J. Software Engineering & Applications, 2010, 3: 167-175
doi:10.4236/jsea.2010.32021 Published Online February 2010 (http://www.SciRP.org/journal/jsea)
Copyright © 2010 SciRes JSEA
167
A Novel Method of Using API to Generate Liaison
Relationships from an Assembly
Arun Tom Mathew1*, C. S. P. Rao2
1Research Scholar, Department of Mechanical Engineering, National Institute of Technology, Warangal, India; 2Professor, Depart-
ment of Mechanical Engineering, National Institute of Technology, Warangal, India.
Email: aruntom123@gmail.com
Received November 17th, 2009; revised December 3rd, 2009; accepted December 20th, 2009
ABSTRACT
A mechanical assembly is a compos ition of interrelated parts. Assembly data base stores the geometric models of indi-
vidual parts, the spatial positions and orientations of the parts in the assembly, and the relationships between parts. An
assembly of parts can be represented by its liaison which has a description of its relationships between the various
parts in the assembly. The problem is to not only make the information available but also use the relevant information
for making decisions, especially determination of the assembly sequence plan. The method described in this paper ex-
tracts the feature based assembly in formation from CAD mod els of products and bu ild up liaison s to facilitate assembly
planning ap plications. The system works on the a ssumption that the d esigner explicitly defines joints and mating cond i-
tions. Further, a computer representation of mechanical assemblies in the form of liaisons is necessary in order to
automate the generation of assembly plans. A novel method of extracting the assembly information and representing
them in the form of liaisons is p resented in this paper.
Keywords: Assembly, Mate Entities, Liaisons, Solidworks API
1. Introduction
The deployment of product models for planning assem-
bly processes has received significant attention over the
years and considerable research is happening in the area
of assembly planning over the years. But assembly plan-
ning still poses a challeng e like the description of assem-
bly data and information specifically. There is much in-
terest in reducing the cost of assembly activities. Assem-
bly costs account for 10–30% of total industrial product
labor costs [1], and as much as 50% of product manu-
facturing cost [2,3]. One way of achieving this is to im-
prove assembly planning which aims to identify and
evaluate the different ways to identify and evaluate the
different ways to construct a mechanical object from its
component parts. Due to frequent changes in product
design and manufacturing methods, it is desirable to
automate and computerize the planning activity. Se-
quence generation plays an important role in designing
and planning the product assembly process. The choice
of the assembly sequence in which parts of assembly are
put together can drastically affect the efficiency of the
assembly process. Identifying part interdependencies in
assemblies and planning the process of assembly are
examples of complex decision making activities. Assem-
blies contain a very large amount of information and
complex relationships. An assembly planner is a system
based on the geometric description of an assembly model
identifies the parts that are involved in the construction
of the assembly and generates the assembly plan. The
model should provide a representation of parts and rela-
tionships such as contacts, degree of freedom among
parts of assembly. Relational models represent geometric
relations in the form of mating features between individ-
ual parts or subassemblies called liaisons. Of late, these
parts or subassemblies are being designed using CAD
programs, therefore the shape of each part and geometric
information are already available in computer database.
Since an effective description of assembly knowledge is
very necessary, this information if extracted will be
beneficial in identifying interdependencies between parts
of the assembly and represent them in the form of a liai-
son diagram. In this paper, a novel approach of using the
Automatic Programmable Interface (API) of the CAD
software to extract the information which is then used to
generate the liaison diagram which will be useful to gen-
erate assembly plans more efficiently.
The organization of the paper is as follows: Section 2
presents a literature review of assembly sequence gen-
eration; Section 3 describes the Generation of Assembly
Relationships. Section 4 summarizes a different approach
to the generation of assembly relationships using API
A Novel Method of Using API to Generate Liaison Relationships from an Assembly
168
and building of a liaison diagram; Section 5 describes the
system interface and the algorithm to extract relation-
ships and generate liaisons; Section 6 gives an example
of the approach and system and Section 7 gives conclud-
ing remarks.
