Figure 20
The hydraulic grab comprises 12 parts 1198610 1198611 1198612 1198613
1198614 1198615 1198616 1198617 1198618 1198619 11986110 and 119861
11 12 coplanar constraints
119862119900119894119865119865(1198610 1198612)119862119900119894119865119865(119861
0 1198611)119862119900119894119865119865(119861
1 1198613)119862119900119894119865119865(119861
1 1198615)
119862119900119894119865119865(1198611 1198614) 119862119900119894119865119865(119861
4 1198616) 119862119900119894119865119865(119861
4 11986111) 119862119900119894119865119865(119861
4
1198617) 119862119900119894119865119865(119861
4 1198618) 119862119900119894119865119865(119861
9 11986110) 119862119900119894119865119865(119861
7 11986110) and
119862119900119894119865119865(11986110 11986111) and 15 coaxial constraints 119862119900119894119871119871(119861
0 1198612)
119862119900119894119871119871(1198610 1198611) 119862119900119894119871119871(119861
1 1198613) 119862119900119894119871119871(119861
2 1198613) 119862119900119894119871119871(119861
1 1198615)
119862119900119894119871119871(1198611 1198614) 119862119900119894119871119871(119861
4 1198616) 119862119900119894119871119871(119861
5 1198616) 119862119900119894119871119871(119861
4
11986111) 119862119900119894119871119871(119861
4 1198617) 119862119900119894119871119871(119861
4 1198618) 119862119900119894119871119871(119861
8 1198619)
119862119900119894119871119871(1198619 11986110) 119862119900119894119871119871(119861
7 11986110) and 119862119900119894119871119871(119861
10 11986111) See
Figure 21Here we present the processes of recognition among 119861
1
1198612 and 119861
3 as shown in Figure 22 The geometry constraint
between 1198612and 119861
3is 119862119900119894119871119871(119861
2 1198613) Following the process
Mathematical Problems in Engineering 11
CoiLL
CoiFF
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
Figure 21 The constraint graph of a hydraulic grab
h2 g2
f 9984003
g9984003
B2
B1
B0
g3h3
g9984001
f 9984001
B3
x
y
Figure 22 Parts 1198611 1198612 and 119861
3in a hydraulic grab
outlined in Section 331 we find based on the definition ofthe coaxial constraint that the wrench matrix is
W119862119900119894119871119871(119861
21198613)=
[
[
[
[
[
[
0 h2times f3
0 h2times g3
f2 0
g2 0
]
]
]
]
]
]
(54)
The orthogonal complement matrix is
P119862119900119894119871119871(119861
21198613)= [
0 h2
h2 0
] (55)
Then the twist matrix T23
between 1198612and 119861
3has the
following form
T23= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119871119871(11986121198613)) = [
h2 0
0 h2
] (56)
The geometry constraints between 1198611
and 1198613
are119862119900119894119871119871(119861
1 1198613) and 119862119900119894119865119865(119861
1 1198613) Thus the wrench matrices
between 1198611and 119861
3are
W119862119900119894119871119871(119861
11198613)=
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
]
]
]
]
]
]
]
W119862119900119894119865119865(119861
11198613)=
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
h10158403 0
]
]
]
]
(57)
where W13
= W119862119900119894119871119871(119861
11198613)cup W119862119900119894119865119865(119861
11198613) The orthogonal
matrixQ13ofW13is given by
Q13=
[
[
[
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
h10158403 0
]
]
]
]
]
]
]
]
]
]
(58)
Then the orthogonal complement matrix P13ofQ13is
P13= [0 h1015840
3] (59)
Therefore the twist matrix T13between 119861
1and 119861
3is
T13= 119877119890119888119894119901119903119900119888119886119897 (P
13) = [h1015840
3 0] (60)
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system basedon reciprocal screw theory According to the geometryconstraints between any two parts in the hydraulic graband Algorithm 1 we obtain the screw system as shown inFigure 23
As mentioned in [4ndash29] we can perform some differenttypes of kinematic analyses using the screw system such asdetermining the degrees of freedom of 119861
1and 119861
4 that is
T1198611
= T10cap (T20cup T32cup T31)
T1198614
= (T1198611
cup T41) cap (T
1198611
cup T51cup T65cup T64)
(61)
where the resultant of the motions that the end-effector ofa purely serial chain may perform due to the action of eachjoint acting alone is the union of that set and the resultant ofthe set of twists allowed separately by multiple contacts upona body is the intersection of the individual twists Thus thedegrees of freedom of 119861
1are equal to the dimension of the
orthogonal twist space of1198611 denoted byT
1198611
In the samewaywe can obtain the degrees of freedomof119861
4and any part in the
hydraulic grab
12 Mathematical Problems in Engineering
T twist
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
T32
T31
T10
T20
T109
T1110
T114
T98
T74 T8
4
T41
T64
T65
T51
T107
Figure 23 The screw system of a hydraulic grab
R revolute jointC cylindrical joint
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
R31
R10
R20
1110
R114
C98
R74 R8
4
R41
R64 R5
1
R107
R
R109 C3
2
C65
Figure 24 The kinematic joint model of a hydraulic grab
According to Sections 222 and 224 the kinematic jointsbetween 119861
2and 119861
3and 119861
1and 119861
3are the cylindrical joint and
revolute joint As a result we can obtain the kinematic jointmodel See Figure 24 If the corresponding kinematic jointbetween any two parts cannot be recognized we can easilyanalyze the modelrsquos geometry constraints
6 Conclusions
This paper investigates a method for determining the kine-matic joints of assemblies restricted by arbitrary combina-tions of geometry constraints using reciprocal screw theoryWe achieve the process of recognition in terms of thebasic components of geometry constraints and kinematicjoints We describe the three basic components of geometricconstraints in terms of wrench At the same time the specificoperation of a matrix space is applied to eliminate the redun-dant components between different geometry constraintsthat act upon the same part A user of this method can auto-matically and easily convert any assembly model into a screwsystem in any CAD system Afterwards the screw system canbe easily used for validating kinematic functions analyzingover- or underconstrained assembly configurations and soon
Additional Points
(i) We provide a new method for converting any 3Dassembly model into a screw system based on recip-rocal screw theory
(ii) We describe the three basic components of geometricconstraints in terms of wrench
(iii) We eliminate the redundant components of differentgeometry constraints that act upon the same part bythe specific operation of a matrix space
(iv) We establish the maps between geometry constraintsand kinematic joints as the reciprocal product of atwist and wrench which is equal to zero
(v) We present several application examples
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
The authors would like to thank the Ministry of Scienceand Technology of the Peoplersquos Republic of China (nos2011ZX02403-005 2011CB706502 and 2013AA041301) Thiswork is also supported by the National Natural ScienceFoundation of China (nos 61370182 and 51175198)
References
[1] A H Slocum Precision Machine Design Prentice Hall NewYork NY USA 1991
[2] M Rajh S Glodez J Flasker K Gotlih and T KostanjevecldquoDesign and analysis of an fMRI compatible haptic robotrdquoRobotics and Computer-Integrated Manufacturing vol 27 no 2pp 267ndash275 2011
[3] S Awtar and G Parmar ldquoDesign of a large range XY nanoposi-tioning systemrdquo Journal of Mechanisms and Robotics vol 5 no2 Article ID 021008 2013
[4] R S Ball A Treatise on the Theory of Screws CambridgeUniversity Press Cambridge UK 1998
[5] K H Hunt Kinematic Geometry of Mechanisms ClarendonPress Oxford UK 1978
[6] K J Waldron ldquoThe constraint analysis of mechanismsrdquo Journalof Mechanisms vol 1 no 2 pp 101ndash114 1966
[7] R Konkar Incremental kinematic analysis and symbolic syn-thesis of mechanisms [PhD dissertation] Stanford UniversityStanford Calif USA 1993
[8] J-S Zhao Z-J Feng and J-X Dong ldquoComputation of theconfiguration degree of freedomof a spatial parallel mechanismby using reciprocal screw theoryrdquo Mechanism and MachineTheory vol 41 no 12 pp 1486ndash1504 2006
[9] J Eddie Baker ldquoScrew system algebra applied to special linkageconfigurationsrdquoMechanism and Machine Theory vol 15 no 4pp 255ndash265 1980
[10] M TMason and J K Salisbury Robot Hands and theMechanicsof Manipulation MIT Press Cambridge Mass USA 1985
[11] Y Fei and X Zhao ldquoAn assembly process modeling and analysisfor robotic multiple peg-in-holerdquo Journal of Intelligent andRobotic Systems vol 36 no 2 pp 175ndash189 2003
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
Mathematical Problems in Engineering 5
z
x
yo
M
Force system
Instantaneous screw axis (ISA)
Fi
F1 F2
F3
Fr
Mr
ri
Line of action of Fr
Fr
Figure 8 Wrench
is forbidden The first triplet describes the resultant force ina global reference frame The second triplet represents theresultant moment of the force system about the origin of theglobal frame See Figure 8
Both a wrench and geometry constraint are structuralconstraints Therefore we can use a wrench to characterizea component of the geometry constraints between two partsAfterwards the geometry constraints can be transformedinto a corresponding wrench matrix
32 Basic Components of Geometric Constraints When wedescribe a part in space the orientation and position shouldbe specified [34] The orientation cannot be described inabsolute terms An orientation is given by a rotation fromsome known reference orientationThe amount of rotation isknown as an angular displacement In other words describ-ing an orientation is mathematically equivalent to describingan angular displacement In the same way when we specifythe position of a part we cannot do so in absolute terms wemust always do so within the context of a specific referenceframeTherefore specifying a position is actually the same asspecifying an amount of translation from a given referencepoint (usually the origin of some coordinate system)
In practice there are many geometry constraints suchas align and point on line We can use one or arbitrarycombinations of geometric constraints to restrict the relativeorientation and position between two parts As we knowcommon geometry constraints such as coaxial constraintsand spherical constraints usually have a high degree ofconstraint (doc) However these geometry constraints are allcombined by the three different types of basic constraintsdot-1 constraint dot-2 constraint and the distance constraint[35] which can be divided into two categories the orientationpart and the position part
Consider a pair of rigid bodies denoted as bodies 119894 and119895 in Figure 9 Reference points P
119894and P
119895 reference frames
f119894-g119894-h119894and f
119895-g119895-h119895 and nonzero unit vectors a
119894and a
119895
are fixed in bodies 119894 and 119895 respectively Vector d119894119895connects
z
x
y
i
j
ai
Pihi
gi
f i
ri
sPidij
fj
hjgj
sPj
Pj aj
rj
Figure 9 Vectors fixed in and between bodies
P119894and P
119895between bodies The basic constraints are often
characterized by such elements
321 Dot-1 Constraint Dot-1 constraint restricts the relativeorientation of a pair of bodies that is
Φ
1198891
(a119894 a119895 1198621) equiv a119894
Ta119895minus 1198621= 0 119862
1isin (minus1 1) (18)
where doc(Φ1198891) = 1 and dot-1 constraint is symmetric withregard to bodies 119894 and 119895 Note that dot-1 constraint would notbe a basic constraint if119862
1= plusmn1Thismeans that the vectors a
119894
and a119895are parallel in the same or opposite directions In this
case doc(Φ1198891(a119894 a119895 plusmn1)) = 2
Writing the vectors a119894and a119895in terms of their respective
body reference transformationmatricesA119894andA
119895and body-
fixed constant vectors a1015840119894and a1015840
119895 where a
119894= A119894a1015840119894and a
119895=
A119895a1015840119895 (18) has the form
Φ
1198891
(a119894 a119895 1198621) equiv a1015840119894
TA119894
TA119895a1015840119895minus 1198621= 0
1198621isin (minus1 1)
(19)
where the transformation matrices A119894and A
119895in (19) depend
on the orientation generalized coordinates of bodies 119894 and 119895respectively
As a result the wrench representation for dot-1 constraintshould also freeze the relative orientation of the vectors a
119894and
a119895 Therefore the wrench matrix is given by
W1198891 = (0 a119894times a119895) (20)
If we use the body reference transformationmatrices (20)has the form
W1198891 = (0 (a1015840119894
TA119894
T) times (A
119895a1015840119895)) (21)
322 Dot-2 Constraint Dot-2 constraint restricts the relativeposition of a pair of bodies that is
Φ
1198892
(a119894 d119894119895 1198622) equiv a119894
Td119894119895minus 1198622= 0 d
119894119895= 0 (22)
6 Mathematical Problems in Engineering
where doc(Φ1198892) = 1 the same as that with dot-1 constraintalthough dot-2 constraint is not symmetric with regard tobodies 119894 and 119895 Analytically it is important to recall that dot-2constraint breaks down if d
119894119895= 0 Write the vector d
119894119895as
d119894119895= r119895+ A119895s1015840119895
Pminus r119894minus A119894s1015840119894
P (23)
Equation (22) becomes
Φ
1198892
(a119894 d119894119895 1198622) equiv a1015840119894
TA119894
T(r119895+ A119895s1015840119895
Pminus r119894minus A119894s1015840119894
P)
minus 1198622= 0
(24)
As a result the wrench representation for dot-2 constraintmust freeze the relative translation along the vector a
119894
Therefore the wrench matrix is given by
W1198892 = (a119894 0) (25)
Equation (25) also has the form
W1198892 = (A119894a1015840119894 0) (26)
323 Distance Constraint In practice we often require a pairof points on two bodies to coincide or maintain a determineddistance that is
Φ
119904
(119875119894 119875119895 1198623) equiv d119894119895
Td119894119895minus 1198623sdot 1198623= 0 (27)
where doc(Φ119904) would be different depending on whether 1198623
is zero Therefore we have
doc (Φ119904) = 1 if 1198623
= 0
doc (Φ119904) = 3 if 1198623= 0
(28)
As a result the wrench representation for the distanceconstraint should have two types of expressions as shown in(29) and (30)
W119904 = [d119894119895 0] if 119862
3= 0 (29)
where the wrench representation for the distance constraintshould freeze the relative translation along the vector d
119894119895if
1198623
= 0 However if 1198623= 0 the wrench representation has
the form
W119904 = [[[
1 0 0 00 1 0 00 0 1 0
]
]
]
if 1198623= 0 (30)
where the wrench which is illustrated in (30) freezes all therelative translations along the axes 119909 119910 and 119911
33 Wrench Representations of Geometry Constraints Tointroduce geometry constraints we give Figure 10
In Figure 10 reference points P1and P
2 reference frames
f1-g1-h1and f2-g2-h2 and central axes L
1and L
2are fixed in
cylinders 1198611and 119861
2 respectively Vector d
12connects P
1and
P2between the cylinders
L1
B1
h1
P1
f1
F1 g1
L2
B2
h2
P2
f2F2
g2
d12
Figure 10 Instance of geometry constraints
The reference point P1has the form
P1= (1198751199091 1198751199101 1198751199111) (31)
The central axis L1has the form
L1= P1+ 119886 sdot h
1 (32)
where 119886 is a real number The planar surface F1has the form
F1= P1+ 119887 sdot f
1+ 119888 sdot g
1 (33)
where 119887 and 119888 are both real numbers At the same timethe elements P
2 L2 and F
2have the same forms The
following geometry constraints are often characterized bysuch elements
We can divide all different types of geometry constraintswhich have more than one doc into six categories coaxialconstraint coplanar constraint spherical constraint align con-straint distance between line and surface and point on lineWe have modeled the three basic components of geometryconstraints in terms of wrench Thus the wrench models ofthese six types of geometry constraints can be automaticallydeduced
331 Coaxial Constraint Weuse119862119900119894119871119871(L1 L2) to denote the
coaxial constraint of axes L1and L
2 as shown in Figure 11
The doc of 119862119900119894119871119871(L1 L2) is four which restricts two
relative orientations and two relative translations In otherwords it comprises two dot-1 constraints and two dot-2constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(f1 d12 0)
Φ4
1198892
(g1 d12 0)
(34)
Then the wrench matrix of 119862119900119894119871119871(L1 L2) is given by
W119862119900119894119871119871(L
1L2)=
[
[
[
[
[
[
0 h1times f2
0 h1times g2
f1 0
g1 0
]
]
]
]
]
]
(35)
Mathematical Problems in Engineering 7
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
CoiLL(L1 L2)
Figure 11 Coaxial constraint
f2
P2
h2
F2
g2
L2
B2
f1P1
h1
F1
g1
L1
B1
CoiFF(F1 F2)
d12
Figure 12 Coplanar constraint
The corresponding kinematic joint of coaxial constraint iscylindrical joint
332 Coplanar Constraint We use 119862119900119894119865119865(F1 F2) to denote
the coplanar constraint of planar surfaces F1and F
2 as shown
in Figure 12The doc of 119862119900119894119865119865(F
1 F2) is three which restricts two
