knowledge model as an integral way to reuse the knowledge for fixture design process

Journal of Materials Processing Technology 164–165 (2005) 1510–1518

Knowledge model as an integral way to reusethe knowledge for fixture design process

R. Huntera,∗, A. Vizan, J. Perezb, J. Rıosb

a Mechanical Engineering Department, La Frontera University, Av. Francisco Salazar 01145, Temuco, Chileb Mechanical and Manufacturing Engineering Department, Polytechnic University of Madrid,

Jose Gutierrez Abascal 2, 28006 Madrid, Spain

Abstract

The fixture design is considered a complex process that demands the knowledge of different areas, such as geometry, tolerances, dimensions,processes and manufacturing resources. Nowadays, the fixture design process is oriented to automated systems based on knowledge models.These models describe the characteristics and relationships of the physical elements together with the inference processes that allow carryingout the activity of fixture design. With the employment of the knowledge models, besides the automation, it is possible to systematize andstructure the knowledge of the fixture design process.

l, for its usei sses to use int sign processo©

K

1

rdrsekkfatH

as

earchctivi-ance

era-thein the

ng oftureesigniqueys-

ch-ems,n iscon-

ships

0d

With the use of specific methodologies, as the knowledge template, it is possible to reuse the knowledge represented in a moden a different design process. The knowledge template represents a pattern that defines the common entities and inference procehe design process. In this work, with the use of knowledge template we propose the reuse of the knowledge described in the def fixtures for machining to other types of fixtures uses like inspection, assembly or welding.2005 Elsevier B.V. All rights reserved.

eywords:Knowledge model; Knowledge template; Fixture design

. Introduction

The continuous challenge that involves the knowledge rep-esentation has oriented to many different research groups toevelop methodologies that describe stages for capture andepresentation of the knowledge in design and manufacturingystems[1–3]. This has allowed to define knowledge mod-ls as a tool that helps us to clarify the structure of intensivenowledge and information-processing tasks. In this sense, anowledge model provides a specification of the data and in-erence processes required by the system of study[4]. A firstpproach in the development of knowledge models applied

o machining fixtures design process has been proposed byunter[5].During the last decade, the use of modelling techniques has

llowed us to represent the fixture design process employed inome manufacturing operations, such as machining, assem-

∗ Corresponding author.E-mail address:[email protected] (R. Hunter).

bly and inspection, etc.[6]. Due to the complexity and thwide scope of the fixture design process, different resegroups have been focused in the analysis of specific aties of this process, such as fixture configuration, toleranalyses, stability and accessibility.

A great number of investigations has taken in considtion the way in which represent the knowledge used infixture design process. These researches are focuseddocumentation of the design parameters, the structurithe information of the fixture and the description of the fixelements used in fixture design[2,7]. On the other hand, thimplementation of the knowledge used in the fixture decan be classified regarding the artificial intelligence techn(AI) used[8,9] and on the automation level of the design stem[2].

However, whatever it is the artificial intelligence tenique used and the automation level of this type of systthe process of knowledge modelling in the fixture desigimportant for several reasons: the need to specify thecepts used in the fixture design; to establish the relation

924-0136/$ – see front matter © 2005 Elsevier B.V. All rights reserved.oi:10.1016/j.jmatprotec.2005.02.181

R. Hunter et al. / Journal of Materials Processing Technology 164–165 (2005) 1510–1518 1511

Fig. 1. The structure of the work.

among different knowledge groups; to develop knowledgebased systems (KBS), and finally, to provide a conceptualbase for reusing the knowledge.

In this sense, the entities and structures defined in a knowl-edge model for design process of machining fixtures can bepartially reused to develop new models for fixture designprocess, as the inspection or assembly fixture. The entitiesand structures reused has been defined using the method ofknowledge template[4].

The work presented is a detailed proposition of the knowl-edge model for machining fixture design and the definitionof the knowledge groups that can be reused in the inspec-tion fixture design process, using the knowledge templatemethod.Fig. 1presents a general view of the contents of thisexplanation.