2. Literature Survey
Over the years, a considerable progress has been made in
the area of assembly planning specifically in the genera-
tion of assembly sequences. In general, assembly se-
quence planning consists of assembly modeling and as-
sembly sequence generation. A lot of relevant informa-
tion regarding the assembly could not be modeled and
stored in the product while the assembly operation is
done. The efficiency of assembly planning depends on
the way the assembly information is modeled. Assembly
information modeling is the base in this research, for
generating assembly sequences. In the modeling of an
assembly, the relation between the connected compo-
nents must be established. The most commonly used
method of assembly modeling is graph based called part
mating graph [4] which represents the topological rela-
tionship between components of the assembly, where in
the nodes represent the components and the arcs establish
the relationship between the components. The mating
conditions between two components provided by the
designer are captured by a Virtual link mating graph [5].
Relational model graph includes parts, contacts and at-
tachment relationships in a model [6]. De Fazio and
Whitney called these mating graph s as liaison graphs [7].
Commercial CAD systems interpret assembly modeling
as a means of providing functionality to the designer to
easily position components with respect to each other [8].
Various detailed assembly representations have evolved
including feature based [9], kinematics based [10] and
geometry based [11]. Gottipolu and Ghosh [12,13] gen-
erated relationships by analyzing contact and mobility
constraints. Laperriere and ElMaraghy [14] generated
relationships using geometric and accessibility constrai-
nts. Generation of assembly relationships was also at-
tempted using solid models by Chang [15]. Linn and Liu
[16] described an algorithm developed to identify part
liaison relationships presented in the commercial pack-
age, I-DEAS where the program processes geometric and
topological data. Completely disassembling an assembly
component based on the geometric contact relations re-
sults in the components explosion graph. Chen used the
“contact above” concept to construct the Above Graph
and then derive the Explosion Graph [17]. In the joint-
based method [18], the assembly constraints are assigned
on the components, but not on the geometric elements of
the components. The method generates assembly models
from kinematic joint constraints by extracting feasible
joint mating features for each mating component, and
then generates the assembly configuration for a set of
joint constraints. Geometry-based representations capture
the surface mating constraints like fit, coplanar, etc to
establish the relations of precedence and feasibility [19].
The connection-semantics based assembly tree hierarchy
provides a way to consider both geometric information
and non-geometric knowledge of the assembly and ob-
tain the degree of freedom of the mating entities [20].
Product semantic information model which is made of
semantic information is structured into a three level se-
mantic abstract, from which the relevant information is
retrieved [21]. During assembly operation the designer
specifies the relative location and the orientation of the
components and surface mating constraints in order to
accomplish the desired joints. All information regarding
relationships between parts should be captured and used
during the assembly planning process. In this paper, a
novel method to extract this information between the
assembly, subassembly and components using the API of
the CAD modeling package and building the liaison ma-
trix of the assembly is proposed.
3. Generation of Assembly Relationships
SolidWorks, the commercial CAD system, is used as the
main feature-based design environment. The benefit of
using SolidWorks is, it includes an entire API with func-
tions that can be called from Visual Basic. In addition,
SolidWorks shares the same solid modeling engine as
Unigraphics and several other CAD systems like the
Pro/Engineer and Catia. In addition, these CAD systems
account for large user and application bases.
The description of the relationships among the features
of various parts is required for an assembly component.
These features can be classified into assembly features
and primal features. It is the primal features that partici-
pate in assembly constraints. The assembly module au-
tomatically determines which relationship is meant by
the user based upon the features involved in the relation-
ship and updates the degrees of freedom accordingly.
The primary mating conditions are align, mate, mate en-
tity, align offset, insert, orient etc. The align condition
requires that the axial centre lines of two parts be collin-
ear. The mate condition requires that the two mating
faces lie in the same plane with their outward normal
opposing each other. The offset condition requires that
the two faces lie in parallel planes with their outward
normal in the same direction. The relationships between
a pair of parts are specified by the user in terms of their
features and the mating conditions between them. The
individual parts in an assembly are created before the
assembly module is invoked. The assembly modeling
module requires information about the relationships be-
tween the part features. The information specified for
each mating condition includes the ID of the mating fea-
ture and the type of mating conditions.
To build a list of all the ch aracteristics of an assembly
the assembly format is developed to store all the charact-
Copyright © 2010 SciRes JSEA
A Novel Method of Using API to Generate Liaison Relationships from an Assembly 169
Figure 1. Flow of information
eristics in an assembly as its signature. The method ex-
plores in depth the assembly tree and extracts assembly
related information for each part.
a) The method retrieves the constraints used to specify
the position of the part.
b) It identifies which entities are used to constrain the
part or subassembly .
c) It identifies the parent features and part of each
geometrical entity in use.