relative orientations and one relative translation In otherwords it comprises two dot-1 constraints and one dot-2constraint as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(h1 d12 0) (36)
Then the wrench matrix of 119862119900119894119865119865(F1 F2) is given by
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
(37)
The corresponding kinematic joint of coplanar constraintis planar joint
333 Spherical Constraint We use 119862119900119894119875119875(P1P2) to denote
the spherical constraint of two points P1and P
2 as shown in
Figure 13
f2
P2
h2
F2 g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPP(P1P2)
Figure 13 Spherical constraint
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
ParFF(F1 F2)
Figure 14 Align constraint
The doc of 119862119900119894119875119875(P1P2) is three which restricts three
relative translations in three unrelated directions In otherwords it comprises one distance constraint which has 119862
3= 0
in Section 323 as shown in
Φ
119904
(119875119894 119875119895 0) (38)
Then the wrench matrix of 119862119900119894119875119875(P1P2) is given by
W119862119900119894119875119875(P
1P2)=
[
[
[
1 0 0 00 1 0 00 0 1 0
]
]
]
(39)
The corresponding kinematic joint of spherical constraintis spherical joint
334 Align Constraint Align constraint (119875119886119903119865119865) is used tomake two planar surfaces F
1and F
2parallel as shown in
Figure 14The doc of 119875119886119903119865119865(F
1 F2) is two which restricts two
relative orientations In other words it comprises two dot-1constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) (40)
8 Mathematical Problems in Engineering
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
DisLF(L1 F2 D)
Figure 15 Distance between line and surface
Then the wrench matrix of 119875119886119903119865119865(F1 F2) is given by
W119875119886119903119865119865(F
1F2)= [
0 h1times f2
0 h1times g2
] (41)
We cannot find a corresponding kinematic joint for alignconstraint among the common kinematic joints
335 Distance between Line and Surface We use 119863119894119904119871119865(L1
F2 119863) to denote the axis L
1and the surface F
2that maintain
a certain distance119863 as shown in Figure 15The doc of 119863119894119904119871119865(L
1 F2 119863) is two which restricts one
relative orientation and one relative translation In otherwords it comprises one dot-1 constraint and one dot-2constraint as shown in
Φ1
1198891
(h1 h2 0) Φ
2
1198892
(h2 d12 119863) (42)
Then the wrench matrix of119863119894119904119871119865(L1 F2 119863) is given by
W119863119894119904119871119865(L
1F2119863)
= [
0 h1times h2
h2 0
] (43)
We also cannot find a corresponding kinematic jointfor distance between line and surface among the commonkinematic joints
336 Point on Line Weuse119862119900119894119875119871(P1 L2) to denote the point
P1on line L
2 as shown in Figure 16
The doc of 119862119900119894119875119871(P1 L2) is two which restricts two
relative translations In other words it comprises two dot-2constraints as shown in
Φ1
1198892
(f2 d12 0) Φ
2
1198892
(g2 d12 0) (44)
Then the wrench matrix of 119862119900119894119875119871(P1 L2) is given by
W119863119894119904119871119865(L
1F2119863)
= [
f2 0
g2 0
] (45)
We still cannot find a corresponding kinematic joint forpoint on line among the common kinematic joints
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPL(P1 L2)
Figure 16 Point on line
4 Recognition of Kinematic Joints
The key to recognizing kinematic joints in terms of geometryconstraints is the reciprocity relation We show how to definegeometry constraints using wrench The space of screws thatdefines the range of wrenches that can be exerted upon abody and the space of screws that defines the range of twiststhat the body can perform without breaking any contacts inthe joints of the mechanism are mutually reciprocal screwspaces When we know the wrenches acting upon a body wecan determine the motions it can perform Conversely if weknow the motions a body is capable of performing then theconstraints that are exerted upon it can also be deduced
41 Algorithm We assume that 119860 is a part in a 3D assemblymodel restricted by parts 119861
1 1198612 119861
119899 The twist space of 119860
is the space of screws reciprocal to the wrenches of contactacting upon it The space of wrenches acting upon 119860 is theunion of all the individual wrenches acting upon it fromthe parts 119861
119894 These individual wrenches are to be considered
as acting alone Algorithm 1 provides the twist space of 119860denoted by T119860
The 119877119890119888119894119901119903119900119888119886119897(P) in Algorithm 1 means to exchangethe first triplet and the second triplet of P For exampleif P = [119886 119887 119888 119889 119890 119891] then we have 119877119890119888119894119901119903119900119888119886119897(P) =
[119889 119890 119891 119886 119887 119888] The recursive nature of the procedurerequires the computation of the wrench space and twistspace of each component This proves to be computationallyexpensive because wrench space and twist space are definedin terms of the reciprocal of one another The operation ofderiving reciprocal screws involves computing the null spaceof matrices The speed of the algorithm is linearly related tothe number of 119861
119894 That is the algorithm is of the order 119874(119899)
42 Reciprocal Product According to reciprocal screw the-ory two screws (a twist and a wrench) are said to be mutuallyreciprocal if their reciprocal product is zero The reciprocal
Mathematical Problems in Engineering 9
BEGINFOR each part Bi which restrict AConvert all the geometry constraints between A and Bi into basic components 119862119895
119860 119895 = 1 119897
Determine the wrenchesW119896119860 119896= 1 119898 acting upon A through C119895
119860 We havem not less than l
Collect all the wrenchesW119896119860into a matrixW
END FORCompute the orthogonal matrixQ in terms ofW by Gram-Schmidt orthogonalization [31]IF Rank(Q) = 6 THENReturn ldquoIMMOBILErdquo
ELSECompute the orthogonal complement matrix P of Q
END IFTA =Reciprocal(P)
END
Algorithm 1 Function twist space
f2
A2
h2F2
g2
B2
f1A1
F1
g1B1
Figure 17 Redundant constraint
product of a twist T1and a wrenchW
1thatisequal to zero is
given by
T1∘W1= T11minus3
sdotW24minus6
+ T14minus6
sdotW21minus3
= [0 0] (46)
Mathematically because reciprocity imposes the relationT1∘W1 we can compute one of these quantities if the other is
known This allows us to compute the twists that a body mayexecute while in contact with other bodies without breakingthe contacts
43 Redundant Constraint When more than one geometryconstraint acts upon the same part because some compo-nents of these constraints may be redundant the resultantmotion is the logical intersection of the individual wrenchmatrix that defines each constraint for example when wemodel a revolute joint 119877
12with two links 119861
1and 119861
2 The
constraints used are 119862119900119894119871119871(A1A2) and 119862119900119894119865119865(F
1 F2) See
Figure 17Also dof(119877
12) = 1Thismeans that doc(119861
1 1198612) = 5 How-
ever we have doc(119862119900119894119871119871(A1A2)) = 4 and doc(119862119900119894119865119865(F
1
F2)) = 3 In other words the dimension of the orthog-
onal wrench space which comprises W119862119900119894119871119871(A
1A2)and
f2
h2
F2g2
B2
f1h1
g1
B1
F1
Figure 18 Recognition of planar joint
W119862119900119894119865119865(F
1F2) is 5 Therefore the motions that a body is
capable of performing while under the action of multiplewrenches are the logical intersection of the motions it canperform under the action of each wrench acting alone Thuswe can eliminate the redundant components of differentgeometry constraints that act upon the same part by usingGram-Schmidt orthogonalization in Algorithm 1
5 Examples
In each of the examples below different types of kinematicjoints are recognized
Example 1 (planar joint) Figure 18 shows a planar jointconsisting of two parts 119861
1and 119861
2 and a coplanar constraint
119862119900119894119865119865(F1 F2)
Following the process outlined in Section 332 we findbased on the definition of the coplanar constraint that thewrench matrix is
W119862119900119894119865119865(F
1F2)=
[
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
]
(47)
10 Mathematical Problems in Engineering
f2h2
F2
g2
B2
f1
h1 g1
B1
F1
f 9984002
h9984002g9984002
F9984002
f 9984001
h9984001
g9984001
F9984001
Figure 19 Recognition of prismatic joint
whereW119862119900119894119865119865(F
1F2)is already an orthogonalmatrixThen the
orthogonal complement matrix P119862119900119894119865119865(F
1F2)is
P119862119900119894119865119865(F
1F2)=
[
[
[
f1 0
g1 0
0 h1
]
]
]
(48)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119865119865(F1F2)) =
[
[
[
0 f1
0 g1
h1 0
]
]
]
(49)
where T12
represents the motion of a planar joint which isillustrated in Section 226 Moreover the recognition processof a cylindrical joint or spherical joint is approximately thesame as that of the planar joint
Example 2 (prismatic joint) Figure 19 shows a prismaticjoint consisting of two parts 119861
1and 119861
2 and two coplanar
constraints 119862119900119894119865119865(F1 F2) and 119862119900119894119865119865(F1015840
1 F10158402)
Following the process outlined in Example 1 we find thatthe wrench matrices between 119861
1and 119861
2are
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
W119862119900119894119865119865(F1015840
1F10158402)=
[
[
[
[
0 h10158401times f10158402
0 h10158401times g10158402
h10158401 0
]
]
]
]
(50)
B0
B1
B2
B3
B4
B5B6
B7
B8
B9
B10
B11
x
y
Figure 20 Hydraulic grab
where W12
= W119862119900119894119865119865(F
1F2)cup W119862119900119894119865119865(F1015840
1F10158402) The orthogonal
matrixQ12ofW12is
Q12=
[
[
[
[
[
[
[
[
[
[
[
0 h1times f2
0 h1times g2
0 h10158401times g10158402
h1 0
h10158401 0
]
]
]
]
]
]
]
]
]
]
]
(51)
The orthogonal complement matrix P12ofQ12is
P12= [g1 0] (52)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
12) = [0 g
1] (53)
where T12
represents the motion of a prismatic joint whichis shown in Section 221 Moreover the recognition processof a revolute joint is approximately the same as that of theprismatic joint
Example 3 (hydraulic grab) Here we present a projectexample A hydraulic grab with three degrees of freedom inCartesian coordinates is illustrated in Figure 20
The hydraulic grab comprises 12 parts 1198610 1198611 1198612 1198613
1198614 1198615 1198616 1198617 1198618 1198619 11986110 and 119861
11 12 coplanar constraints
119862119900119894119865119865(1198610 1198612)119862119900119894119865119865(119861
0 1198611)119862119900119894119865119865(119861
1 1198613)119862119900119894119865119865(119861
1 1198615)
119862119900119894119865119865(1198611 1198614) 119862119900119894119865119865(119861
4 1198616) 119862119900119894119865119865(119861
4 11986111) 119862119900119894119865119865(119861
4
1198617) 119862119900119894119865119865(119861
4 1198618) 119862119900119894119865119865(119861
9 11986110) 119862119900119894119865119865(119861
7 11986110) and
119862119900119894119865119865(11986110 11986111) and 15 coaxial constraints 119862119900119894119871119871(119861
0 1198612)
119862119900119894119871119871(1198610 1198611) 119862119900119894119871119871(119861
1 1198613) 119862119900119894119871119871(119861
2 1198613) 119862119900119894119871119871(119861
1 1198615)
119862119900119894119871119871(1198611 1198614) 119862119900119894119871119871(119861
4 1198616) 119862119900119894119871119871(119861
5 1198616) 119862119900119894119871119871(119861
4
11986111) 119862119900119894119871119871(119861
4 1198617) 119862119900119894119871119871(119861
4 1198618) 119862119900119894119871119871(119861
8 1198619)
119862119900119894119871119871(1198619 11986110) 119862119900119894119871119871(119861
7 11986110) and 119862119900119894119871119871(119861
10 11986111) See
Figure 21Here we present the processes of recognition among 119861
1
1198612 and 119861
3 as shown in Figure 22 The geometry constraint
between 1198612and 119861
3is 119862119900119894119871119871(119861
2 1198613) Following the process
Mathematical Problems in Engineering 11
CoiLL
CoiFF
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
Figure 21 The constraint graph of a hydraulic grab
h2 g2
f 9984003
g9984003
B2
B1
B0
g3h3
g9984001
f 9984001
B3
x
y
Figure 22 Parts 1198611 1198612 and 119861
3in a hydraulic grab
outlined in Section 331 we find based on the definition ofthe coaxial constraint that the wrench matrix is
W119862119900119894119871119871(119861
21198613)=
[
[
[
[
[
[
0 h2times f3
0 h2times g3
f2 0
g2 0
]
]
]
]
]
]
(54)
The orthogonal complement matrix is
P119862119900119894119871119871(119861
21198613)= [
0 h2
h2 0
] (55)
Then the twist matrix T23
between 1198612and 119861
3has the
following form
T23= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119871119871(11986121198613)) = [
h2 0
0 h2
] (56)
The geometry constraints between 1198611
and 1198613
are119862119900119894119871119871(119861
1 1198613) and 119862119900119894119865119865(119861
1 1198613) Thus the wrench matrices
between 1198611and 119861
3are
W119862119900119894119871119871(119861
11198613)=
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
]
]
]
]
]
]
]
W119862119900119894119865119865(119861
11198613)=
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
h10158403 0
]
]
]
]
(57)
where W13
= W119862119900119894119871119871(119861
11198613)cup W119862119900119894119865119865(119861
11198613) The orthogonal
matrixQ13ofW13is given by
Q13=
[
[
[
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
h10158403 0
]
]
]
]
]
]
]
]
]
]
(58)
Then the orthogonal complement matrix P13ofQ13is
P13= [0 h1015840
3] (59)
Therefore the twist matrix T13between 119861
1and 119861
3is
T13= 119877119890119888119894119901119903119900119888119886119897 (P
13) = [h1015840
3 0] (60)
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system basedon reciprocal screw theory According to the geometryconstraints between any two parts in the hydraulic graband Algorithm 1 we obtain the screw system as shown inFigure 23
As mentioned in [4ndash29] we can perform some differenttypes of kinematic analyses using the screw system such asdetermining the degrees of freedom of 119861
1and 119861
4 that is
T1198611
= T10cap (T20cup T32cup T31)
T1198614
= (T1198611
cup T41) cap (T
1198611
cup T51cup T65cup T64)
(61)
where the resultant of the motions that the end-effector ofa purely serial chain may perform due to the action of eachjoint acting alone is the union of that set and the resultant ofthe set of twists allowed separately by multiple contacts upona body is the intersection of the individual twists Thus thedegrees of freedom of 119861
1are equal to the dimension of the
orthogonal twist space of1198611 denoted byT
1198611
In the samewaywe can obtain the degrees of freedomof119861
4and any part in the
hydraulic grab
12 Mathematical Problems in Engineering
T twist
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
T32
T31
T10
T20
T109
T1110
T114
T98
T74 T8
4
T41
T64
T65
T51
T107
Figure 23 The screw system of a hydraulic grab
R revolute jointC cylindrical joint
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
R31
R10
R20
1110
R114
C98
R74 R8
4
R41
R64 R5
1
R107
R
R109 C3
2
C65
Figure 24 The kinematic joint model of a hydraulic grab
According to Sections 222 and 224 the kinematic jointsbetween 119861
2and 119861
3and 119861
1and 119861
3are the cylindrical joint and
revolute joint As a result we can obtain the kinematic jointmodel See Figure 24 If the corresponding kinematic jointbetween any two parts cannot be recognized we can easilyanalyze the modelrsquos geometry constraints
6 Conclusions
This paper investigates a method for determining the kine-matic joints of assemblies restricted by arbitrary combina-tions of geometry constraints using reciprocal screw theoryWe achieve the process of recognition in terms of thebasic components of geometry constraints and kinematicjoints We describe the three basic components of geometricconstraints in terms of wrench At the same time the specificoperation of a matrix space is applied to eliminate the redun-dant components between different geometry constraintsthat act upon the same part A user of this method can auto-matically and easily convert any assembly model into a screwsystem in any CAD system Afterwards the screw system canbe easily used for validating kinematic functions analyzingover- or underconstrained assembly configurations and soon
Additional Points
(i) We provide a new method for converting any 3Dassembly model into a screw system based on recip-rocal screw theory
(ii) We describe the three basic components of geometricconstraints in terms of wrench
(iii) We eliminate the redundant components of differentgeometry constraints that act upon the same part bythe specific operation of a matrix space
(iv) We establish the maps between geometry constraintsand kinematic joints as the reciprocal product of atwist and wrench which is equal to zero