2. Present state of fixture design process knowledgemodelling

The fixture concept arises from the need to establish aphysical connection between part, and tool, and part andmachine-tool. This connection should fulfil some require-ments for support the machining operation to carry out. Themainly functionality of the fixture is to support, locate andc in-t thefi ic in-f sifi-c

thatd weene def-i olo-g reuset andC

oft cture

the knowledge of a system. The first model, uses a group offorms (ICARE: Illustrations, Constraints, Activities, Rules,Entities) that allow to capture and to represent the knowledgein a semi-structured way; the second model allows to repre-sent knowledge in a structured way, using the extension ofUML [10].

The CommonKADS methodology proposes the use oftools and techniques to carry out the capture and represen-tation of the knowledge. In the case of reusing knowledge,CommonKADS proposes the use of the knowledge template:a knowledge template is a piece of a knowledge model, inwhich the data and reasoning processes can be employed inthe development of other applications[3].

3. Methodology for development a knowledge model:structural model

The methodology proposed for development of a knowl-edge model includes the realization of two stages. The firststage represents the knowledge of the objects like part geome-try, machining process, functional and detailed fixture design,and fixture resources (seeTable 1). The second stage de-scribes the inference process (design and interpretation rules)needed to obtain a first solution for the machining fixture.T l andb edgem y int ured

truc-t de-s h the

TK

K

P

M

F

F r

lamp the part to the machine tool. However, in order toerpreting correctly the needed knowledge for developxture design process, it is necessary to define the basormation related with this process according to the clasation exposed inTable 1.

All this information has been represented in modelsescribe the entities, attributes and relationships betach knowledge group in the fixture design process. The

nition of these models can be carried out using methodies that describe the activities to capture, represent and

he knowledge of a design system, for example MOKAommonKADS.The MOKA methodology is based on the definition

wo models. These models allow to capture and to stru

hese two stages allow to describe the structural modeehaviour model of the objects that compose the knowlodel for machining fixtures. This work is focused mainl

he description of the structural model for machining fixtesign process.

The proposed structural model contains a general sure of the knowledge groups related with the fixtureign process. The description of the aspects related wit

able 1nowledge group for machining fixture

nowledge group Characteristics

art geometry Geometry: holes, slots, etc.DimensionsTolerances

achining process Type of machining processMachining phase and sub phaseMachining operations

unctional and detailedfixture design process

Methodology of design

Design rulesInterpretation rulesDesign constrains

ixture resources(functional elementsand commercialelements)

Type of fixture (modular, general, odedicated)

Type of fixture elements (support,locate, clamp, etc.)Type of machine tool (verticalmilling machine, horizontal millingmachine, etc.)

1512 R. Hunter et al. / Journal of Materials Processing Technology 164–165 (2005) 1510–1518

Fig. 2. Structural model for machining fixture.

knowledge groups are presented inTable 1. Fig. 2 showsthe general structure of the knowledge groups for machiningfixture design process.

Due to the complexity of the fixture design process, thefixture design cannot be considered as an independent pro-cess with respect to the manufacturing process. In this sense,the information of the manufacturing process is directly rep-resented in the fixture design process. In a similar way, the re-sources involved in the manufacturing process have a narrowrelationship with the fixture resources, in terms of machine-tools and commercial elements of fixture. The definition ofeach knowledge groups (seeTable 1) has taken into con-sideration the attributes and necessary operations to repre-sent the knowledge relative to these knowledge groups. The

applications of these models are presented in the followingsections.

4. Knowledge template model

In this section, we describe those pieces of the knowl-edge model that can be reused in other applications using themethod of knowledge template. From a conceptual point ofview, a knowledge template describes a piece of the knowl-edge model in which the inference and the knowledge tasksare defined with the objective of reuse this knowledge inother similar applications. In this sense, it is necessary todistinguish among the analytic and the synthetic tasks. The

ks base
Fig. 3. Knowledge tas d on the structural model.


Fig. 4. Knowledge template for part entity (analytic task).

analytic tasks define the classification of the objects involvedin the fixture design process. The synthetic tasks have re-lationship with the reasoning way procedure from which afixture solution is obtained.

Using these two types of tasks, a first approach has beenestablished to define the knowledge groups that can be clas-sified under the analytic and synthetic tasks.Fig. 3shows theobjects of the structural model that describe the analytic andsynthetic tasks of the machining fixture design process.