4. Application Programming Interface (API)
Application programming Interface is an interface that
defines the ways by which an application program may
request services from libraries and/or operating systems.
An API determines the terminology and calling conven-
tions the programmer should employ to use the services.
It may include specifications for routines, data structures,
object classes and protocols used to communicate be-
tween the requesting software and the library.
An API may be Language-dependent, available only in
a given programming language, using the syntax and
elements of that language to make the API convenient to
use in this context. It can also be Language-independent,
written in a way that means it can be called from several
programming languages. An API itself is largely abstract
in that it specifies an interface and controls the behavior
of the objects specified in that interface. The software
that provides the functionality described by an API is
said to be an execution of the API. An API is typically
defined in terms of the programming language used to
build the application. The API functions used in this pa-
per are SolidWorks functions. The API functions are
essential for developing the application software. The
names of the mate features, the types, identities and the
types of the mate surfaces, the mate clearances and the
reference features etc are included in the mate informa-
tion. There are three main SolidWorks document types
namely Part Document, Assembly Document and Draw-
ing Document. Each document type has its own object
(PartDoc, DrawingDoc and AssemblyDoc) with its own
set of related functions. For example, the Assembly-
Doc::AddComponent4 method exists on the Assembly-
Doc object because adding components is specific to
assembly documents. The SolidWorks API also has
functions that are common to all document types. For
example, printing, saving, or determining the file name
associated with a document would be common opera-
tions. To expose common document-level functions, the
SolidWorks API uses the ModelDoc2 object. The Model-
Doc2 object provides direct access to the PartDoc, Dr-
awingDoc, and AssemblyDoc objects. As a general rule,
the AssemblyDoc object provides access to functions that
perform assembly operations; for example, adding new
components, adding mate conditions, hiding and explod-
ing components. The SolidWorks API also includes fun-
ctions that are common to all document types; for exam-
ple, determining the file name associated with a docu-
ment is a common operation. To expose common docu-
ment-level functions, the SolidWorks API uses the Mo-
delDoc2 object. The structure of the assembly document
is shown in Figure 2. The AssemblyDoc object is derived
from the ModelDoc2 object. Therefore, an AssemblyDoc
object has access to all of the functions on the Model-
Doc2 object. The following objects related to the mate
information like Mate, Mate Feature, Face and Surface.
The Mate Object allows access to various assembly mate
parameters. The MateEntity object enables access to ma-
ted objects and the assembly mate d e finition. The Feature
Object allows access to the feature type, name, parameter
data and to the next feature in the FeatureManager design
tree. The mate information accessed through these obje-
cts in the FeatureManager design tree.
5. Mate Information
Most CAD systems represent the assembly using constr-
Figure 2. Detailed structure of SolidWorks assembly docu-
ment
Copyright © 2010 SciRes JSEA
A Novel Method of Using API to Generate Liaison Relationships from an Assembly
Copyright © 2010 SciRes JSEA
170
aint relationships between parts. A mate feature tree is
obtained after the parts are assembled together using
SolidWorks. With the help of SolidWorks API functions
the mate constraints information is extracted by travers-
ing the mate feature tree object. The information gener-
ated is represented in an object oriented way to generate
assembly strategies especially assembly sequence plans.
the Solidworks assembly module.
The relationships between parts in the assembly are
extracted using the Assembly Mate Extraction Algorithm
and stored in a database. In add ition to the mate type and
mate entities types, the algorithm returns the follo- wing
array of doubles: [ pointX, pointY, pointZ, vectorI,
vectorJ, vectorK, halfangle, radius ]
A mate surface is a geometric surface of a part model
that has mate relationships with other part models. A part
model is composed of many surfaces, but only a few
surfaces have mate relationships with other parts. Each
surface is mated with the other using the mate entities
(Table 1) generating a constraint relation. The constraint
list includes all the geometrical constraints defined in the
CAD model; each constraint in the list includes the con-
straint type, such as concentric, against and collin ear, and
mate tolerance, etc and its corresponding code as dis-
played in the Table 2.
where
pointX is the X location of this mate entity in
the assembly model space
pointY is the Y location of this mate entity in the
assembly model space
pointZ is the Z location of this mate entity in the
assembly model space
vectorI is the i component of the assembly mate
vector
vectorJ is the j component of the assembly mate
vector
The interface consists of a provision to load Solid-
works.exe and create part models and assemble them.