(v) We present several application examples
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
The authors would like to thank the Ministry of Scienceand Technology of the Peoplersquos Republic of China (nos2011ZX02403-005 2011CB706502 and 2013AA041301) Thiswork is also supported by the National Natural ScienceFoundation of China (nos 61370182 and 51175198)
References
[1] A H Slocum Precision Machine Design Prentice Hall NewYork NY USA 1991
[2] M Rajh S Glodez J Flasker K Gotlih and T KostanjevecldquoDesign and analysis of an fMRI compatible haptic robotrdquoRobotics and Computer-Integrated Manufacturing vol 27 no 2pp 267ndash275 2011
[3] S Awtar and G Parmar ldquoDesign of a large range XY nanoposi-tioning systemrdquo Journal of Mechanisms and Robotics vol 5 no2 Article ID 021008 2013
[4] R S Ball A Treatise on the Theory of Screws CambridgeUniversity Press Cambridge UK 1998
[5] K H Hunt Kinematic Geometry of Mechanisms ClarendonPress Oxford UK 1978
[6] K J Waldron ldquoThe constraint analysis of mechanismsrdquo Journalof Mechanisms vol 1 no 2 pp 101ndash114 1966
[7] R Konkar Incremental kinematic analysis and symbolic syn-thesis of mechanisms [PhD dissertation] Stanford UniversityStanford Calif USA 1993
[8] J-S Zhao Z-J Feng and J-X Dong ldquoComputation of theconfiguration degree of freedomof a spatial parallel mechanismby using reciprocal screw theoryrdquo Mechanism and MachineTheory vol 41 no 12 pp 1486ndash1504 2006
[9] J Eddie Baker ldquoScrew system algebra applied to special linkageconfigurationsrdquoMechanism and Machine Theory vol 15 no 4pp 255ndash265 1980
[10] M TMason and J K Salisbury Robot Hands and theMechanicsof Manipulation MIT Press Cambridge Mass USA 1985
[11] Y Fei and X Zhao ldquoAn assembly process modeling and analysisfor robotic multiple peg-in-holerdquo Journal of Intelligent andRobotic Systems vol 36 no 2 pp 175ndash189 2003
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
6 Mathematical Problems in Engineering
where doc(Φ1198892) = 1 the same as that with dot-1 constraintalthough dot-2 constraint is not symmetric with regard tobodies 119894 and 119895 Analytically it is important to recall that dot-2constraint breaks down if d
119894119895= 0 Write the vector d
119894119895as
d119894119895= r119895+ A119895s1015840119895
Pminus r119894minus A119894s1015840119894
P (23)
Equation (22) becomes
Φ
1198892
(a119894 d119894119895 1198622) equiv a1015840119894
TA119894
T(r119895+ A119895s1015840119895
Pminus r119894minus A119894s1015840119894
P)
minus 1198622= 0
(24)
As a result the wrench representation for dot-2 constraintmust freeze the relative translation along the vector a
119894
Therefore the wrench matrix is given by
W1198892 = (a119894 0) (25)
Equation (25) also has the form
W1198892 = (A119894a1015840119894 0) (26)
323 Distance Constraint In practice we often require a pairof points on two bodies to coincide or maintain a determineddistance that is
Φ
119904
(119875119894 119875119895 1198623) equiv d119894119895
Td119894119895minus 1198623sdot 1198623= 0 (27)
where doc(Φ119904) would be different depending on whether 1198623
is zero Therefore we have
doc (Φ119904) = 1 if 1198623
= 0
doc (Φ119904) = 3 if 1198623= 0
(28)
As a result the wrench representation for the distanceconstraint should have two types of expressions as shown in(29) and (30)
W119904 = [d119894119895 0] if 119862
3= 0 (29)
where the wrench representation for the distance constraintshould freeze the relative translation along the vector d
119894119895if
1198623
= 0 However if 1198623= 0 the wrench representation has
the form
W119904 = [[[
1 0 0 00 1 0 00 0 1 0
]
]
]
if 1198623= 0 (30)
where the wrench which is illustrated in (30) freezes all therelative translations along the axes 119909 119910 and 119911
33 Wrench Representations of Geometry Constraints Tointroduce geometry constraints we give Figure 10
In Figure 10 reference points P1and P
2 reference frames
f1-g1-h1and f2-g2-h2 and central axes L
1and L
2are fixed in
cylinders 1198611and 119861
2 respectively Vector d
12connects P
1and
P2between the cylinders
L1
B1
h1
P1
f1
F1 g1
L2
B2
h2
P2
f2F2
g2
d12
Figure 10 Instance of geometry constraints
The reference point P1has the form
P1= (1198751199091 1198751199101 1198751199111) (31)
The central axis L1has the form
L1= P1+ 119886 sdot h
1 (32)
where 119886 is a real number The planar surface F1has the form
F1= P1+ 119887 sdot f
1+ 119888 sdot g
1 (33)
where 119887 and 119888 are both real numbers At the same timethe elements P
2 L2 and F
2have the same forms The
following geometry constraints are often characterized bysuch elements
We can divide all different types of geometry constraintswhich have more than one doc into six categories coaxialconstraint coplanar constraint spherical constraint align con-straint distance between line and surface and point on lineWe have modeled the three basic components of geometryconstraints in terms of wrench Thus the wrench models ofthese six types of geometry constraints can be automaticallydeduced
331 Coaxial Constraint Weuse119862119900119894119871119871(L1 L2) to denote the
coaxial constraint of axes L1and L
2 as shown in Figure 11
The doc of 119862119900119894119871119871(L1 L2) is four which restricts two
relative orientations and two relative translations In otherwords it comprises two dot-1 constraints and two dot-2constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(f1 d12 0)
Φ4
1198892
(g1 d12 0)
(34)
Then the wrench matrix of 119862119900119894119871119871(L1 L2) is given by
W119862119900119894119871119871(L
1L2)=
[
[
[
[
[
[
0 h1times f2
0 h1times g2
f1 0
g1 0
]
]
]
]
]
]
(35)
Mathematical Problems in Engineering 7
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
CoiLL(L1 L2)
Figure 11 Coaxial constraint
f2
P2
h2
F2
g2
L2
B2
f1P1
h1
F1
g1
L1
B1
CoiFF(F1 F2)
d12
Figure 12 Coplanar constraint
The corresponding kinematic joint of coaxial constraint iscylindrical joint
332 Coplanar Constraint We use 119862119900119894119865119865(F1 F2) to denote
the coplanar constraint of planar surfaces F1and F
2 as shown
in Figure 12The doc of 119862119900119894119865119865(F
1 F2) is three which restricts two
relative orientations and one relative translation In otherwords it comprises two dot-1 constraints and one dot-2constraint as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(h1 d12 0) (36)
Then the wrench matrix of 119862119900119894119865119865(F1 F2) is given by
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
(37)
The corresponding kinematic joint of coplanar constraintis planar joint
333 Spherical Constraint We use 119862119900119894119875119875(P1P2) to denote
the spherical constraint of two points P1and P
2 as shown in
Figure 13
f2
P2
h2
F2 g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPP(P1P2)
Figure 13 Spherical constraint
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
ParFF(F1 F2)
Figure 14 Align constraint
The doc of 119862119900119894119875119875(P1P2) is three which restricts three
relative translations in three unrelated directions In otherwords it comprises one distance constraint which has 119862
3= 0
in Section 323 as shown in
Φ
119904
(119875119894 119875119895 0) (38)
Then the wrench matrix of 119862119900119894119875119875(P1P2) is given by
W119862119900119894119875119875(P
1P2)=
[
[
[
1 0 0 00 1 0 00 0 1 0
]
]
]
(39)
The corresponding kinematic joint of spherical constraintis spherical joint
334 Align Constraint Align constraint (119875119886119903119865119865) is used tomake two planar surfaces F
1and F
2parallel as shown in
Figure 14The doc of 119875119886119903119865119865(F
1 F2) is two which restricts two
relative orientations In other words it comprises two dot-1constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) (40)
8 Mathematical Problems in Engineering
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
DisLF(L1 F2 D)
Figure 15 Distance between line and surface
Then the wrench matrix of 119875119886119903119865119865(F1 F2) is given by
W119875119886119903119865119865(F
1F2)= [
0 h1times f2
0 h1times g2
] (41)
We cannot find a corresponding kinematic joint for alignconstraint among the common kinematic joints
335 Distance between Line and Surface We use 119863119894119904119871119865(L1
F2 119863) to denote the axis L
1and the surface F
2that maintain
a certain distance119863 as shown in Figure 15The doc of 119863119894119904119871119865(L
1 F2 119863) is two which restricts one
relative orientation and one relative translation In otherwords it comprises one dot-1 constraint and one dot-2constraint as shown in
Φ1
1198891
(h1 h2 0) Φ
2
1198892
(h2 d12 119863) (42)
Then the wrench matrix of119863119894119904119871119865(L1 F2 119863) is given by
W119863119894119904119871119865(L
1F2119863)
= [
0 h1times h2
h2 0
] (43)
We also cannot find a corresponding kinematic jointfor distance between line and surface among the commonkinematic joints
336 Point on Line Weuse119862119900119894119875119871(P1 L2) to denote the point
P1on line L
2 as shown in Figure 16
The doc of 119862119900119894119875119871(P1 L2) is two which restricts two
relative translations In other words it comprises two dot-2constraints as shown in
Φ1
1198892
(f2 d12 0) Φ
2
1198892
(g2 d12 0) (44)
Then the wrench matrix of 119862119900119894119875119871(P1 L2) is given by
W119863119894119904119871119865(L
1F2119863)
= [
f2 0
g2 0
] (45)
We still cannot find a corresponding kinematic joint forpoint on line among the common kinematic joints
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPL(P1 L2)
Figure 16 Point on line
4 Recognition of Kinematic Joints
The key to recognizing kinematic joints in terms of geometryconstraints is the reciprocity relation We show how to definegeometry constraints using wrench The space of screws thatdefines the range of wrenches that can be exerted upon abody and the space of screws that defines the range of twiststhat the body can perform without breaking any contacts inthe joints of the mechanism are mutually reciprocal screwspaces When we know the wrenches acting upon a body wecan determine the motions it can perform Conversely if weknow the motions a body is capable of performing then theconstraints that are exerted upon it can also be deduced
41 Algorithm We assume that 119860 is a part in a 3D assemblymodel restricted by parts 119861
1 1198612 119861
119899 The twist space of 119860
is the space of screws reciprocal to the wrenches of contactacting upon it The space of wrenches acting upon 119860 is theunion of all the individual wrenches acting upon it fromthe parts 119861
119894 These individual wrenches are to be considered
as acting alone Algorithm 1 provides the twist space of 119860denoted by T119860
The 119877119890119888119894119901119903119900119888119886119897(P) in Algorithm 1 means to exchangethe first triplet and the second triplet of P For exampleif P = [119886 119887 119888 119889 119890 119891] then we have 119877119890119888119894119901119903119900119888119886119897(P) =
[119889 119890 119891 119886 119887 119888] The recursive nature of the procedurerequires the computation of the wrench space and twistspace of each component This proves to be computationallyexpensive because wrench space and twist space are definedin terms of the reciprocal of one another The operation ofderiving reciprocal screws involves computing the null spaceof matrices The speed of the algorithm is linearly related tothe number of 119861
119894 That is the algorithm is of the order 119874(119899)
42 Reciprocal Product According to reciprocal screw the-ory two screws (a twist and a wrench) are said to be mutuallyreciprocal if their reciprocal product is zero The reciprocal
Mathematical Problems in Engineering 9
BEGINFOR each part Bi which restrict AConvert all the geometry constraints between A and Bi into basic components 119862119895
119860 119895 = 1 119897
Determine the wrenchesW119896119860 119896= 1 119898 acting upon A through C119895
119860 We havem not less than l
Collect all the wrenchesW119896119860into a matrixW
END FORCompute the orthogonal matrixQ in terms ofW by Gram-Schmidt orthogonalization [31]IF Rank(Q) = 6 THENReturn ldquoIMMOBILErdquo
ELSECompute the orthogonal complement matrix P of Q
END IFTA =Reciprocal(P)
END
Algorithm 1 Function twist space
f2
A2
h2F2
g2
B2
f1A1
F1
g1B1
Figure 17 Redundant constraint
product of a twist T1and a wrenchW
1thatisequal to zero is
given by
T1∘W1= T11minus3
sdotW24minus6
+ T14minus6
sdotW21minus3
= [0 0] (46)
Mathematically because reciprocity imposes the relationT1∘W1 we can compute one of these quantities if the other is
known This allows us to compute the twists that a body mayexecute while in contact with other bodies without breakingthe contacts
43 Redundant Constraint When more than one geometryconstraint acts upon the same part because some compo-nents of these constraints may be redundant the resultantmotion is the logical intersection of the individual wrenchmatrix that defines each constraint for example when wemodel a revolute joint 119877
12with two links 119861
1and 119861
2 The
constraints used are 119862119900119894119871119871(A1A2) and 119862119900119894119865119865(F
1 F2) See
Figure 17Also dof(119877
12) = 1Thismeans that doc(119861
1 1198612) = 5 How-
ever we have doc(119862119900119894119871119871(A1A2)) = 4 and doc(119862119900119894119865119865(F
1
F2)) = 3 In other words the dimension of the orthog-
onal wrench space which comprises W119862119900119894119871119871(A
1A2)and
f2
h2
F2g2
B2
f1h1
g1
B1
F1
Figure 18 Recognition of planar joint
W119862119900119894119865119865(F
1F2) is 5 Therefore the motions that a body is
capable of performing while under the action of multiplewrenches are the logical intersection of the motions it canperform under the action of each wrench acting alone Thuswe can eliminate the redundant components of differentgeometry constraints that act upon the same part by usingGram-Schmidt orthogonalization in Algorithm 1
5 Examples
In each of the examples below different types of kinematicjoints are recognized
Example 1 (planar joint) Figure 18 shows a planar jointconsisting of two parts 119861
1and 119861
2 and a coplanar constraint
119862119900119894119865119865(F1 F2)
Following the process outlined in Section 332 we findbased on the definition of the coplanar constraint that thewrench matrix is
W119862119900119894119865119865(F
1F2)=
[
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
]
(47)
10 Mathematical Problems in Engineering
f2h2
F2
g2
B2
f1
h1 g1
B1
F1
f 9984002
h9984002g9984002
F9984002
f 9984001
h9984001
g9984001
F9984001
Figure 19 Recognition of prismatic joint
whereW119862119900119894119865119865(F
1F2)is already an orthogonalmatrixThen the
orthogonal complement matrix P119862119900119894119865119865(F
1F2)is
P119862119900119894119865119865(F
1F2)=
[
[
[
f1 0
g1 0
0 h1
]
]
]
(48)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119865119865(F1F2)) =
[
[
[
0 f1
0 g1
h1 0
]
]
]
(49)
where T12
represents the motion of a planar joint which isillustrated in Section 226 Moreover the recognition processof a cylindrical joint or spherical joint is approximately thesame as that of the planar joint
Example 2 (prismatic joint) Figure 19 shows a prismaticjoint consisting of two parts 119861
1and 119861
2 and two coplanar
constraints 119862119900119894119865119865(F1 F2) and 119862119900119894119865119865(F1015840
1 F10158402)
Following the process outlined in Example 1 we find thatthe wrench matrices between 119861
1and 119861
2are
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
W119862119900119894119865119865(F1015840
1F10158402)=
[
[
[
[
0 h10158401times f10158402
0 h10158401times g10158402
h10158401 0
]
]
]
]
(50)
B0
B1
B2
B3
B4
B5B6
B7
B8
B9
B10
B11
x
y
Figure 20 Hydraulic grab
where W12
= W119862119900119894119865119865(F
1F2)cup W119862119900119894119865119865(F1015840
1F10158402) The orthogonal
matrixQ12ofW12is
Q12=
[
[
[
[
[
[
[
[
[
[
[
0 h1times f2
0 h1times g2
0 h10158401times g10158402
h1 0
h10158401 0
]
]
]
]
]
]
]
]
]
]
]
(51)
The orthogonal complement matrix P12ofQ12is
P12= [g1 0] (52)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
12) = [0 g
1] (53)
where T12
represents the motion of a prismatic joint whichis shown in Section 221 Moreover the recognition processof a revolute joint is approximately the same as that of theprismatic joint
Example 3 (hydraulic grab) Here we present a projectexample A hydraulic grab with three degrees of freedom inCartesian coordinates is illustrated in Figure 20
The hydraulic grab comprises 12 parts 1198610 1198611 1198612 1198613
1198614 1198615 1198616 1198617 1198618 1198619 11986110 and 119861
11 12 coplanar constraints
119862119900119894119865119865(1198610 1198612)119862119900119894119865119865(119861
0 1198611)119862119900119894119865119865(119861
1 1198613)119862119900119894119865119865(119861
1 1198615)
119862119900119894119865119865(1198611 1198614) 119862119900119894119865119865(119861