The division of these two tasks allows to set in a first levelthe knowledge groups, that it objects and attributes that can beemployed in the development of new applications. Also, thisseparation allows us to identify those knowledge groups thatdescribe inference procedures in the design process, as thefunctional fixture design and the detailed fixture design. Thissection presents the definition of the tasks of the knowledgemodel classified under the concept of analytic and synthetictasks that can be reused in other applications.

4.1. Analytic task definition

The entities (or classes) defined under this category can beclassified regarding the level of dependence level that presentthe objects involved in the machining fixture design. Thefirst level defines those knowledge groups that are not conse-quence of the fixture design process, as geometry, dimensionsa om-p theirs -p usedi

The second level describes those entities that present asimilar structure and relationships in the fixture design pro-cess (fixture functions and commercial elements for machin-ing fixtures). In this level, can be reused only a portion ofthe structure and relationships that are not conditioned by thefixture design process.

The third level describes those entities that present a com-plete dependence to the fixture design process. In this level,cannot be reused the knowledge previously defined (struc-tures, relationships and attributes), due to dependence of theprocess developed.

4.2. Synthetic task definition

The definition of the synthetic tasks involves the identifi-cation of those objects linked with the inference procedurecarried out in the fixture design process. In this type of tasks,it cannot have a total reutilization of the knowledge, becausethe inference process carried out using a group of productionrules that depend of the type of process executed.

Under this classification, the knowledge group of func-tional design establish the functional solution of the fixturedefinition: the supporting surfaces, locating and clampingof the part. The definition of these surfaces is depending tothe manufacturing process developed. This last characteris-tic makes that the functional design possesses depend of them ng oft up isl ints,l oft p of

nd tolerances of the part. In this level, the entities that cose these knowledge groups can be totally reused intructure, relationships and attributes.Fig. 4shows an examle of the knowledge group of geometry that can be re

n other applications.

achining processes developed during the manufacturihe part. In this sense, the sharing knowledge of this groimited to the definition of the surfaces and supporting poocating, clamping for machining fixture and to selectionhe functional elements. However, the knowledge grou


Fig. 5. Knowledge template for functional elements entity.

functional elements can be reused in other applications, dueto the functional elements can be employed in multiple do-mains in the fixture design process.Fig. 5shows an exampleof the knowledge template for functional elements used inthe fixture design process.

In the detailed design occur similar situations to those ofthe functional design. In this case, the detailed design dependson the fixture functional design through a correspondence be-tween functional and commercial elements. The knowledgegroup for fixture elements can be partially reused to define anew group of fixture elements. For it, we must use the struc-ture, relationships and entities defined for the following cat-egories, base, support, locate, clamp and auxiliary elements.

5. Application of the knowledge model

In the next two sections, we present the application ofthe knowledge model for machining and inspection fixturedesign. These models taking into consideration two differentparts, because we wish express the potentiality of the use ofknowledge template.

The implementation of the structural model, discussed inSection3, is based on the instantiation of each attribute de-

fined in the knowledge groups that compose this model. Theinstantiation is defined as the assignment of a concrete valuefor a specific attribute. For it, the initial conditions are ex-posed for the application of the knowledge model, whichinclude the description of the initial geometry, final geome-try, lists of machining operations, machine-tool and fixtureresources.Table 2shows the initial information for the ap-plication of the knowledge model for machining fixture.

The execution sequence for the knowledge model hasbeen carried out having present the information provided bythe part (geometry, dimensions and tolerances), the machin-ing process and the manufacturing resources (machine-tools)suitable for the manufacturing process of the part. This infor-mation has been modelled using the unified modelling lan-guage UML, and later on, coded in the language C++. Forexample,Fig. 6 shows a piece of the code generated by theapplication of the manufacturing process group.

Fig. 6. Drilling operation code.


Table 2Initial information for fixture machining

Information Characteristics

Initial geometry

Final geometry

Machining operations Face millingSide millingDrilling

Manufacturing resources Vertical milling machineFixture resources Modular fixture elements

The definition of knowledge group of functional require-ments, presents the declaration of the functional requirementsthat should satisfy the fixture. These requirements have beendeclared using the anatomy proposed by Stevens and Alexan-der[11].