The part models are saved as .sldprt file and assembly
models are saved as .sldasm file. A provision is provided
to assemble the part models directly if they already exist.
vectorK is the k component of the assembly
mate vector
halfangle is the value for the half angle
radius is the value for the radius
Table 1. List of mate entity types
5.1 Assembly Mate Extraction (AME)
Algorithm MATE TYPES CODE
swMateUnsupported 0
swMatePoint 1
swMateLine 2
swMatePlane 3
swMateCylinder 4
swMateCone 5
swMateSphere 6
swMateCircle 7
An algorithm is developed and executed using Visual
Basic for Applications in order to access feature tree of
Table 2. List of mate types
MATE TYPES CODE
swMateCOINCIDENT 0
swMateCONCENTRIC 1
swMatePERPENDICULAR 2
swMatePARALLEL 3
swMateTANGENT 4
swMateDISTANCE 5
swMateANGLE 6
swMateUNKNOWN 7
swMateSYMMETRIC 8
swMateCAMFOLLOWER 9
swMateGEAR 10
Figure 3. Process flow to generate relational database
A Novel Method of Using API to Generate Liaison Relationships from an Assembly 171
Figure 4. User interface between SolidWorks and Visual Basic
Figure 5. Flowchart for AME algorithm
Copyright © 2010 SciRes JSEA
A Novel Method of Using API to Generate Liaison Relationships from an Assembly
172
To define the mate entity, the following information
shown in Table 3 is returned based on the mate type. All
coordinate information is given in terms of the assembly
coordinate system where the mate resides.
The database contains the relations (or the Mategroups
in Solidworks) between various parts of the assembly.
There can be more than one relation between parts such
as the axis of the two parts is aligned to each other and
one face of the first part is coincident to the other. In ad-
dition to the basic relations, the database also contains
the type of relation or mate whether it is a point contact,
line contact or a plane contact as in the case. It can also
be inferred whether the component is a cylindrical com-
ponent or a prismatic component. If the program returns
a value for the radius then the component is a cylindrical
component else it is a prismatic component. The mass
properties such as the weight, volume of individual
components taking part in the process of assembly is also
considered. The above information is necessary to assist
in the selection of the base component for the assembly
and to analyze the assembly plan with respect the Design
for Assembly .
5.2 Symbolic Representation of Liaisons
Liaison graph can portray different logical and physical
contact relations among the parts of the assembly. Liai-
son graph is a two-tuples G = (P,L), where P is a set of
nodes that represent parts, and L a set of edges that rep-
resent any of certain user defined relations between parts
called liaisons. Given the liaisons graph , a decomposition
method is used to systematically generate the assembly
plans. The graph representation is difficult to process
using a computer, but can easily handle the information
in a matrix form. The table of liaisons or the liaison ma-
trix is used to represent the contact information between
Table 3. Information is returned based on the mate type
Mate Type Information Returned
swMatePoint pointX, pointY, pointZ
swMateLine
pointX, pointY, pointZ, vectorI, vectorJ,
vectorK where the point is a point on the
line and the vector represents the line
direction.
swMatePlane
pointX, pointY, pointZ, vectorI, vectorJ,
vectorK where the point is a point on the
plane and the vector represents the plane
normal.
swMateCylinder
pointX, pointY, pointZ, vectorI, vectorJ,
vectorK, halfangle where the point is a
point on the cylinder axis and the vector
represents the cylinder axis.
swMateCone
pointX, pointY, pointZ, vectorI, vectorJ,
vectorK, halfangle, radius where the point
is a point on the cone axis and the vector
represents the cone axis.
Figure 6. Flowchart for LTG algorithm
the components. If one or more relation is found to exist
between 2 pairs of components of the assembly, then that
pair will have a value one else zero.
i.e, if Li,j = 1, then relation exist between component ‘i’
and ‘j’
if Li,j = 0, then no relation exist betw een component ‘i’
and ‘j’
The Matrix Cell ij contains the value of L as 0 or 1.