4 1198616) 119862119900119894119865119865(119861
4 11986111) 119862119900119894119865119865(119861
4
1198617) 119862119900119894119865119865(119861
4 1198618) 119862119900119894119865119865(119861
9 11986110) 119862119900119894119865119865(119861
7 11986110) and
119862119900119894119865119865(11986110 11986111) and 15 coaxial constraints 119862119900119894119871119871(119861
0 1198612)
119862119900119894119871119871(1198610 1198611) 119862119900119894119871119871(119861
1 1198613) 119862119900119894119871119871(119861
2 1198613) 119862119900119894119871119871(119861
1 1198615)
119862119900119894119871119871(1198611 1198614) 119862119900119894119871119871(119861
4 1198616) 119862119900119894119871119871(119861
5 1198616) 119862119900119894119871119871(119861
4
11986111) 119862119900119894119871119871(119861
4 1198617) 119862119900119894119871119871(119861
4 1198618) 119862119900119894119871119871(119861
8 1198619)
119862119900119894119871119871(1198619 11986110) 119862119900119894119871119871(119861
7 11986110) and 119862119900119894119871119871(119861
10 11986111) See
Figure 21Here we present the processes of recognition among 119861
1
1198612 and 119861
3 as shown in Figure 22 The geometry constraint
between 1198612and 119861
3is 119862119900119894119871119871(119861
2 1198613) Following the process
Mathematical Problems in Engineering 11
CoiLL
CoiFF
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
Figure 21 The constraint graph of a hydraulic grab
h2 g2
f 9984003
g9984003
B2
B1
B0
g3h3
g9984001
f 9984001
B3
x
y
Figure 22 Parts 1198611 1198612 and 119861
3in a hydraulic grab
outlined in Section 331 we find based on the definition ofthe coaxial constraint that the wrench matrix is
W119862119900119894119871119871(119861
21198613)=
[
[
[
[
[
[
0 h2times f3
0 h2times g3
f2 0
g2 0
]
]
]
]
]
]
(54)
The orthogonal complement matrix is
P119862119900119894119871119871(119861
21198613)= [
0 h2
h2 0
] (55)
Then the twist matrix T23
between 1198612and 119861
3has the
following form
T23= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119871119871(11986121198613)) = [
h2 0
0 h2
] (56)
The geometry constraints between 1198611
and 1198613
are119862119900119894119871119871(119861
1 1198613) and 119862119900119894119865119865(119861
1 1198613) Thus the wrench matrices
between 1198611and 119861
3are
W119862119900119894119871119871(119861
11198613)=
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
]
]
]
]
]
]
]
W119862119900119894119865119865(119861
11198613)=
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
h10158403 0
]
]
]
]
(57)
where W13
= W119862119900119894119871119871(119861
11198613)cup W119862119900119894119865119865(119861
11198613) The orthogonal
matrixQ13ofW13is given by
Q13=
[
[
[
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
h10158403 0
]
]
]
]
]
]
]
]
]
]
(58)
Then the orthogonal complement matrix P13ofQ13is
P13= [0 h1015840
3] (59)
Therefore the twist matrix T13between 119861
1and 119861
3is
T13= 119877119890119888119894119901119903119900119888119886119897 (P
13) = [h1015840
3 0] (60)
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system basedon reciprocal screw theory According to the geometryconstraints between any two parts in the hydraulic graband Algorithm 1 we obtain the screw system as shown inFigure 23
As mentioned in [4ndash29] we can perform some differenttypes of kinematic analyses using the screw system such asdetermining the degrees of freedom of 119861
1and 119861
4 that is
T1198611
= T10cap (T20cup T32cup T31)
T1198614
= (T1198611
cup T41) cap (T
1198611
cup T51cup T65cup T64)
(61)
where the resultant of the motions that the end-effector ofa purely serial chain may perform due to the action of eachjoint acting alone is the union of that set and the resultant ofthe set of twists allowed separately by multiple contacts upona body is the intersection of the individual twists Thus thedegrees of freedom of 119861
1are equal to the dimension of the
orthogonal twist space of1198611 denoted byT
1198611
In the samewaywe can obtain the degrees of freedomof119861
4and any part in the
hydraulic grab
12 Mathematical Problems in Engineering
T twist
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
T32
T31
T10
T20
T109
T1110
T114
T98
T74 T8
4
T41
T64
T65
T51
T107
Figure 23 The screw system of a hydraulic grab
R revolute jointC cylindrical joint
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
R31
R10
R20
1110
R114
C98
R74 R8
4
R41
R64 R5
1
R107
R
R109 C3
2
C65
Figure 24 The kinematic joint model of a hydraulic grab
According to Sections 222 and 224 the kinematic jointsbetween 119861
2and 119861
3and 119861
1and 119861
3are the cylindrical joint and
revolute joint As a result we can obtain the kinematic jointmodel See Figure 24 If the corresponding kinematic jointbetween any two parts cannot be recognized we can easilyanalyze the modelrsquos geometry constraints
6 Conclusions
This paper investigates a method for determining the kine-matic joints of assemblies restricted by arbitrary combina-tions of geometry constraints using reciprocal screw theoryWe achieve the process of recognition in terms of thebasic components of geometry constraints and kinematicjoints We describe the three basic components of geometricconstraints in terms of wrench At the same time the specificoperation of a matrix space is applied to eliminate the redun-dant components between different geometry constraintsthat act upon the same part A user of this method can auto-matically and easily convert any assembly model into a screwsystem in any CAD system Afterwards the screw system canbe easily used for validating kinematic functions analyzingover- or underconstrained assembly configurations and soon
Additional Points
(i) We provide a new method for converting any 3Dassembly model into a screw system based on recip-rocal screw theory
(ii) We describe the three basic components of geometricconstraints in terms of wrench
(iii) We eliminate the redundant components of differentgeometry constraints that act upon the same part bythe specific operation of a matrix space
(iv) We establish the maps between geometry constraintsand kinematic joints as the reciprocal product of atwist and wrench which is equal to zero
(v) We present several application examples
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
The authors would like to thank the Ministry of Scienceand Technology of the Peoplersquos Republic of China (nos2011ZX02403-005 2011CB706502 and 2013AA041301) Thiswork is also supported by the National Natural ScienceFoundation of China (nos 61370182 and 51175198)
References
[1] A H Slocum Precision Machine Design Prentice Hall NewYork NY USA 1991
[2] M Rajh S Glodez J Flasker K Gotlih and T KostanjevecldquoDesign and analysis of an fMRI compatible haptic robotrdquoRobotics and Computer-Integrated Manufacturing vol 27 no 2pp 267ndash275 2011
[3] S Awtar and G Parmar ldquoDesign of a large range XY nanoposi-tioning systemrdquo Journal of Mechanisms and Robotics vol 5 no2 Article ID 021008 2013
[4] R S Ball A Treatise on the Theory of Screws CambridgeUniversity Press Cambridge UK 1998
[5] K H Hunt Kinematic Geometry of Mechanisms ClarendonPress Oxford UK 1978
[6] K J Waldron ldquoThe constraint analysis of mechanismsrdquo Journalof Mechanisms vol 1 no 2 pp 101ndash114 1966
[7] R Konkar Incremental kinematic analysis and symbolic syn-thesis of mechanisms [PhD dissertation] Stanford UniversityStanford Calif USA 1993
[8] J-S Zhao Z-J Feng and J-X Dong ldquoComputation of theconfiguration degree of freedomof a spatial parallel mechanismby using reciprocal screw theoryrdquo Mechanism and MachineTheory vol 41 no 12 pp 1486ndash1504 2006
[9] J Eddie Baker ldquoScrew system algebra applied to special linkageconfigurationsrdquoMechanism and Machine Theory vol 15 no 4pp 255ndash265 1980
[10] M TMason and J K Salisbury Robot Hands and theMechanicsof Manipulation MIT Press Cambridge Mass USA 1985
[11] Y Fei and X Zhao ldquoAn assembly process modeling and analysisfor robotic multiple peg-in-holerdquo Journal of Intelligent andRobotic Systems vol 36 no 2 pp 175ndash189 2003
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
Mathematical Problems in Engineering 7
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
CoiLL(L1 L2)
Figure 11 Coaxial constraint
f2
P2
h2
F2
g2
L2
B2
f1P1
h1
F1
g1
L1
B1
CoiFF(F1 F2)
d12
Figure 12 Coplanar constraint
The corresponding kinematic joint of coaxial constraint iscylindrical joint
332 Coplanar Constraint We use 119862119900119894119865119865(F1 F2) to denote
the coplanar constraint of planar surfaces F1and F
2 as shown
in Figure 12The doc of 119862119900119894119865119865(F
1 F2) is three which restricts two
relative orientations and one relative translation In otherwords it comprises two dot-1 constraints and one dot-2constraint as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) Φ
3
1198892
(h1 d12 0) (36)
Then the wrench matrix of 119862119900119894119865119865(F1 F2) is given by
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
(37)
The corresponding kinematic joint of coplanar constraintis planar joint
333 Spherical Constraint We use 119862119900119894119875119875(P1P2) to denote
the spherical constraint of two points P1and P
2 as shown in
Figure 13
f2
P2
h2
F2 g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPP(P1P2)
Figure 13 Spherical constraint
f2 P2 h2
F2g2
L2
B2
f1 P1 h1
F1g1
L1
B1
ParFF(F1 F2)
Figure 14 Align constraint
The doc of 119862119900119894119875119875(P1P2) is three which restricts three
relative translations in three unrelated directions In otherwords it comprises one distance constraint which has 119862
3= 0
in Section 323 as shown in
Φ
119904
(119875119894 119875119895 0) (38)
Then the wrench matrix of 119862119900119894119875119875(P1P2) is given by
W119862119900119894119875119875(P
1P2)=
[
[
[
1 0 0 00 1 0 00 0 1 0
]
]
]
(39)
The corresponding kinematic joint of spherical constraintis spherical joint
334 Align Constraint Align constraint (119875119886119903119865119865) is used tomake two planar surfaces F
1and F
2parallel as shown in
Figure 14The doc of 119875119886119903119865119865(F
1 F2) is two which restricts two
relative orientations In other words it comprises two dot-1constraints as shown in
Φ1
1198891
(h1 f2 0) Φ
2
1198891
(h1 g2 0) (40)
8 Mathematical Problems in Engineering
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
DisLF(L1 F2 D)
Figure 15 Distance between line and surface
Then the wrench matrix of 119875119886119903119865119865(F1 F2) is given by
W119875119886119903119865119865(F
1F2)= [
0 h1times f2
0 h1times g2
] (41)
We cannot find a corresponding kinematic joint for alignconstraint among the common kinematic joints
335 Distance between Line and Surface We use 119863119894119904119871119865(L1
F2 119863) to denote the axis L
1and the surface F
2that maintain
a certain distance119863 as shown in Figure 15The doc of 119863119894119904119871119865(L
1 F2 119863) is two which restricts one
relative orientation and one relative translation In otherwords it comprises one dot-1 constraint and one dot-2constraint as shown in
Φ1
1198891
(h1 h2 0) Φ
2
1198892
(h2 d12 119863) (42)
Then the wrench matrix of119863119894119904119871119865(L1 F2 119863) is given by
W119863119894119904119871119865(L
1F2119863)
= [
0 h1times h2
h2 0
] (43)
We also cannot find a corresponding kinematic jointfor distance between line and surface among the commonkinematic joints
336 Point on Line Weuse119862119900119894119875119871(P1 L2) to denote the point
P1on line L
2 as shown in Figure 16
The doc of 119862119900119894119875119871(P1 L2) is two which restricts two
relative translations In other words it comprises two dot-2constraints as shown in
Φ1
1198892
(f2 d12 0) Φ
2
1198892
(g2 d12 0) (44)
Then the wrench matrix of 119862119900119894119875119871(P1 L2) is given by
W119863119894119904119871119865(L
1F2119863)
= [
f2 0
g2 0
] (45)
We still cannot find a corresponding kinematic joint forpoint on line among the common kinematic joints
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPL(P1 L2)
Figure 16 Point on line
4 Recognition of Kinematic Joints
The key to recognizing kinematic joints in terms of geometryconstraints is the reciprocity relation We show how to definegeometry constraints using wrench The space of screws thatdefines the range of wrenches that can be exerted upon abody and the space of screws that defines the range of twiststhat the body can perform without breaking any contacts inthe joints of the mechanism are mutually reciprocal screwspaces When we know the wrenches acting upon a body wecan determine the motions it can perform Conversely if weknow the motions a body is capable of performing then theconstraints that are exerted upon it can also be deduced
41 Algorithm We assume that 119860 is a part in a 3D assemblymodel restricted by parts 119861
1 1198612 119861
119899 The twist space of 119860
is the space of screws reciprocal to the wrenches of contactacting upon it The space of wrenches acting upon 119860 is theunion of all the individual wrenches acting upon it fromthe parts 119861
119894 These individual wrenches are to be considered
as acting alone Algorithm 1 provides the twist space of 119860denoted by T119860
The 119877119890119888119894119901119903119900119888119886119897(P) in Algorithm 1 means to exchangethe first triplet and the second triplet of P For exampleif P = [119886 119887 119888 119889 119890 119891] then we have 119877119890119888119894119901119903119900119888119886119897(P) =
[119889 119890 119891 119886 119887 119888] The recursive nature of the procedurerequires the computation of the wrench space and twistspace of each component This proves to be computationallyexpensive because wrench space and twist space are definedin terms of the reciprocal of one another The operation ofderiving reciprocal screws involves computing the null spaceof matrices The speed of the algorithm is linearly related tothe number of 119861
119894 That is the algorithm is of the order 119874(119899)
42 Reciprocal Product According to reciprocal screw the-ory two screws (a twist and a wrench) are said to be mutuallyreciprocal if their reciprocal product is zero The reciprocal
Mathematical Problems in Engineering 9
BEGINFOR each part Bi which restrict AConvert all the geometry constraints between A and Bi into basic components 119862119895
119860 119895 = 1 119897
Determine the wrenchesW119896119860 119896= 1 119898 acting upon A through C119895
119860 We havem not less than l
Collect all the wrenchesW119896119860into a matrixW
END FORCompute the orthogonal matrixQ in terms ofW by Gram-Schmidt orthogonalization [31]IF Rank(Q) = 6 THENReturn ldquoIMMOBILErdquo
ELSECompute the orthogonal complement matrix P of Q
END IFTA =Reciprocal(P)
END
Algorithm 1 Function twist space
f2
A2
h2F2
g2
B2
f1A1
F1
g1B1
Figure 17 Redundant constraint
product of a twist T1and a wrenchW
1thatisequal to zero is
given by
T1∘W1= T11minus3
sdotW24minus6
+ T14minus6
sdotW21minus3
= [0 0] (46)
Mathematically because reciprocity imposes the relationT1∘W1 we can compute one of these quantities if the other is
known This allows us to compute the twists that a body mayexecute while in contact with other bodies without breakingthe contacts
43 Redundant Constraint When more than one geometryconstraint acts upon the same part because some compo-nents of these constraints may be redundant the resultantmotion is the logical intersection of the individual wrenchmatrix that defines each constraint for example when wemodel a revolute joint 119877
12with two links 119861
1and 119861
2 The
constraints used are 119862119900119894119871119871(A1A2) and 119862119900119894119865119865(F
1 F2) See
Figure 17Also dof(119877
12) = 1Thismeans that doc(119861
1 1198612) = 5 How-
ever we have doc(119862119900119894119871119871(A1A2)) = 4 and doc(119862119900119894119865119865(F
1
F2)) = 3 In other words the dimension of the orthog-
onal wrench space which comprises W119862119900119894119871119871(A
1A2)and
f2
h2
F2g2
B2
f1h1
g1
B1
F1
Figure 18 Recognition of planar joint
W119862119900119894119865119865(F
1F2) is 5 Therefore the motions that a body is
capable of performing while under the action of multiplewrenches are the logical intersection of the motions it canperform under the action of each wrench acting alone Thuswe can eliminate the redundant components of differentgeometry constraints that act upon the same part by usingGram-Schmidt orthogonalization in Algorithm 1
5 Examples
In each of the examples below different types of kinematicjoints are recognized
Example 1 (planar joint) Figure 18 shows a planar jointconsisting of two parts 119861
1and 119861
2 and a coplanar constraint
119862119900119894119865119865(F1 F2)
Following the process outlined in Section 332 we findbased on the definition of the coplanar constraint that thewrench matrix is
W119862119900119894119865119865(F
1F2)=
[
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