The fixture functions knowledge group, use the infor-mation described by geometry, dimensions, tolerances andmanufacturing process to define the attributes of each fix-ture function. This information allows to define the neededknowledge to execute the operations of each fixture function,such as definition of the support surfaces, definition of thelocating elements and definition of the clamping elements.In this case, we have defined five functions: guide, support,locate, clamp and centre function.

The functional fixture design is carried out applying agroup of production rules that allow to define the surfaces,points and symbolic elements for supporting, locating andclamping of the part. On the other hand, the detailed fixturedesign is obtained applying a group of interpretation rules (se-lection of supporting, locating and clamping elements) that

Fig. 7. Functional solution.

define the commercial elements employed in the detailed fix-ture design representation.

Both designs (functional and detailed) describe the lowerlevel of functional decomposition in the knowledge model.These basic information-processing units are called inferencein knowledge modelling. An inference carries out a primitivereasoning step and typically it uses knowledge contained insome knowledge base to derive new information from a spe-cific design[4].

The graphic visualization of the fixture solution has beencarried out using two representations. The first ones illustratesthe functional solution of the fixture using a group of sym-bolic elements that represent the functionalities of the fixtureelements[12]; while the second one illustrates the detailedsolution of the fixture using a group of commercial elements.Fig. 7shows the functional fixture use case solution obtainedapplying the rules for the functional design with arrangementto the initial conditions of the fixture design.

It is necessary to highlight that the detailed solution de-pends on the functional fixture solution. This dependence ispresented between the functional and commercial elements.To establish the correspondence between both types of ele-ments we have defined a group of identificators that represent

Table 3Correspondence between functional and commercial elements

Functional element Co

d

ted

IIII

d. Type of technology: reversible supportd. Surface class: machiningd. Contact surface: sphericald. Function: support element for machining fixture

mmercial elements selected type Definition dimensions

If r > 9 mm, then element not selecteIf r≤ 9 mm, then element selectedIf L> 40 mm, then element selectedIf L≤ 40 mm, then element not selecIf r1 = 10 mm, then drillØ10 mm


Fig. 8. Detailed fixture design.

the different functionalities that should have a commercialelement.Table 3describes the identifier used in the corre-spondence between the functional and commercial elements.Additionally, this table present an example for the definitionof the type and dimension of the support elements used in thedetailed fixture solution.

Finally, using the interpretation rules and the rules of def-inition of dimensions of the commercial elements, the finalconfiguration of the fixture is obtained.Fig. 8 illustrates thedetailed fixture design.

It is important to highlight that the structural model hasbeen conceived thinking of defining a way to presenting thenecessary functions for the execution of certain activitiesrelatives to the fixture design, and that logically theyhave to drive to a compatible solution with the functionalrequirements and fixture constraints.

5.1. Knowledge template model for inspection fixturedesign

In Section4, we have been defined the knowledge groupsthat can be reused in other applications of the fixture design,as it is the inspection fixture design. The result of the appli-cation of the knowledge template is presented inFig. 9. Thisfigure presents the general structure of the inspection fixturedesign.

The basic difference between these two knowledge mod-els (machining and inspection) proposed in this article re-sides in the description of the type of process needed todevelop the fixture design. For it, the following knowledgegroups have been changed: manufacturing process by in-spection operation planning, and commercial elements formachining fixtures by commercial elements for inspectionfixtures.

Using the knowledge template proposed inFig. 9, theprocess of implementation of the knowledge model forthe inspection fixture, begins with the declaration of therequirements of the fixture inspection (orientation, sup-port and clamp), the information provided by geometricrepresentation of the piece allow to define the dimen-sions, the dimensional tolerances, and the geometric tol-erances. Other information is provided by the elementsthat should be inspected, the inspection operations and re-ss dgeg

thep signf fini-

e mode
Fig. 9. Knowledge templat
ources used in the inspection process.Table 4illustrates theummary of the information contained in these knowleroups.

Taking into consideration the information described inrevious table, has been defined the functional fixture de

or inspection applying a group of production rules (de

l for inspection fixture design.