It assumed that between two parts there is only one
edge representing a liaison that includes all contacts such
as collinear, coincident etc. There is no liaison joining a
vertex to itself because it is impossible to assemble a
component to itself. A component belongs to the product
it has at least a liaison with another component of the
assembly. If a component pi can be assembled to pj, the
reverse is also true. An algorithm named Liaison Table
Generator (LTG) Algorithm is developed to generate the
liaison matrix based on the relationships existing among
parts and is described in Figure 6.
Copyright © 2010 SciRes JSEA
A Novel Method of Using API to Generate Liaison Relationships from an Assembly
Copyright © 2010 SciRes JSEA
173
6. Implementation with an Example
The clamping fixture (Figure 7) taken for the present
study has 7 components namely the base, plunger, link,
knob, lever, clamp-end and pin. SolidWorks software
was used to model the assembly. The individual compo-
nents were created as separate geometric models in the
part models in the part mode and saved as “.sldprt” files.
Next, the assembly modeling mode is invoked and the
Base is taken as the support component. After specifying
the assembly constraints, the assembly was built by add-
ing the remaining components to the casting. All the
components are assembled using the mate attributes like
the Coincident, Parallel, Perpendicular, Tangent, Con-
centric, Distance and Angle. The completed assembly
model is then saved as a “.sldasm” file.
The feature tree of the clamping fixture modeled and
assembled is generated in the assembly module after
adding all the mating conditions is shown in Figure 8.
This feature tree appears on the left side of the Assembly
module. The mate relationship information stored in this
tree is extracted using the AME algorithm and stored in a
notepad file as shown in Figure 9. The AME algorithm is
implemented in a computer program coded in Visual
Basic for Applications (VBA) in the Microsoft Windows
platform. VBA is chosen, essentially for availability and
portability reasons in addition to the ability to interface
with SolidWorks. The machine used to run the program
was a notebook with a Pentium Centrino Duo 1.6 GHz
processor, with 1 GB DDR RAM.
Figure 7. Example of clamping fixture (exploded view)
Once the notepad file is generated, it acts as a database,
from which the liaisons between the various components
of the assembly can be generated. The LTG algorithm is
applied to the clamping fixture assembly and the liaison
table with relations is generated. The LTG algorithm is
implemented JAVA and resultant liaison table is shown
in Figure 10. The output of the program gives the number
of parts in the assembly, mating relationships that exist
between various parts of the assembly and finally dis-
plays the liaison diagram or table showing the relations.
7. Conclusions
The liaison information contains relationships between
parts is the basic information needed for assembly se-
quence planning. Although many modeling packages
provide information about the solid models, the informa-
tion regarding to the relationships between parts in the
assembly is not explicitly available. In this paper, a nov el
method of determining the relationship liaison diagram is
proposed. The method first extracts the assembly infor-
mation, processes it and then generates liaisons. The
mating feature matrix of two contacting parts in the as-
sembly is established. The technique is fully automated,
simplifies the process of extracting geometrical con-
straints for any given assembly considering the relation-
ships between components of a CAD model using the
Figure 8. Feature tree of the assembly created in Solid-
Works
Automated Programmable Interface of the CAD software.
The main constraint in this approach is that the designer
should model the assembly components and define the
mating conditions while assembling the product. Once
the database containing the relationships is extracted, it
can be used to generate the liaison diagram and liaison
tree. Taking the precedence relations into consideration,
feasible assembly sequences of a given assembly can be
generated.
A Novel Method of Using API to Generate Liaison Relationships from an Assembly
174
Figure 9. Snapshot of the relational database ge ne r a te d
Figure 10. Liaison table generated
Copyright © 2010 SciRes JSEA
A Novel Method of Using API to Generate Liaison Relationships from an Assembly 175
REFERENCES
[1] J. L. Nevins and D. E. Whitney, “Concurrent design of
product and processes,” McGraw-Hill, New York, 1989.
[2] U. Rembold, C. Blume, and R. Dillmann, “Computer-
integrated manufacturing technology and systems,” Mar-
cel Dekker, New York, 1985.