]
(47)
10 Mathematical Problems in Engineering
f2h2
F2
g2
B2
f1
h1 g1
B1
F1
f 9984002
h9984002g9984002
F9984002
f 9984001
h9984001
g9984001
F9984001
Figure 19 Recognition of prismatic joint
whereW119862119900119894119865119865(F
1F2)is already an orthogonalmatrixThen the
orthogonal complement matrix P119862119900119894119865119865(F
1F2)is
P119862119900119894119865119865(F
1F2)=
[
[
[
f1 0
g1 0
0 h1
]
]
]
(48)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119865119865(F1F2)) =
[
[
[
0 f1
0 g1
h1 0
]
]
]
(49)
where T12
represents the motion of a planar joint which isillustrated in Section 226 Moreover the recognition processof a cylindrical joint or spherical joint is approximately thesame as that of the planar joint
Example 2 (prismatic joint) Figure 19 shows a prismaticjoint consisting of two parts 119861
1and 119861
2 and two coplanar
constraints 119862119900119894119865119865(F1 F2) and 119862119900119894119865119865(F1015840
1 F10158402)
Following the process outlined in Example 1 we find thatthe wrench matrices between 119861
1and 119861
2are
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
W119862119900119894119865119865(F1015840
1F10158402)=
[
[
[
[
0 h10158401times f10158402
0 h10158401times g10158402
h10158401 0
]
]
]
]
(50)
B0
B1
B2
B3
B4
B5B6
B7
B8
B9
B10
B11
x
y
Figure 20 Hydraulic grab
where W12
= W119862119900119894119865119865(F
1F2)cup W119862119900119894119865119865(F1015840
1F10158402) The orthogonal
matrixQ12ofW12is
Q12=
[
[
[
[
[
[
[
[
[
[
[
0 h1times f2
0 h1times g2
0 h10158401times g10158402
h1 0
h10158401 0
]
]
]
]
]
]
]
]
]
]
]
(51)
The orthogonal complement matrix P12ofQ12is
P12= [g1 0] (52)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
12) = [0 g
1] (53)
where T12
represents the motion of a prismatic joint whichis shown in Section 221 Moreover the recognition processof a revolute joint is approximately the same as that of theprismatic joint
Example 3 (hydraulic grab) Here we present a projectexample A hydraulic grab with three degrees of freedom inCartesian coordinates is illustrated in Figure 20
The hydraulic grab comprises 12 parts 1198610 1198611 1198612 1198613
1198614 1198615 1198616 1198617 1198618 1198619 11986110 and 119861
11 12 coplanar constraints
119862119900119894119865119865(1198610 1198612)119862119900119894119865119865(119861
0 1198611)119862119900119894119865119865(119861
1 1198613)119862119900119894119865119865(119861
1 1198615)
119862119900119894119865119865(1198611 1198614) 119862119900119894119865119865(119861
4 1198616) 119862119900119894119865119865(119861
4 11986111) 119862119900119894119865119865(119861
4
1198617) 119862119900119894119865119865(119861
4 1198618) 119862119900119894119865119865(119861
9 11986110) 119862119900119894119865119865(119861
7 11986110) and
119862119900119894119865119865(11986110 11986111) and 15 coaxial constraints 119862119900119894119871119871(119861
0 1198612)
119862119900119894119871119871(1198610 1198611) 119862119900119894119871119871(119861
1 1198613) 119862119900119894119871119871(119861
2 1198613) 119862119900119894119871119871(119861
1 1198615)
119862119900119894119871119871(1198611 1198614) 119862119900119894119871119871(119861
4 1198616) 119862119900119894119871119871(119861
5 1198616) 119862119900119894119871119871(119861
4
11986111) 119862119900119894119871119871(119861
4 1198617) 119862119900119894119871119871(119861
4 1198618) 119862119900119894119871119871(119861
8 1198619)
119862119900119894119871119871(1198619 11986110) 119862119900119894119871119871(119861
7 11986110) and 119862119900119894119871119871(119861
10 11986111) See
Figure 21Here we present the processes of recognition among 119861
1
1198612 and 119861
3 as shown in Figure 22 The geometry constraint
between 1198612and 119861
3is 119862119900119894119871119871(119861
2 1198613) Following the process
Mathematical Problems in Engineering 11
CoiLL
CoiFF
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
Figure 21 The constraint graph of a hydraulic grab
h2 g2
f 9984003
g9984003
B2
B1
B0
g3h3
g9984001
f 9984001
B3
x
y
Figure 22 Parts 1198611 1198612 and 119861
3in a hydraulic grab
outlined in Section 331 we find based on the definition ofthe coaxial constraint that the wrench matrix is
W119862119900119894119871119871(119861
21198613)=
[
[
[
[
[
[
0 h2times f3
0 h2times g3
f2 0
g2 0
]
]
]
]
]
]
(54)
The orthogonal complement matrix is
P119862119900119894119871119871(119861
21198613)= [
0 h2
h2 0
] (55)
Then the twist matrix T23
between 1198612and 119861
3has the
following form
T23= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119871119871(11986121198613)) = [
h2 0
0 h2
] (56)
The geometry constraints between 1198611
and 1198613
are119862119900119894119871119871(119861
1 1198613) and 119862119900119894119865119865(119861
1 1198613) Thus the wrench matrices
between 1198611and 119861
3are
W119862119900119894119871119871(119861
11198613)=
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
]
]
]
]
]
]
]
W119862119900119894119865119865(119861
11198613)=
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
h10158403 0
]
]
]
]
(57)
where W13
= W119862119900119894119871119871(119861
11198613)cup W119862119900119894119865119865(119861
11198613) The orthogonal
matrixQ13ofW13is given by
Q13=
[
[
[
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
h10158403 0
]
]
]
]
]
]
]
]
]
]
(58)
Then the orthogonal complement matrix P13ofQ13is
P13= [0 h1015840
3] (59)
Therefore the twist matrix T13between 119861
1and 119861
3is
T13= 119877119890119888119894119901119903119900119888119886119897 (P
13) = [h1015840
3 0] (60)
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system basedon reciprocal screw theory According to the geometryconstraints between any two parts in the hydraulic graband Algorithm 1 we obtain the screw system as shown inFigure 23
As mentioned in [4ndash29] we can perform some differenttypes of kinematic analyses using the screw system such asdetermining the degrees of freedom of 119861
1and 119861
4 that is
T1198611
= T10cap (T20cup T32cup T31)
T1198614
= (T1198611
cup T41) cap (T
1198611
cup T51cup T65cup T64)
(61)
where the resultant of the motions that the end-effector ofa purely serial chain may perform due to the action of eachjoint acting alone is the union of that set and the resultant ofthe set of twists allowed separately by multiple contacts upona body is the intersection of the individual twists Thus thedegrees of freedom of 119861
1are equal to the dimension of the
orthogonal twist space of1198611 denoted byT
1198611
In the samewaywe can obtain the degrees of freedomof119861
4and any part in the
hydraulic grab
12 Mathematical Problems in Engineering
T twist
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
T32
T31
T10
T20
T109
T1110
T114
T98
T74 T8
4
T41
T64
T65
T51
T107
Figure 23 The screw system of a hydraulic grab
R revolute jointC cylindrical joint
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
R31
R10
R20
1110
R114
C98
R74 R8
4
R41
R64 R5
1
R107
R
R109 C3
2
C65
Figure 24 The kinematic joint model of a hydraulic grab
According to Sections 222 and 224 the kinematic jointsbetween 119861
2and 119861
3and 119861
1and 119861
3are the cylindrical joint and
revolute joint As a result we can obtain the kinematic jointmodel See Figure 24 If the corresponding kinematic jointbetween any two parts cannot be recognized we can easilyanalyze the modelrsquos geometry constraints
6 Conclusions
This paper investigates a method for determining the kine-matic joints of assemblies restricted by arbitrary combina-tions of geometry constraints using reciprocal screw theoryWe achieve the process of recognition in terms of thebasic components of geometry constraints and kinematicjoints We describe the three basic components of geometricconstraints in terms of wrench At the same time the specificoperation of a matrix space is applied to eliminate the redun-dant components between different geometry constraintsthat act upon the same part A user of this method can auto-matically and easily convert any assembly model into a screwsystem in any CAD system Afterwards the screw system canbe easily used for validating kinematic functions analyzingover- or underconstrained assembly configurations and soon
Additional Points
(i) We provide a new method for converting any 3Dassembly model into a screw system based on recip-rocal screw theory
(ii) We describe the three basic components of geometricconstraints in terms of wrench
(iii) We eliminate the redundant components of differentgeometry constraints that act upon the same part bythe specific operation of a matrix space
(iv) We establish the maps between geometry constraintsand kinematic joints as the reciprocal product of atwist and wrench which is equal to zero
(v) We present several application examples
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
The authors would like to thank the Ministry of Scienceand Technology of the Peoplersquos Republic of China (nos2011ZX02403-005 2011CB706502 and 2013AA041301) Thiswork is also supported by the National Natural ScienceFoundation of China (nos 61370182 and 51175198)
References
[1] A H Slocum Precision Machine Design Prentice Hall NewYork NY USA 1991
[2] M Rajh S Glodez J Flasker K Gotlih and T KostanjevecldquoDesign and analysis of an fMRI compatible haptic robotrdquoRobotics and Computer-Integrated Manufacturing vol 27 no 2pp 267ndash275 2011
[3] S Awtar and G Parmar ldquoDesign of a large range XY nanoposi-tioning systemrdquo Journal of Mechanisms and Robotics vol 5 no2 Article ID 021008 2013
[4] R S Ball A Treatise on the Theory of Screws CambridgeUniversity Press Cambridge UK 1998
[5] K H Hunt Kinematic Geometry of Mechanisms ClarendonPress Oxford UK 1978
[6] K J Waldron ldquoThe constraint analysis of mechanismsrdquo Journalof Mechanisms vol 1 no 2 pp 101ndash114 1966
[7] R Konkar Incremental kinematic analysis and symbolic syn-thesis of mechanisms [PhD dissertation] Stanford UniversityStanford Calif USA 1993
[8] J-S Zhao Z-J Feng and J-X Dong ldquoComputation of theconfiguration degree of freedomof a spatial parallel mechanismby using reciprocal screw theoryrdquo Mechanism and MachineTheory vol 41 no 12 pp 1486ndash1504 2006
[9] J Eddie Baker ldquoScrew system algebra applied to special linkageconfigurationsrdquoMechanism and Machine Theory vol 15 no 4pp 255ndash265 1980
[10] M TMason and J K Salisbury Robot Hands and theMechanicsof Manipulation MIT Press Cambridge Mass USA 1985
[11] Y Fei and X Zhao ldquoAn assembly process modeling and analysisfor robotic multiple peg-in-holerdquo Journal of Intelligent andRobotic Systems vol 36 no 2 pp 175ndash189 2003
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
8 Mathematical Problems in Engineering
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
DisLF(L1 F2 D)
Figure 15 Distance between line and surface
Then the wrench matrix of 119875119886119903119865119865(F1 F2) is given by
W119875119886119903119865119865(F
1F2)= [
0 h1times f2
0 h1times g2
] (41)
We cannot find a corresponding kinematic joint for alignconstraint among the common kinematic joints
335 Distance between Line and Surface We use 119863119894119904119871119865(L1
F2 119863) to denote the axis L
1and the surface F
2that maintain
a certain distance119863 as shown in Figure 15The doc of 119863119894119904119871119865(L
1 F2 119863) is two which restricts one
relative orientation and one relative translation In otherwords it comprises one dot-1 constraint and one dot-2constraint as shown in
Φ1
1198891
(h1 h2 0) Φ
2
1198892
(h2 d12 119863) (42)
Then the wrench matrix of119863119894119904119871119865(L1 F2 119863) is given by
W119863119894119904119871119865(L
1F2119863)
= [
0 h1times h2
h2 0
] (43)
We also cannot find a corresponding kinematic jointfor distance between line and surface among the commonkinematic joints
336 Point on Line Weuse119862119900119894119875119871(P1 L2) to denote the point
P1on line L
2 as shown in Figure 16
The doc of 119862119900119894119875119871(P1 L2) is two which restricts two
relative translations In other words it comprises two dot-2constraints as shown in
Φ1
1198892
(f2 d12 0) Φ
2
1198892
(g2 d12 0) (44)
Then the wrench matrix of 119862119900119894119875119871(P1 L2) is given by
W119863119894119904119871119865(L
1F2119863)
= [
f2 0
g2 0
] (45)
We still cannot find a corresponding kinematic joint forpoint on line among the common kinematic joints
f2 P2 h2
F2g2
L2
B2
f1
P1
h1
F1
g1
L1
B1
CoiPL(P1 L2)
Figure 16 Point on line
4 Recognition of Kinematic Joints
The key to recognizing kinematic joints in terms of geometryconstraints is the reciprocity relation We show how to definegeometry constraints using wrench The space of screws thatdefines the range of wrenches that can be exerted upon abody and the space of screws that defines the range of twiststhat the body can perform without breaking any contacts inthe joints of the mechanism are mutually reciprocal screwspaces When we know the wrenches acting upon a body wecan determine the motions it can perform Conversely if weknow the motions a body is capable of performing then theconstraints that are exerted upon it can also be deduced
41 Algorithm We assume that 119860 is a part in a 3D assemblymodel restricted by parts 119861
1 1198612 119861
119899 The twist space of 119860
is the space of screws reciprocal to the wrenches of contactacting upon it The space of wrenches acting upon 119860 is theunion of all the individual wrenches acting upon it fromthe parts 119861
119894 These individual wrenches are to be considered
as acting alone Algorithm 1 provides the twist space of 119860denoted by T119860
The 119877119890119888119894119901119903119900119888119886119897(P) in Algorithm 1 means to exchangethe first triplet and the second triplet of P For exampleif P = [119886 119887 119888 119889 119890 119891] then we have 119877119890119888119894119901119903119900119888119886119897(P) =
[119889 119890 119891 119886 119887 119888] The recursive nature of the procedurerequires the computation of the wrench space and twistspace of each component This proves to be computationallyexpensive because wrench space and twist space are definedin terms of the reciprocal of one another The operation ofderiving reciprocal screws involves computing the null spaceof matrices The speed of the algorithm is linearly related tothe number of 119861
119894 That is the algorithm is of the order 119874(119899)
42 Reciprocal Product According to reciprocal screw the-ory two screws (a twist and a wrench) are said to be mutuallyreciprocal if their reciprocal product is zero The reciprocal
Mathematical Problems in Engineering 9
BEGINFOR each part Bi which restrict AConvert all the geometry constraints between A and Bi into basic components 119862119895
119860 119895 = 1 119897
Determine the wrenchesW119896119860 119896= 1 119898 acting upon A through C119895
119860 We havem not less than l
Collect all the wrenchesW119896119860into a matrixW
END FORCompute the orthogonal matrixQ in terms ofW by Gram-Schmidt orthogonalization [31]IF Rank(Q) = 6 THENReturn ldquoIMMOBILErdquo
ELSECompute the orthogonal complement matrix P of Q
END IFTA =Reciprocal(P)
END
Algorithm 1 Function twist space
f2
A2
h2F2
g2
B2
f1A1
F1
g1B1
Figure 17 Redundant constraint
product of a twist T1and a wrenchW
1thatisequal to zero is
given by
T1∘W1= T11minus3
sdotW24minus6
+ T14minus6
sdotW21minus3
= [0 0] (46)
Mathematically because reciprocity imposes the relationT1∘W1 we can compute one of these quantities if the other is
known This allows us to compute the twists that a body mayexecute while in contact with other bodies without breakingthe contacts
43 Redundant Constraint When more than one geometryconstraint acts upon the same part because some compo-nents of these constraints may be redundant the resultantmotion is the logical intersection of the individual wrenchmatrix that defines each constraint for example when wemodel a revolute joint 119877
12with two links 119861
1and 119861
2 The
constraints used are 119862119900119894119871119871(A1A2) and 119862119900119894119865119865(F
1 F2) See
Figure 17Also dof(119877
12) = 1Thismeans that doc(119861
1 1198612) = 5 How-
ever we have doc(119862119900119894119871119871(A1A2)) = 4 and doc(119862119900119894119865119865(F
1
F2)) = 3 In other words the dimension of the orthog-
onal wrench space which comprises W119862119900119894119871119871(A
1A2)and
f2
h2
F2g2
B2
f1h1
g1
B1
F1
Figure 18 Recognition of planar joint
W119862119900119894119865119865(F
1F2) is 5 Therefore the motions that a body is
capable of performing while under the action of multiplewrenches are the logical intersection of the motions it canperform under the action of each wrench acting alone Thuswe can eliminate the redundant components of differentgeometry constraints that act upon the same part by usingGram-Schmidt orthogonalization in Algorithm 1
5 Examples
In each of the examples below different types of kinematicjoints are recognized