Table 4Description of knowledge groups for inspection fixture

Knowledge groups Description

Geometry

Dimension D1: height and diameter hole: 12 mm and10 mmD2: height and diameter hole: 12 mm and10 mmD3: height and diameter hole: 12 mm and10 mm

Dimensions tolerances T1: hole tolerance: +0.03/−0.05T2: hole tolerance: +0.02/−0.06T3: hole tolerance: +0.02/−0.06

Geometric tolerances T1: cylindrical tolerance: 0.01 mmT2: cylindrical tolerance: 0.03 mmT3: cylindrical tolerance: 0.03 mm

Inspection elements

Inspection operations Cylinder measureD1: three point evenly dis-tributes under hole surfaceSurface measureD1: four point evenly dis-tributes under the plane surface ofD1

Cylinder measureD2: three point evenly dis-tributes under hole surfaceSurface measureD2: four point evenly dis-tributes under the plane surface ofD2

Cylinder measureD3: three point evenly dis-tributes under hole surfaceSurface measureD3: four point evenly dis-tributes under the plane surface ofD2

Inspection resources Vertical coordinate measure machiningCalibration elements RenishawStar style: five pointModular elements for inspection fixture

tion of support, locate and clamp surfaces rules, etc.). Theserules define the support and clamping surfaces; support andclamping points; and the selection of the functional elementsfor support and clamp.Fig. 10 illustrates the result of thefunctional fixture design for inspection.

In the case of detailed fixture design for inspection,a group of interpretation rules is used (selection of sup-port, locate and clamp elements) to define the correspond-ing commercial elements.Fig. 11 illustrates the detailedfixture design for inspection using a group of modularelements.

Fig. 10. Functional design of inspection fixture.

Fig. 11. Detailed design for inspection fixture.

6. Conclusion

An approach for the partial reusing of a knowledge modelin new applications relative to the fixture design process, us-ing the concept knowledge template has been presented. Thisapproach provides a way to reusing the knowledge definedin the different knowledge groups that integrate a model forfixture design. The use of knowledge template provides areduction time spend to development a knowledge model.This saving is traduced in the reuse of entities, attributes andrelationships defined in previous knowledge models.

The concept of knowledge template has demonstrated tobe very useful in the generation of new models that share anduse knowledge for developing similar design processes.

References

[1] J. Barreiro, J. Labarga, A. Vizan, J. Rios, Information model forthe integration of inspection activity in a concurrent engineeringframework, Int. J. Mach. Tools Manuf. 43 (2003) 797–809.

[2] N. Bugtai, R.I.M. Young, Information models in an integrated fixturedecision support tool, J. Mater. Process. Technol. 76 (1998) 29–35.


[3] C. Grabowick, R. Knosala, The method of knowledge representationfor a CAPP system, J. Mater. Process. Technol. 5805 (2002) 1–9.

[4] G. Schreiber, et al., Knowledge engineering management, in: TheCommonKADS Methodology, MIT Press, 1999.

[5] R. Hunter, Definicion e integracion de requisitos y funciones defabricacion en un sistema basado en el conocimiento (KBE) apli-cado al diseno de utillajes de mecanizado, PhD Thesis, UniversidadPolitecnica de Madrid, 2004.

[6] A. Nee, A. Senthil, K. Whybrew, Advanced Fixture Design for FMS,Springer-Verlag, 1995.

[7] J. Kakish, Toward the design and development of a knowledge baseduniversal nodular jigs and fixture design, J. Intell. Manuf. 11 (2000)381–401.

[8] W. Li, P. Li, Y. Rong, Case-based agile fixture design, J. Mater.Process. Technol. 128 (2002) 7–18.

[9] A. Senthil, V. Subramaniam, K. Seow, Conceptual design of fixtureusing genetic algorithms, T. Int. J. Adv. Manuf. Technol. 15 (1999)79–84.

[10] M. Stokes, et al. (Eds.), Managing Engineering Knowledge:MOKA Methodology for Knowledge-based Engineering Applica-tions, ASME Press, 2001.

[11] R. Stevens, I. Alexander, Writing Better Requirements, Addison-Wesley, 2002.

[12] AFNOR—NF E 04-013 (Dessins Techniques—Dessinsd’operations—Symbolisation des prises de pieces), AFNOR,1985.

knowledge model as an integral way to reuse the knowledge for fixture design process

Documents