[3] S. S. F. Smith, “Using multiple genetic operators to re-
duce premature convergence in genetic assembly plan-
ning,” Computers in Industry, Vol. 54, Iss. 1, pp. 35–49,
May 2004.
[4] C. M. Eastman, “The design of assembly,” SAE Techni-
cal Paper Series 0148-7191/81/0223-0197, 1981.
[5] H. Ko and K. Lee, “Automatic assembly procedure gen-
eration from mating conditions,” Computer Aided Design,
Vol. 19, pp. 3–10, 1987.
[6] L. S. Homen De Mello, and A. C. Sanderson, “Represen-
tations of mechanical assembly sequences,” IEEE Trans-
actions on Robotics and Automation, Vol. 7, No. 2, pp.
211–227, 1991.
[7] D. F. Baldwin, T. E. Abell, M. C. M. Lui, T. L. D. Fazio,
and D. E. Whitney, “An integrated computer aid for gen-
erating and evaluating assembly sequences for mechani-
cal products,” IEEE International Conference on Robotics
and Automation, Vo1. 7, No. 1, pp. 78–94, 1991.
[8] K. W. Lyons, V. N. Rajan, and R. Sreerangam, “Repre-
sentations and methodologies for Assembly Modeling,”
NIST Int. Rep., Gaithersburg, MD, 1996.
[9] J. J. Shah and M. T. Rogers, “Assembly modeling as an
extension of feature based design,” Recent Trends in En-
gineering Design, Vol. 3, No. 3 & 4, pp. 218–237, 1993.
[10] C. Mascle, “Features modeling in assembly sequence and
resource planning,” In Proceedings IEEE International
Symposium on Assembly and Task Planning, Pittsburgh,
PA, pp. 232–237, 1995.
[11] R. Anantha, G. A. Kramer, and R.H. Crawford, “Assem-
bly modeling by geometric constraint satisfaction,” Co-
mputer Aided Design, Vol. 28, No. 9, pp. 707–722, 1996.
[12] R. B. Gottipolu and K. Ghosh, “An Integrated approach
to the generation of assembly sequences,” International
Journal of Computer Applications in Technology, Vol. 8,
No. 3–4, pp. 125–138, 1995.
[13] R. B. Gottipolu and K. Ghosh, “A simplified and efficient
representation for evaluation and selection of assembly
sequences,” Computers in Industry, Vol. 50, pp. 251–264,
2003.
[14] L. Laperriere and H. A. ElMaraghy, “Assembly sequen-
ces planning for simultaneous engineering applications,”
International Journal of Advanced Manufacturing Tech-
nology, Vol. 9, pp. 231–244, 1994.
[15] A. C. Lin and T. C. Chang, “3D MAPS: Three dimen-
sional mechanical assembly planning system,” Journal of
Manufacturing Systems, Vol. 12, No. 6, pp. 437–456,
1993.
[16] R. J. Linn and H. Liu, “An automatic assembly liaison
extraction method and assembly liaison model,” IIE Tr-
ansactions, Vol. 31, pp. 353–363, 1996.
[17] R. S. Chen, K. Y. Lu, and P. H. Tai, “Optimizing assem-
bly planning through a three-stage integrated approach,”
International Journal of Production Economics, Vol. 88,
pp. 243–256, 2004.
[18] J. S. Kim, K. S. Kim, J. Y. Lee, and J. H. Jeong, “Gen-
eration of assembly models from kinematic constraints,”
International Journal of Advanced Manufacturing Tech-
nology, Vol. 26, pp. 131–137, 2005.
[19] R. Sudarsan, Y. H. Han, S. Foufou, S. C. Feng, U. Roy, F.
Wang, R. D. Sriram, and K. Lyons, “A model for captur-
ing product assembly information,” Journal of Computing
and Information Science in Engineering, Vol. 6, No. 1, pp.
11–21, 2006.
[20] T. Dong, R. Tong, Ling, and J. Dong, “A Knowledge
based approach to assembly sequence planning,” Interna-
tional Journal of Advanced Manufacturing Technology,
Vol. 32, pp. 1232–1244, 2007.
[21] H. Wang, D. Xiang, G. Duan, and L. Zhang, “Assembly
planning based on semantic modelling approach,” Com-
puters in Industry, Vol. 58, pp. 227–239, 2007.
Copyright © 2010 SciRes JSEA