Example 1 (planar joint) Figure 18 shows a planar jointconsisting of two parts 119861
1and 119861
2 and a coplanar constraint
119862119900119894119865119865(F1 F2)
Following the process outlined in Section 332 we findbased on the definition of the coplanar constraint that thewrench matrix is
W119862119900119894119865119865(F
1F2)=
[
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
]
(47)
10 Mathematical Problems in Engineering
f2h2
F2
g2
B2
f1
h1 g1
B1
F1
f 9984002
h9984002g9984002
F9984002
f 9984001
h9984001
g9984001
F9984001
Figure 19 Recognition of prismatic joint
whereW119862119900119894119865119865(F
1F2)is already an orthogonalmatrixThen the
orthogonal complement matrix P119862119900119894119865119865(F
1F2)is
P119862119900119894119865119865(F
1F2)=
[
[
[
f1 0
g1 0
0 h1
]
]
]
(48)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119865119865(F1F2)) =
[
[
[
0 f1
0 g1
h1 0
]
]
]
(49)
where T12
represents the motion of a planar joint which isillustrated in Section 226 Moreover the recognition processof a cylindrical joint or spherical joint is approximately thesame as that of the planar joint
Example 2 (prismatic joint) Figure 19 shows a prismaticjoint consisting of two parts 119861
1and 119861
2 and two coplanar
constraints 119862119900119894119865119865(F1 F2) and 119862119900119894119865119865(F1015840
1 F10158402)
Following the process outlined in Example 1 we find thatthe wrench matrices between 119861
1and 119861
2are
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
W119862119900119894119865119865(F1015840
1F10158402)=
[
[
[
[
0 h10158401times f10158402
0 h10158401times g10158402
h10158401 0
]
]
]
]
(50)
B0
B1
B2
B3
B4
B5B6
B7
B8
B9
B10
B11
x
y
Figure 20 Hydraulic grab
where W12
= W119862119900119894119865119865(F
1F2)cup W119862119900119894119865119865(F1015840
1F10158402) The orthogonal
matrixQ12ofW12is
Q12=
[
[
[
[
[
[
[
[
[
[
[
0 h1times f2
0 h1times g2
0 h10158401times g10158402
h1 0
h10158401 0
]
]
]
]
]
]
]
]
]
]
]
(51)
The orthogonal complement matrix P12ofQ12is
P12= [g1 0] (52)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
12) = [0 g
1] (53)
where T12
represents the motion of a prismatic joint whichis shown in Section 221 Moreover the recognition processof a revolute joint is approximately the same as that of theprismatic joint
Example 3 (hydraulic grab) Here we present a projectexample A hydraulic grab with three degrees of freedom inCartesian coordinates is illustrated in Figure 20
The hydraulic grab comprises 12 parts 1198610 1198611 1198612 1198613
1198614 1198615 1198616 1198617 1198618 1198619 11986110 and 119861
11 12 coplanar constraints
119862119900119894119865119865(1198610 1198612)119862119900119894119865119865(119861
0 1198611)119862119900119894119865119865(119861
1 1198613)119862119900119894119865119865(119861
1 1198615)
119862119900119894119865119865(1198611 1198614) 119862119900119894119865119865(119861
4 1198616) 119862119900119894119865119865(119861
4 11986111) 119862119900119894119865119865(119861
4
1198617) 119862119900119894119865119865(119861
4 1198618) 119862119900119894119865119865(119861
9 11986110) 119862119900119894119865119865(119861
7 11986110) and
119862119900119894119865119865(11986110 11986111) and 15 coaxial constraints 119862119900119894119871119871(119861
0 1198612)
119862119900119894119871119871(1198610 1198611) 119862119900119894119871119871(119861
1 1198613) 119862119900119894119871119871(119861
2 1198613) 119862119900119894119871119871(119861
1 1198615)
119862119900119894119871119871(1198611 1198614) 119862119900119894119871119871(119861
4 1198616) 119862119900119894119871119871(119861
5 1198616) 119862119900119894119871119871(119861
4
11986111) 119862119900119894119871119871(119861
4 1198617) 119862119900119894119871119871(119861
4 1198618) 119862119900119894119871119871(119861
8 1198619)
119862119900119894119871119871(1198619 11986110) 119862119900119894119871119871(119861
7 11986110) and 119862119900119894119871119871(119861
10 11986111) See
Figure 21Here we present the processes of recognition among 119861
1
1198612 and 119861
3 as shown in Figure 22 The geometry constraint
between 1198612and 119861
3is 119862119900119894119871119871(119861
2 1198613) Following the process
Mathematical Problems in Engineering 11
CoiLL
CoiFF
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
Figure 21 The constraint graph of a hydraulic grab
h2 g2
f 9984003
g9984003
B2
B1
B0
g3h3
g9984001
f 9984001
B3
x
y
Figure 22 Parts 1198611 1198612 and 119861
3in a hydraulic grab
outlined in Section 331 we find based on the definition ofthe coaxial constraint that the wrench matrix is
W119862119900119894119871119871(119861
21198613)=
[
[
[
[
[
[
0 h2times f3
0 h2times g3
f2 0
g2 0
]
]
]
]
]
]
(54)
The orthogonal complement matrix is
P119862119900119894119871119871(119861
21198613)= [
0 h2
h2 0
] (55)
Then the twist matrix T23
between 1198612and 119861
3has the
following form
T23= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119871119871(11986121198613)) = [
h2 0
0 h2
] (56)
The geometry constraints between 1198611
and 1198613
are119862119900119894119871119871(119861
1 1198613) and 119862119900119894119865119865(119861
1 1198613) Thus the wrench matrices
between 1198611and 119861
3are
W119862119900119894119871119871(119861
11198613)=
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
]
]
]
]
]
]
]
W119862119900119894119865119865(119861
11198613)=
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
h10158403 0
]
]
]
]
(57)
where W13
= W119862119900119894119871119871(119861
11198613)cup W119862119900119894119865119865(119861
11198613) The orthogonal
matrixQ13ofW13is given by
Q13=
[
[
[
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
h10158403 0
]
]
]
]
]
]
]
]
]
]
(58)
Then the orthogonal complement matrix P13ofQ13is
P13= [0 h1015840
3] (59)
Therefore the twist matrix T13between 119861
1and 119861
3is
T13= 119877119890119888119894119901119903119900119888119886119897 (P
13) = [h1015840
3 0] (60)
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system basedon reciprocal screw theory According to the geometryconstraints between any two parts in the hydraulic graband Algorithm 1 we obtain the screw system as shown inFigure 23
As mentioned in [4ndash29] we can perform some differenttypes of kinematic analyses using the screw system such asdetermining the degrees of freedom of 119861
1and 119861
4 that is
T1198611
= T10cap (T20cup T32cup T31)
T1198614
= (T1198611
cup T41) cap (T
1198611
cup T51cup T65cup T64)
(61)
where the resultant of the motions that the end-effector ofa purely serial chain may perform due to the action of eachjoint acting alone is the union of that set and the resultant ofthe set of twists allowed separately by multiple contacts upona body is the intersection of the individual twists Thus thedegrees of freedom of 119861
1are equal to the dimension of the
orthogonal twist space of1198611 denoted byT
1198611
In the samewaywe can obtain the degrees of freedomof119861
4and any part in the
hydraulic grab
12 Mathematical Problems in Engineering
T twist
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
T32
T31
T10
T20
T109
T1110
T114
T98
T74 T8
4
T41
T64
T65
T51
T107
Figure 23 The screw system of a hydraulic grab
R revolute jointC cylindrical joint
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
R31
R10
R20
1110
R114
C98
R74 R8
4
R41
R64 R5
1
R107
R
R109 C3
2
C65
Figure 24 The kinematic joint model of a hydraulic grab
According to Sections 222 and 224 the kinematic jointsbetween 119861
2and 119861
3and 119861
1and 119861
3are the cylindrical joint and
revolute joint As a result we can obtain the kinematic jointmodel See Figure 24 If the corresponding kinematic jointbetween any two parts cannot be recognized we can easilyanalyze the modelrsquos geometry constraints
6 Conclusions
This paper investigates a method for determining the kine-matic joints of assemblies restricted by arbitrary combina-tions of geometry constraints using reciprocal screw theoryWe achieve the process of recognition in terms of thebasic components of geometry constraints and kinematicjoints We describe the three basic components of geometricconstraints in terms of wrench At the same time the specificoperation of a matrix space is applied to eliminate the redun-dant components between different geometry constraintsthat act upon the same part A user of this method can auto-matically and easily convert any assembly model into a screwsystem in any CAD system Afterwards the screw system canbe easily used for validating kinematic functions analyzingover- or underconstrained assembly configurations and soon
Additional Points
(i) We provide a new method for converting any 3Dassembly model into a screw system based on recip-rocal screw theory
(ii) We describe the three basic components of geometricconstraints in terms of wrench
(iii) We eliminate the redundant components of differentgeometry constraints that act upon the same part bythe specific operation of a matrix space
(iv) We establish the maps between geometry constraintsand kinematic joints as the reciprocal product of atwist and wrench which is equal to zero
(v) We present several application examples
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
The authors would like to thank the Ministry of Scienceand Technology of the Peoplersquos Republic of China (nos2011ZX02403-005 2011CB706502 and 2013AA041301) Thiswork is also supported by the National Natural ScienceFoundation of China (nos 61370182 and 51175198)
References
[1] A H Slocum Precision Machine Design Prentice Hall NewYork NY USA 1991
[2] M Rajh S Glodez J Flasker K Gotlih and T KostanjevecldquoDesign and analysis of an fMRI compatible haptic robotrdquoRobotics and Computer-Integrated Manufacturing vol 27 no 2pp 267ndash275 2011
[3] S Awtar and G Parmar ldquoDesign of a large range XY nanoposi-tioning systemrdquo Journal of Mechanisms and Robotics vol 5 no2 Article ID 021008 2013
[4] R S Ball A Treatise on the Theory of Screws CambridgeUniversity Press Cambridge UK 1998
[5] K H Hunt Kinematic Geometry of Mechanisms ClarendonPress Oxford UK 1978
[6] K J Waldron ldquoThe constraint analysis of mechanismsrdquo Journalof Mechanisms vol 1 no 2 pp 101ndash114 1966
[7] R Konkar Incremental kinematic analysis and symbolic syn-thesis of mechanisms [PhD dissertation] Stanford UniversityStanford Calif USA 1993
[8] J-S Zhao Z-J Feng and J-X Dong ldquoComputation of theconfiguration degree of freedomof a spatial parallel mechanismby using reciprocal screw theoryrdquo Mechanism and MachineTheory vol 41 no 12 pp 1486ndash1504 2006
[9] J Eddie Baker ldquoScrew system algebra applied to special linkageconfigurationsrdquoMechanism and Machine Theory vol 15 no 4pp 255ndash265 1980
[10] M TMason and J K Salisbury Robot Hands and theMechanicsof Manipulation MIT Press Cambridge Mass USA 1985
[11] Y Fei and X Zhao ldquoAn assembly process modeling and analysisfor robotic multiple peg-in-holerdquo Journal of Intelligent andRobotic Systems vol 36 no 2 pp 175ndash189 2003
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
Mathematical Problems in Engineering 9
BEGINFOR each part Bi which restrict AConvert all the geometry constraints between A and Bi into basic components 119862119895
119860 119895 = 1 119897
Determine the wrenchesW119896119860 119896= 1 119898 acting upon A through C119895
119860 We havem not less than l
Collect all the wrenchesW119896119860into a matrixW
END FORCompute the orthogonal matrixQ in terms ofW by Gram-Schmidt orthogonalization [31]IF Rank(Q) = 6 THENReturn ldquoIMMOBILErdquo
ELSECompute the orthogonal complement matrix P of Q
END IFTA =Reciprocal(P)
END
Algorithm 1 Function twist space
f2
A2
h2F2
g2
B2
f1A1
F1
g1B1
Figure 17 Redundant constraint
product of a twist T1and a wrenchW
1thatisequal to zero is
given by
T1∘W1= T11minus3
sdotW24minus6
+ T14minus6
sdotW21minus3
= [0 0] (46)
Mathematically because reciprocity imposes the relationT1∘W1 we can compute one of these quantities if the other is
known This allows us to compute the twists that a body mayexecute while in contact with other bodies without breakingthe contacts
43 Redundant Constraint When more than one geometryconstraint acts upon the same part because some compo-nents of these constraints may be redundant the resultantmotion is the logical intersection of the individual wrenchmatrix that defines each constraint for example when wemodel a revolute joint 119877
12with two links 119861
1and 119861
2 The
constraints used are 119862119900119894119871119871(A1A2) and 119862119900119894119865119865(F
1 F2) See
Figure 17Also dof(119877
12) = 1Thismeans that doc(119861
1 1198612) = 5 How-
ever we have doc(119862119900119894119871119871(A1A2)) = 4 and doc(119862119900119894119865119865(F
1
F2)) = 3 In other words the dimension of the orthog-
onal wrench space which comprises W119862119900119894119871119871(A
1A2)and
f2
h2
F2g2
B2
f1h1
g1
B1
F1
Figure 18 Recognition of planar joint
W119862119900119894119865119865(F
1F2) is 5 Therefore the motions that a body is
capable of performing while under the action of multiplewrenches are the logical intersection of the motions it canperform under the action of each wrench acting alone Thuswe can eliminate the redundant components of differentgeometry constraints that act upon the same part by usingGram-Schmidt orthogonalization in Algorithm 1
5 Examples
In each of the examples below different types of kinematicjoints are recognized
Example 1 (planar joint) Figure 18 shows a planar jointconsisting of two parts 119861
1and 119861
2 and a coplanar constraint
119862119900119894119865119865(F1 F2)
Following the process outlined in Section 332 we findbased on the definition of the coplanar constraint that thewrench matrix is
W119862119900119894119865119865(F
1F2)=
[
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
]
(47)
10 Mathematical Problems in Engineering
f2h2
F2
g2
B2
f1
h1 g1
B1
F1
f 9984002
h9984002g9984002
F9984002
f 9984001
h9984001
g9984001
F9984001
Figure 19 Recognition of prismatic joint
whereW119862119900119894119865119865(F
1F2)is already an orthogonalmatrixThen the
orthogonal complement matrix P119862119900119894119865119865(F
1F2)is
P119862119900119894119865119865(F
1F2)=
[
[
[
f1 0
g1 0
0 h1
]
]
]
(48)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119865119865(F1F2)) =
[
[
[
0 f1
0 g1
h1 0
]
]
]
(49)
where T12
represents the motion of a planar joint which isillustrated in Section 226 Moreover the recognition processof a cylindrical joint or spherical joint is approximately thesame as that of the planar joint
Example 2 (prismatic joint) Figure 19 shows a prismaticjoint consisting of two parts 119861
1and 119861
2 and two coplanar
constraints 119862119900119894119865119865(F1 F2) and 119862119900119894119865119865(F1015840
1 F10158402)
Following the process outlined in Example 1 we find thatthe wrench matrices between 119861
1and 119861
2are
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
W119862119900119894119865119865(F1015840
1F10158402)=
[
[
[
[
0 h10158401times f10158402
0 h10158401times g10158402
h10158401 0
]
]
]
]
(50)
B0
B1
B2
B3
B4
B5B6
B7
B8
B9
B10
B11
x
y
Figure 20 Hydraulic grab
where W12
= W119862119900119894119865119865(F
1F2)cup W119862119900119894119865119865(F1015840
1F10158402) The orthogonal
matrixQ12ofW12is
Q12=
[
[
[
[
[
[
[
[
[
[
[
0 h1times f2
0 h1times g2
0 h10158401times g10158402
h1 0
h10158401 0
]
]
]
]
]
]
]
]
]
]
]
(51)
The orthogonal complement matrix P12ofQ12is
P12= [g1 0] (52)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
12) = [0 g
1] (53)
where T12
represents the motion of a prismatic joint whichis shown in Section 221 Moreover the recognition processof a revolute joint is approximately the same as that of theprismatic joint
Example 3 (hydraulic grab) Here we present a projectexample A hydraulic grab with three degrees of freedom inCartesian coordinates is illustrated in Figure 20
The hydraulic grab comprises 12 parts 1198610 1198611 1198612 1198613
1198614 1198615 1198616 1198617 1198618 1198619 11986110 and 119861
11 12 coplanar constraints
119862119900119894119865119865(1198610 1198612)119862119900119894119865119865(119861
0 1198611)119862119900119894119865119865(119861
1 1198613)119862119900119894119865119865(119861
1 1198615)
119862119900119894119865119865(1198611 1198614) 119862119900119894119865119865(119861
4 1198616) 119862119900119894119865119865(119861
4 11986111) 119862119900119894119865119865(119861
4
1198617) 119862119900119894119865119865(119861
4 1198618) 119862119900119894119865119865(119861
9 11986110) 119862119900119894119865119865(119861
7 11986110) and
119862119900119894119865119865(11986110 11986111) and 15 coaxial constraints 119862119900119894119871119871(119861
0 1198612)
119862119900119894119871119871(1198610 1198611) 119862119900119894119871119871(119861
1 1198613) 119862119900119894119871119871(119861
2 1198613) 119862119900119894119871119871(119861
1 1198615)
119862119900119894119871119871(1198611 1198614) 119862119900119894119871119871(119861
4 1198616) 119862119900119894119871119871(119861
5 1198616) 119862119900119894119871119871(119861
4
11986111) 119862119900119894119871119871(119861
4 1198617) 119862119900119894119871119871(119861
4 1198618) 119862119900119894119871119871(119861
8 1198619)
119862119900119894119871119871(1198619 11986110) 119862119900119894119871119871(119861
7 11986110) and 119862119900119894119871119871(119861
10 11986111) See
Figure 21Here we present the processes of recognition among 119861
1
1198612 and 119861
3 as shown in Figure 22 The geometry constraint
between 1198612and 119861
3is 119862119900119894119871119871(119861
2 1198613) Following the process
Mathematical Problems in Engineering 11
CoiLL
CoiFF
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
Figure 21 The constraint graph of a hydraulic grab
h2 g2
f 9984003
g9984003
B2
B1
B0
g3h3
g9984001
f 9984001
B3
x
y
Figure 22 Parts 1198611 1198612 and 119861
3in a hydraulic grab
outlined in Section 331 we find based on the definition ofthe coaxial constraint that the wrench matrix is
W119862119900119894119871119871(119861
21198613)=
[
[
[
[
[
[
0 h2times f3
0 h2times g3
f2 0
g2 0
]
]
]
]
]
]
(54)
The orthogonal complement matrix is
P119862119900119894119871119871(119861
21198613)= [
0 h2
h2 0
] (55)
Then the twist matrix T23
between 1198612and 119861
3has the
following form
T23= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119871119871(11986121198613)) = [
h2 0
0 h2
] (56)
The geometry constraints between 1198611
and 1198613
are119862119900119894119871119871(119861
1 1198613) and 119862119900119894119865119865(119861
1 1198613) Thus the wrench matrices
between 1198611and 119861
3are
W119862119900119894119871119871(119861
11198613)=
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
]
]
]
]
]
]
]
W119862119900119894119865119865(119861
11198613)=
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
h10158403 0
]
]
]
]
(57)
where W13
= W119862119900119894119871119871(119861
11198613)cup W119862119900119894119865119865(119861
11198613) The orthogonal
matrixQ13ofW13is given by
Q13=
[
[
[
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
h10158403 0
]
]
]
]
]
]
]
]
]
]
(58)
Then the orthogonal complement matrix P13ofQ13is
P13= [0 h1015840
3] (59)
Therefore the twist matrix T13between 119861
1and 119861
3is
T13= 119877119890119888119894119901119903119900119888119886119897 (P
13) = [h1015840
3 0] (60)
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system basedon reciprocal screw theory According to the geometryconstraints between any two parts in the hydraulic graband Algorithm 1 we obtain the screw system as shown inFigure 23
As mentioned in [4ndash29] we can perform some differenttypes of kinematic analyses using the screw system such asdetermining the degrees of freedom of 119861
1and 119861
4 that is
T1198611
= T10cap (T20cup T32cup T31)
T1198614
= (T1198611
cup T41) cap (T
1198611
cup T51cup T65cup T64)
(61)
where the resultant of the motions that the end-effector ofa purely serial chain may perform due to the action of eachjoint acting alone is the union of that set and the resultant ofthe set of twists allowed separately by multiple contacts upona body is the intersection of the individual twists Thus thedegrees of freedom of 119861
1are equal to the dimension of the
orthogonal twist space of1198611 denoted byT
1198611
In the samewaywe can obtain the degrees of freedomof119861
4and any part in the
hydraulic grab
12 Mathematical Problems in Engineering
T twist
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
T32
T31
T10
T20
T109
T1110
T114
T98
T74 T8
4
T41
T64
T65
T51
T107
Figure 23 The screw system of a hydraulic grab
R revolute jointC cylindrical joint
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
R31
R10
R20
1110
R114
C98
R74 R8
4
R41
R64 R5
1
R107
R
R109 C3
2
C65
Figure 24 The kinematic joint model of a hydraulic grab
According to Sections 222 and 224 the kinematic jointsbetween 119861
2and 119861
3and 119861
1and 119861
3are the cylindrical joint and
revolute joint As a result we can obtain the kinematic jointmodel See Figure 24 If the corresponding kinematic jointbetween any two parts cannot be recognized we can easilyanalyze the modelrsquos geometry constraints
6 Conclusions
This paper investigates a method for determining the kine-matic joints of assemblies restricted by arbitrary combina-tions of geometry constraints using reciprocal screw theoryWe achieve the process of recognition in terms of thebasic components of geometry constraints and kinematicjoints We describe the three basic components of geometricconstraints in terms of wrench At the same time the specificoperation of a matrix space is applied to eliminate the redun-dant components between different geometry constraintsthat act upon the same part A user of this method can auto-matically and easily convert any assembly model into a screwsystem in any CAD system Afterwards the screw system canbe easily used for validating kinematic functions analyzingover- or underconstrained assembly configurations and soon
Additional Points
(i) We provide a new method for converting any 3Dassembly model into a screw system based on recip-rocal screw theory
(ii) We describe the three basic components of geometricconstraints in terms of wrench
(iii) We eliminate the redundant components of differentgeometry constraints that act upon the same part bythe specific operation of a matrix space
(iv) We establish the maps between geometry constraintsand kinematic joints as the reciprocal product of atwist and wrench which is equal to zero
(v) We present several application examples
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
The authors would like to thank the Ministry of Scienceand Technology of the Peoplersquos Republic of China (nos2011ZX02403-005 2011CB706502 and 2013AA041301) Thiswork is also supported by the National Natural ScienceFoundation of China (nos 61370182 and 51175198)
References
[1] A H Slocum Precision Machine Design Prentice Hall NewYork NY USA 1991
[2] M Rajh S Glodez J Flasker K Gotlih and T KostanjevecldquoDesign and analysis of an fMRI compatible haptic robotrdquoRobotics and Computer-Integrated Manufacturing vol 27 no 2pp 267ndash275 2011
[3] S Awtar and G Parmar ldquoDesign of a large range XY nanoposi-tioning systemrdquo Journal of Mechanisms and Robotics vol 5 no2 Article ID 021008 2013
[4] R S Ball A Treatise on the Theory of Screws CambridgeUniversity Press Cambridge UK 1998
[5] K H Hunt Kinematic Geometry of Mechanisms ClarendonPress Oxford UK 1978
[6] K J Waldron ldquoThe constraint analysis of mechanismsrdquo Journalof Mechanisms vol 1 no 2 pp 101ndash114 1966
[7] R Konkar Incremental kinematic analysis and symbolic syn-thesis of mechanisms [PhD dissertation] Stanford UniversityStanford Calif USA 1993
[8] J-S Zhao Z-J Feng and J-X Dong ldquoComputation of theconfiguration degree of freedomof a spatial parallel mechanismby using reciprocal screw theoryrdquo Mechanism and MachineTheory vol 41 no 12 pp 1486ndash1504 2006
[9] J Eddie Baker ldquoScrew system algebra applied to special linkageconfigurationsrdquoMechanism and Machine Theory vol 15 no 4pp 255ndash265 1980
[10] M TMason and J K Salisbury Robot Hands and theMechanicsof Manipulation MIT Press Cambridge Mass USA 1985
[11] Y Fei and X Zhao ldquoAn assembly process modeling and analysisfor robotic multiple peg-in-holerdquo Journal of Intelligent andRobotic Systems vol 36 no 2 pp 175ndash189 2003
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
10 Mathematical Problems in Engineering
f2h2
F2
g2
B2
f1
h1 g1
B1
F1
f 9984002
h9984002g9984002
F9984002
f 9984001
h9984001
g9984001
F9984001
Figure 19 Recognition of prismatic joint
whereW119862119900119894119865119865(F
1F2)is already an orthogonalmatrixThen the
orthogonal complement matrix P119862119900119894119865119865(F
1F2)is
P119862119900119894119865119865(F
1F2)=
[
[
[
f1 0
g1 0
0 h1
]
]
]
(48)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119865119865(F1F2)) =
[
[
[
0 f1
0 g1
h1 0
]
]
]
(49)
where T12
represents the motion of a planar joint which isillustrated in Section 226 Moreover the recognition processof a cylindrical joint or spherical joint is approximately thesame as that of the planar joint
Example 2 (prismatic joint) Figure 19 shows a prismaticjoint consisting of two parts 119861
1and 119861
2 and two coplanar
constraints 119862119900119894119865119865(F1 F2) and 119862119900119894119865119865(F1015840
1 F10158402)
Following the process outlined in Example 1 we find thatthe wrench matrices between 119861
1and 119861
2are
W119862119900119894119865119865(F
1F2)=
[
[
[
0 h1times f2
0 h1times g2
h1 0
]
]
]
W119862119900119894119865119865(F1015840
1F10158402)=
[
[
[
[
0 h10158401times f10158402
0 h10158401times g10158402
h10158401 0
]
]
]
]
(50)
B0
B1
B2
B3
B4
B5B6
B7
B8
B9
B10
B11
x
y
Figure 20 Hydraulic grab
where W12
= W119862119900119894119865119865(F
1F2)cup W119862119900119894119865119865(F1015840
1F10158402) The orthogonal
matrixQ12ofW12is
Q12=
[
[
[
[
[
[
[
[
[
[
[
0 h1times f2
0 h1times g2
0 h10158401times g10158402
h1 0
h10158401 0
]
]
]
]
]
]
]
]
]
]
]
(51)
The orthogonal complement matrix P12ofQ12is
P12= [g1 0] (52)
Therefore the twist matrixT12between 119861
1and 119861
2has the
following form
T12= 119877119890119888119894119901119903119900119888119886119897 (P
12) = [0 g
1] (53)
where T12
represents the motion of a prismatic joint whichis shown in Section 221 Moreover the recognition processof a revolute joint is approximately the same as that of theprismatic joint
Example 3 (hydraulic grab) Here we present a projectexample A hydraulic grab with three degrees of freedom inCartesian coordinates is illustrated in Figure 20
The hydraulic grab comprises 12 parts 1198610 1198611 1198612 1198613
1198614 1198615 1198616 1198617 1198618 1198619 11986110 and 119861
11 12 coplanar constraints
119862119900119894119865119865(1198610 1198612)119862119900119894119865119865(119861
0 1198611)119862119900119894119865119865(119861
1 1198613)119862119900119894119865119865(119861
1 1198615)
119862119900119894119865119865(1198611 1198614) 119862119900119894119865119865(119861
4 1198616) 119862119900119894119865119865(119861
4 11986111) 119862119900119894119865119865(119861
4
1198617) 119862119900119894119865119865(119861
4 1198618) 119862119900119894119865119865(119861
9 11986110) 119862119900119894119865119865(119861
7 11986110) and
119862119900119894119865119865(11986110 11986111) and 15 coaxial constraints 119862119900119894119871119871(119861
0 1198612)
119862119900119894119871119871(1198610 1198611) 119862119900119894119871119871(119861
1 1198613) 119862119900119894119871119871(119861
2 1198613) 119862119900119894119871119871(119861
1 1198615)
119862119900119894119871119871(1198611 1198614) 119862119900119894119871119871(119861
4 1198616) 119862119900119894119871119871(119861
5 1198616) 119862119900119894119871119871(119861
4
11986111) 119862119900119894119871119871(119861
4 1198617) 119862119900119894119871119871(119861
4 1198618) 119862119900119894119871119871(119861
8 1198619)
119862119900119894119871119871(1198619 11986110) 119862119900119894119871119871(119861
7 11986110) and 119862119900119894119871119871(119861
10 11986111) See
Figure 21Here we present the processes of recognition among 119861
1
1198612 and 119861
3 as shown in Figure 22 The geometry constraint
between 1198612and 119861
3is 119862119900119894119871119871(119861
2 1198613) Following the process
Mathematical Problems in Engineering 11
CoiLL
CoiFF
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
Figure 21 The constraint graph of a hydraulic grab
h2 g2
f 9984003
g9984003
B2
B1
B0
g3h3
g9984001
f 9984001
B3
x
y
Figure 22 Parts 1198611 1198612 and 119861
3in a hydraulic grab
outlined in Section 331 we find based on the definition ofthe coaxial constraint that the wrench matrix is
W119862119900119894119871119871(119861
21198613)=
[
[
[
[
[
[
0 h2times f3
0 h2times g3
f2 0
g2 0
]
]
]
]
]
]
(54)
The orthogonal complement matrix is
P119862119900119894119871119871(119861
21198613)= [
0 h2
h2 0
] (55)
Then the twist matrix T23
between 1198612and 119861
3has the
following form
T23= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119871119871(11986121198613)) = [
h2 0
0 h2
] (56)
The geometry constraints between 1198611
and 1198613
are119862119900119894119871119871(119861
1 1198613) and 119862119900119894119865119865(119861
1 1198613) Thus the wrench matrices
between 1198611and 119861
3are
W119862119900119894119871119871(119861
11198613)=
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
]
]
]
]
]
]
]
W119862119900119894119865119865(119861
11198613)=
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
h10158403 0
]
]
]
]
(57)
where W13
= W119862119900119894119871119871(119861
11198613)cup W119862119900119894119865119865(119861
11198613) The orthogonal
matrixQ13ofW13is given by
Q13=
[
[
[
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
h10158403 0
]
]
]
]
]
]
]
]
]
]
(58)
Then the orthogonal complement matrix P13ofQ13is
P13= [0 h1015840
3] (59)
Therefore the twist matrix T13between 119861
1and 119861
3is
T13= 119877119890119888119894119901119903119900119888119886119897 (P
13) = [h1015840
3 0] (60)
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system basedon reciprocal screw theory According to the geometryconstraints between any two parts in the hydraulic graband Algorithm 1 we obtain the screw system as shown inFigure 23
As mentioned in [4ndash29] we can perform some differenttypes of kinematic analyses using the screw system such asdetermining the degrees of freedom of 119861
1and 119861
4 that is
T1198611
= T10cap (T20cup T32cup T31)
T1198614
= (T1198611
cup T41) cap (T
1198611
cup T51cup T65cup T64)
(61)
where the resultant of the motions that the end-effector ofa purely serial chain may perform due to the action of eachjoint acting alone is the union of that set and the resultant ofthe set of twists allowed separately by multiple contacts upona body is the intersection of the individual twists Thus thedegrees of freedom of 119861
1are equal to the dimension of the
orthogonal twist space of1198611 denoted byT
1198611
In the samewaywe can obtain the degrees of freedomof119861
4and any part in the
hydraulic grab
12 Mathematical Problems in Engineering
T twist
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
T32
T31
T10
T20
T109
T1110
T114
T98
T74 T8
4
T41
T64
T65
T51
T107
Figure 23 The screw system of a hydraulic grab
R revolute jointC cylindrical joint
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
R31
R10
R20
1110
R114
C98
R74 R8
4
R41
R64 R5
1
R107
R
R109 C3
2
C65
Figure 24 The kinematic joint model of a hydraulic grab
According to Sections 222 and 224 the kinematic jointsbetween 119861
2and 119861
3and 119861
1and 119861
3are the cylindrical joint and
revolute joint As a result we can obtain the kinematic jointmodel See Figure 24 If the corresponding kinematic jointbetween any two parts cannot be recognized we can easilyanalyze the modelrsquos geometry constraints
6 Conclusions
This paper investigates a method for determining the kine-matic joints of assemblies restricted by arbitrary combina-tions of geometry constraints using reciprocal screw theoryWe achieve the process of recognition in terms of thebasic components of geometry constraints and kinematicjoints We describe the three basic components of geometricconstraints in terms of wrench At the same time the specificoperation of a matrix space is applied to eliminate the redun-dant components between different geometry constraintsthat act upon the same part A user of this method can auto-matically and easily convert any assembly model into a screwsystem in any CAD system Afterwards the screw system canbe easily used for validating kinematic functions analyzingover- or underconstrained assembly configurations and soon
Additional Points
(i) We provide a new method for converting any 3Dassembly model into a screw system based on recip-rocal screw theory
(ii) We describe the three basic components of geometricconstraints in terms of wrench
(iii) We eliminate the redundant components of differentgeometry constraints that act upon the same part bythe specific operation of a matrix space
(iv) We establish the maps between geometry constraintsand kinematic joints as the reciprocal product of atwist and wrench which is equal to zero
(v) We present several application examples
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
The authors would like to thank the Ministry of Scienceand Technology of the Peoplersquos Republic of China (nos2011ZX02403-005 2011CB706502 and 2013AA041301) Thiswork is also supported by the National Natural ScienceFoundation of China (nos 61370182 and 51175198)
References
[1] A H Slocum Precision Machine Design Prentice Hall NewYork NY USA 1991
[2] M Rajh S Glodez J Flasker K Gotlih and T KostanjevecldquoDesign and analysis of an fMRI compatible haptic robotrdquoRobotics and Computer-Integrated Manufacturing vol 27 no 2pp 267ndash275 2011
[3] S Awtar and G Parmar ldquoDesign of a large range XY nanoposi-tioning systemrdquo Journal of Mechanisms and Robotics vol 5 no2 Article ID 021008 2013
[4] R S Ball A Treatise on the Theory of Screws CambridgeUniversity Press Cambridge UK 1998
[5] K H Hunt Kinematic Geometry of Mechanisms ClarendonPress Oxford UK 1978
[6] K J Waldron ldquoThe constraint analysis of mechanismsrdquo Journalof Mechanisms vol 1 no 2 pp 101ndash114 1966
[7] R Konkar Incremental kinematic analysis and symbolic syn-thesis of mechanisms [PhD dissertation] Stanford UniversityStanford Calif USA 1993
[8] J-S Zhao Z-J Feng and J-X Dong ldquoComputation of theconfiguration degree of freedomof a spatial parallel mechanismby using reciprocal screw theoryrdquo Mechanism and MachineTheory vol 41 no 12 pp 1486ndash1504 2006
[9] J Eddie Baker ldquoScrew system algebra applied to special linkageconfigurationsrdquoMechanism and Machine Theory vol 15 no 4pp 255ndash265 1980
[10] M TMason and J K Salisbury Robot Hands and theMechanicsof Manipulation MIT Press Cambridge Mass USA 1985
[11] Y Fei and X Zhao ldquoAn assembly process modeling and analysisfor robotic multiple peg-in-holerdquo Journal of Intelligent andRobotic Systems vol 36 no 2 pp 175ndash189 2003
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
Mathematical Problems in Engineering 11
CoiLL
CoiFF
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
Figure 21 The constraint graph of a hydraulic grab
h2 g2
f 9984003
g9984003
B2
B1
B0
g3h3
g9984001
f 9984001
B3
x
y
Figure 22 Parts 1198611 1198612 and 119861
3in a hydraulic grab
outlined in Section 331 we find based on the definition ofthe coaxial constraint that the wrench matrix is
W119862119900119894119871119871(119861
21198613)=
[
[
[
[
[
[
0 h2times f3
0 h2times g3
f2 0
g2 0
]
]
]
]
]
]
(54)
The orthogonal complement matrix is
P119862119900119894119871119871(119861
21198613)= [
0 h2
h2 0
] (55)
Then the twist matrix T23
between 1198612and 119861
3has the
following form
T23= 119877119890119888119894119901119903119900119888119886119897 (P
119862119900119894119871119871(11986121198613)) = [
h2 0
0 h2
] (56)
The geometry constraints between 1198611
and 1198613
are119862119900119894119871119871(119861
1 1198613) and 119862119900119894119865119865(119861
1 1198613) Thus the wrench matrices
between 1198611and 119861
3are
W119862119900119894119871119871(119861
11198613)=
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
]
]
]
]
]
]
]
W119862119900119894119865119865(119861
11198613)=
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
h10158403 0
]
]
]
]
(57)
where W13
= W119862119900119894119871119871(119861
11198613)cup W119862119900119894119865119865(119861
11198613) The orthogonal
matrixQ13ofW13is given by
Q13=
[
[
[
[
[
[
[
[
[
[
0 h10158403times f10158401
0 h10158403times g10158401
f10158403 0
g10158403 0
h10158403 0
]
]
]
]
]
]
]
]
]
]
(58)
Then the orthogonal complement matrix P13ofQ13is
P13= [0 h1015840
3] (59)
Therefore the twist matrix T13between 119861
1and 119861
3is
T13= 119877119890119888119894119901119903119900119888119886119897 (P
13) = [h1015840
3 0] (60)
In this paper we propose a new method to automaticallyconvert a 3D assembly model into a screw system basedon reciprocal screw theory According to the geometryconstraints between any two parts in the hydraulic graband Algorithm 1 we obtain the screw system as shown inFigure 23
As mentioned in [4ndash29] we can perform some differenttypes of kinematic analyses using the screw system such asdetermining the degrees of freedom of 119861
1and 119861
4 that is
T1198611
= T10cap (T20cup T32cup T31)
T1198614
= (T1198611
cup T41) cap (T
1198611
cup T51cup T65cup T64)
(61)
where the resultant of the motions that the end-effector ofa purely serial chain may perform due to the action of eachjoint acting alone is the union of that set and the resultant ofthe set of twists allowed separately by multiple contacts upona body is the intersection of the individual twists Thus thedegrees of freedom of 119861
1are equal to the dimension of the
orthogonal twist space of1198611 denoted byT
1198611
In the samewaywe can obtain the degrees of freedomof119861
4and any part in the
hydraulic grab
12 Mathematical Problems in Engineering
T twist
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
T32
T31
T10
T20
T109
T1110
T114
T98
T74 T8
4
T41
T64
T65
T51
T107
Figure 23 The screw system of a hydraulic grab
R revolute jointC cylindrical joint
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
R31
R10
R20
1110
R114
C98
R74 R8
4
R41
R64 R5
1
R107
R
R109 C3
2
C65
Figure 24 The kinematic joint model of a hydraulic grab
According to Sections 222 and 224 the kinematic jointsbetween 119861
2and 119861
3and 119861
1and 119861
3are the cylindrical joint and
revolute joint As a result we can obtain the kinematic jointmodel See Figure 24 If the corresponding kinematic jointbetween any two parts cannot be recognized we can easilyanalyze the modelrsquos geometry constraints
6 Conclusions
This paper investigates a method for determining the kine-matic joints of assemblies restricted by arbitrary combina-tions of geometry constraints using reciprocal screw theoryWe achieve the process of recognition in terms of thebasic components of geometry constraints and kinematicjoints We describe the three basic components of geometricconstraints in terms of wrench At the same time the specificoperation of a matrix space is applied to eliminate the redun-dant components between different geometry constraintsthat act upon the same part A user of this method can auto-matically and easily convert any assembly model into a screwsystem in any CAD system Afterwards the screw system canbe easily used for validating kinematic functions analyzingover- or underconstrained assembly configurations and soon
Additional Points
(i) We provide a new method for converting any 3Dassembly model into a screw system based on recip-rocal screw theory
(ii) We describe the three basic components of geometricconstraints in terms of wrench
(iii) We eliminate the redundant components of differentgeometry constraints that act upon the same part bythe specific operation of a matrix space
(iv) We establish the maps between geometry constraintsand kinematic joints as the reciprocal product of atwist and wrench which is equal to zero
(v) We present several application examples
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
The authors would like to thank the Ministry of Scienceand Technology of the Peoplersquos Republic of China (nos2011ZX02403-005 2011CB706502 and 2013AA041301) Thiswork is also supported by the National Natural ScienceFoundation of China (nos 61370182 and 51175198)
References
[1] A H Slocum Precision Machine Design Prentice Hall NewYork NY USA 1991
[2] M Rajh S Glodez J Flasker K Gotlih and T KostanjevecldquoDesign and analysis of an fMRI compatible haptic robotrdquoRobotics and Computer-Integrated Manufacturing vol 27 no 2pp 267ndash275 2011
[3] S Awtar and G Parmar ldquoDesign of a large range XY nanoposi-tioning systemrdquo Journal of Mechanisms and Robotics vol 5 no2 Article ID 021008 2013
[4] R S Ball A Treatise on the Theory of Screws CambridgeUniversity Press Cambridge UK 1998
[5] K H Hunt Kinematic Geometry of Mechanisms ClarendonPress Oxford UK 1978
[6] K J Waldron ldquoThe constraint analysis of mechanismsrdquo Journalof Mechanisms vol 1 no 2 pp 101ndash114 1966
[7] R Konkar Incremental kinematic analysis and symbolic syn-thesis of mechanisms [PhD dissertation] Stanford UniversityStanford Calif USA 1993
[8] J-S Zhao Z-J Feng and J-X Dong ldquoComputation of theconfiguration degree of freedomof a spatial parallel mechanismby using reciprocal screw theoryrdquo Mechanism and MachineTheory vol 41 no 12 pp 1486ndash1504 2006
[9] J Eddie Baker ldquoScrew system algebra applied to special linkageconfigurationsrdquoMechanism and Machine Theory vol 15 no 4pp 255ndash265 1980
[10] M TMason and J K Salisbury Robot Hands and theMechanicsof Manipulation MIT Press Cambridge Mass USA 1985
[11] Y Fei and X Zhao ldquoAn assembly process modeling and analysisfor robotic multiple peg-in-holerdquo Journal of Intelligent andRobotic Systems vol 36 no 2 pp 175ndash189 2003
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
12 Mathematical Problems in Engineering
T twist
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
T32
T31
T10
T20
T109
T1110
T114
T98
T74 T8
4
T41
T64
T65
T51
T107
Figure 23 The screw system of a hydraulic grab
R revolute jointC cylindrical joint
B0B1
B2B3
B4
B5B6
B7 B8
B9B10
B11
R31
R10
R20
1110
R114
C98
R74 R8
4
R41
R64 R5
1
R107
R
R109 C3
2
C65
Figure 24 The kinematic joint model of a hydraulic grab
According to Sections 222 and 224 the kinematic jointsbetween 119861
2and 119861
3and 119861
1and 119861
3are the cylindrical joint and
revolute joint As a result we can obtain the kinematic jointmodel See Figure 24 If the corresponding kinematic jointbetween any two parts cannot be recognized we can easilyanalyze the modelrsquos geometry constraints
6 Conclusions
This paper investigates a method for determining the kine-matic joints of assemblies restricted by arbitrary combina-tions of geometry constraints using reciprocal screw theoryWe achieve the process of recognition in terms of thebasic components of geometry constraints and kinematicjoints We describe the three basic components of geometricconstraints in terms of wrench At the same time the specificoperation of a matrix space is applied to eliminate the redun-dant components between different geometry constraintsthat act upon the same part A user of this method can auto-matically and easily convert any assembly model into a screwsystem in any CAD system Afterwards the screw system canbe easily used for validating kinematic functions analyzingover- or underconstrained assembly configurations and soon
Additional Points
(i) We provide a new method for converting any 3Dassembly model into a screw system based on recip-rocal screw theory
(ii) We describe the three basic components of geometricconstraints in terms of wrench
(iii) We eliminate the redundant components of differentgeometry constraints that act upon the same part bythe specific operation of a matrix space
(iv) We establish the maps between geometry constraintsand kinematic joints as the reciprocal product of atwist and wrench which is equal to zero
(v) We present several application examples
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
The authors would like to thank the Ministry of Scienceand Technology of the Peoplersquos Republic of China (nos2011ZX02403-005 2011CB706502 and 2013AA041301) Thiswork is also supported by the National Natural ScienceFoundation of China (nos 61370182 and 51175198)
References
[1] A H Slocum Precision Machine Design Prentice Hall NewYork NY USA 1991
[2] M Rajh S Glodez J Flasker K Gotlih and T KostanjevecldquoDesign and analysis of an fMRI compatible haptic robotrdquoRobotics and Computer-Integrated Manufacturing vol 27 no 2pp 267ndash275 2011
[3] S Awtar and G Parmar ldquoDesign of a large range XY nanoposi-tioning systemrdquo Journal of Mechanisms and Robotics vol 5 no2 Article ID 021008 2013
[4] R S Ball A Treatise on the Theory of Screws CambridgeUniversity Press Cambridge UK 1998
[5] K H Hunt Kinematic Geometry of Mechanisms ClarendonPress Oxford UK 1978
[6] K J Waldron ldquoThe constraint analysis of mechanismsrdquo Journalof Mechanisms vol 1 no 2 pp 101ndash114 1966
[7] R Konkar Incremental kinematic analysis and symbolic syn-thesis of mechanisms [PhD dissertation] Stanford UniversityStanford Calif USA 1993
[8] J-S Zhao Z-J Feng and J-X Dong ldquoComputation of theconfiguration degree of freedomof a spatial parallel mechanismby using reciprocal screw theoryrdquo Mechanism and MachineTheory vol 41 no 12 pp 1486ndash1504 2006
[9] J Eddie Baker ldquoScrew system algebra applied to special linkageconfigurationsrdquoMechanism and Machine Theory vol 15 no 4pp 255ndash265 1980
[10] M TMason and J K Salisbury Robot Hands and theMechanicsof Manipulation MIT Press Cambridge Mass USA 1985
[11] Y Fei and X Zhao ldquoAn assembly process modeling and analysisfor robotic multiple peg-in-holerdquo Journal of Intelligent andRobotic Systems vol 36 no 2 pp 175ndash189 2003
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
Mathematical Problems in Engineering 13
[12] B Lee andK Saitou ldquoThree-dimensional assembly synthesis forrobust dimensional integrity based on screw theoryrdquo Journal ofMechanical Design vol 128 no 1 pp 243ndash251 2006
[13] W Kim A P Veloso D Araujo V Vleck and F Joao ldquoAninformational framework to predict reaction of constraintsusing a reciprocally connected knee modelrdquo Computer Methodsin Biomechanics and Biomedical Engineering vol 18 no 1 pp78ndash89 2015
[14] R Konkar andM Cutkosky ldquoIncremental kinematic analysis ofmechanismsrdquo Journal of Mechanical Design Transactions of theASME vol 117 no 4 pp 589ndash596 1995
[15] J D Adams and D E Whitney ldquoApplication of screw theory toconstraint analysis ofmechanical assemblies joined by featuresrdquoJournal ofMechanicalDesign Transactions of theASME vol 123no 1 pp 26ndash32 2001
[16] J D Adams and S Gerbino ldquoApplication of screw theory tomotion analysis of assemblies of rigid partsrdquo in Proceedingsof the IEEE International Symposium on Assembly and TaskPlanning (ISATP rsquo99) pp 75ndash80 Porto Portugal July 1999
[17] J D Adams Feature-based analysis of selective limited motion inassemblies [MS thesis] Mechanical Engineering DepartmentMassachusetts Institute of Technology Cambridge Mass USA1998
[18] S Gerbino and F Arrichiello ldquoHow to investigate constraintsand motions in assemblies by Screw Theoryrdquo in Proceedingsof the CIRP International Seminar on Intelligent Computationin Manufacturing Engineering (CIRP ICME rsquo04) pp 331ndash336Sorrento Italy June-July 2004
[19] P Franciosa and S Gerbino ldquoA CAD-based methodology formotion and constraint analysis according to screw theoryrdquo inProceedings of the ASME International Mechanical EngineeringCongress and Exposition (IMECE rsquo09) Paper no IMECE2009-13159 pp 287ndash296 Lake Buena Vista Fla USA November2009
[20] M Celentano Development of a CAD-based graphic environ-ment for motion and constraint analysis according to ScrewTheory [MS thesis in Mechanical Engineering] University ofNaplesmdashFederico II Naples Italy 2007
[21] H J Su D V Dorozhkin and J M Vance ldquoA screw theoryapproach for the conceptual design of flexible joints for com-pliant mechanismsrdquo Journal of Mechanisms and Robotics vol 1no 4 Article ID 041009 pp 1ndash8 2009
[22] A Dixon and J J Shah ldquoAssembly feature tutor and recognitionalgorithms based onmating face PairsrdquoComputer-Aided Designand Applications vol 7 no 3 pp 319ndash333 2010
[23] D E Whitney R Mantripragada J D Adams and S J RheeldquoDesigning assembliesrdquo Research in Engineering Design vol 11no 4 pp 229ndash253 1999
[24] D K Smith Constraint analysis of assemblies using screw theoryand tolerance sensitivities [Master of Science] Department ofMechanical Engineering Brigham Young University 2001
[25] G Shukla and D E Whitney ldquoThe path method for analyzingmobility and constraint of mechanisms and assembliesrdquo IEEETransactions on Automation Science and Engineering vol 2 no2 pp 184ndash192 2005
[26] L Rusli A Luscher and J Schmiedeler ldquoAnalysis of constraintconfigurations in mechanical assembly via screw theoryrdquo Jour-nal of Mechanical Design vol 134 no 2 Article ID 021006 pp1ndash12 2012
[27] L Rusli A Luscher and J Schmiedeler ldquoOptimization ofconstraint location orientation and quantity in mechanical
assemblyrdquo Journal of Mechanical Design vol 135 no 7 ArticleID 071007 pp 1ndash11 2013
[28] H J Su L Zhou and Y Zhang ldquoMobility analysis and typesynthesis with screw theory from rigid body linkages tocompliant mechanismsrdquoAdvances in Mechanisms Robotics andDesign Education and Research vol 14 pp 67ndash81 2013
[29] H-J Su and H Tari ldquoOn line screw systems and their applica-tion to flexure synthesisrdquo Journal of Mechanisms and Roboticsvol 3 no 1 Article ID 011009 2011
[30] D C Anderson and T C Chang ldquoGeometric reasoning infeature-based design and process planningrdquo Computers andGraphics vol 14 no 2 pp 225ndash235 1990
[31] J Hefferon Linear Algebra 2014 httpjoshuasmcvtedulinea-ralgebra
[32] J Phillips Freedom in MachinerymdashIntroducing Screw Theoryvol 1 Cambridge University Press Cambridge UK 2007
[33] ZHuang Q Li andHDing ldquoBasics of screw theoryrdquo inTheoryof ParallelMechanisms vol 6 pp 1ndash16 Springer DordrechtTheNetherlands 2013
[34] F Dunn and I Parberry 3DMath Primer for Graphics andGameDevelopment CRC Press Boca Raton Fla USA 2nd edition2011
[35] E J HaugComputer Aided Kinematics andDynamics ofMecha-nical Systems vol 1 of Allyn and Bacon Series in EngineeringBasic Method Boston Mass USA 1989
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of