Birendra Prasad (Editor)

CAD/CAM
Robotics and Factories
of the Future
Volume II: Automation of Design,
Analysis and Manufacturing
3rd International Conference on CAD/CAM
Robotics and Factories of the Future
(CARS and FOF'88) Proceedings

With 97 Figures

Springer Verlag Berlin Heidelberg NewYork
London Paris Tokyo Hong Kong


Editorial Board
Chainnan

Birendra Prasad
Senior Engineering Staff
Artificial Intelligence Services
Technical System Development
Electronic Data Systems
General Motors
Post Box 7019
Troy, MI48007-7019

Editors:

S. N. Dwivedi
K. B. Irani
Associate Editors:
M. Botkin
E.J. Carl
R T. Haftka
G.Jones
V. Kumar
P.Liu
M.D.OlifT
P. C. Shrivastava
S. Srivastava

T.H.Caiati
R V. Grandhi
J. Horgan
R T. Johnson
A. Kusiak
E. Nikolaidis
RD.Oza
D. Sriram
H.-P.Wang

ISBN 978-3-642-52325-0

ISBN 978-3-642-52323-6 (eBook)

DOI 10.1007/978-3-642-52323-6
This work is subject to copyright.AII rights are reserved, whetherthe whole orpart ofthe material is
concerned, specifically the rights of translation, reprinting, re-use of illustrations, broadcasting,

reproduction on microfilms or in other ways, and storage in data banks. Duplication of this
publication or parts thereofis only permitted under the provisions of the German Copyright Law
of September 9, 1965, in its version of June 24, 1985, and a copyright fee must always be paid.
Violations fall under the prosecution act of the German Copyright Law.

© Springer-Verlag Berlin, Heidelberg 1989
Softcover reprint of the hardcover 1st edition 1989
The use of registered names, trademarks, etc. in this publication does not imply, even in the
absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.

216113020543210 - Printed on acid-free paper


Conference Objective
Improving cost competitiveness and remaining abreast in high technology are some of the challenges that are faced by a developing
enterprise in the modern times.
In this context, the roles of engineering, manufacturing and plant automation are becoming important
factors to enhance productivity and profitability, and thereby increase market share and product quality. The commuter automobile,
actively controlled car, the U.S. space station, the unmanned platform, and commercial space ventures are all real life examples of a
few explorations now being undertaken on earth and space - requiring a
greater dependence by people on machines. Complete shop floor automation - a "lights out" plant may be unrealistic to many but automating
and integrating the engineering and manufacturing process, where it
makes sense from a cost/benefit stand point, are certainly viable
u~dertakings.

Hence, the objective of the Third International Conference on
CAD/CAM, Robotics and Factories of the Future (FOF) is to bring together researchers and practitioners from government, industries and
academia interested in the multi-disciplinary and inter-organizational
productivity aspects of advanced manufacturing systems utilizing

CAD/CAM, CAE, ClM, Parametric Technology, AI, Robotics, AGV technology, etc..
It also addresses productivity enhancement issues of other
hybrid automated systems that combine machine skills and human intelligence in both manufacturing (aerospace, automotive, civil, electrical, mechanical, industrial, computer, chemical, etc.) and in nonmanufacturing (such as forestry, mining, service and leisure, process
industry, medicine and rehabilitation) areas of application.
such an
exchange is expected to significantly contribute to a better understanding of the available technology, its potential opportunities and
challenges, and how it can be exploited to foster the changing needs
of the industries and the marketplace.


Conference Scope
The conference included the following areas of active research and
application:
CAED:

CAD, CAT, FEM, Kinematics, Dynamics, Simulation,
Computer Graphics, Off-line Programming

Analysis,

CIM:

CAD/CAM, CNC/DNC, FMS, AGV, Integration of CNC, Interactions
between
Robotics, Control, Vision, AI, Machine Intelligence, and other
Automation Equipments, and communications Standards

Design/Build Automation: Parametric Programming, Design, Sensitivity,
optimization, Variational Geometry, Generic Modeling,
Identification, Design Automation, Value Engineering" Art

to Part, Quality, cost & Producibility
Knowledge Automation: Artificial Intelligence, Expert Systems
Robotics: Mechanical Design, Control, Trajectory Planning, Mobility,
End Effecters, Maintenance, Sensory Devices, Work Cells,
Applications, Testing and Standardization
Factory of the Future: Planning of Automation, Management, Organization, Accounting, Plant Design, Informative Systems, Productivity Issues, Socioeconomic Issues, Education, Seminars and
Training.

Conference Theme
The theme of the 3rd International Conference was:

C4 (CAD/CAM/CAE/CIM) Integration, Robotics, and Factory
Automation for improved productivity and cost containment.


Conference Organization
SPONSOR
International Society for Productivity Enhancement
(ISPE), USA
ENDORSING SPONSORS
Association for computing Machinery (ACM), USA
National Science Foundation (NSF), USA
Society of Automotive Engineers (SAE), USA
Automotive Industry Action Group (AIAG), USA
Robotic Industries Association (RIA), USA
Electronic Data Systems (EDS), General
Motors Corporation, USA
The International Association of Vehicle
Design (IAVD), UK
International Society for Computational

Methods in Engineering (ISCME), UK
American Institute of Aeronautics and
Astronautics (MI) (AlAA), USA
American Society of Civil Engineers (MI) (ASCE), USA
Center for Robotics and Advanced Automation
(CRAA), Oakland university, USA
American Society of Engineering Education (ASEE), USA
Engineering Economics Division (EED-ASEE), USA
Japan Technology Transfer Association (JTTAS)
American society of Engineers from India (ASEI), USA
Michigan society of Architects (MSA), USA
CAD/CIM Alert, Massachusetts, USA
Automation and Robotics Research Institute,
university of Texas at Arlington, TX, USA


Committee Chairpersons
Conference General Chairperson: Dr. Siren Prasad,
Electronic Data Systems, GM, USA
Program chairpersons: Dr. Sur en N. Dwivedi,
UWV, USA ; William R. Tanner,
Cresap Manufacturing Cons., USA
Doug Owen, EDS, USA
Technical Chairpersons: Rakesh Mahajan,
Deneb Robotics, Inc., USA;
Dr. Jean M. Mallan, EDS, USA
International Chairpersons: Dr. Ario·Romiti,
Politechnico di Torino, ITALY ;
Dr. Marcel Staroswiecki, Universite De Lille,
FRANCE ; Dr. Jon Trevelyan,

Computational Mechanics Institute, UK
Panel Session Chairpersons: Dr. Frank Bliss,
EDS, USA ; Dr. Subra Ganesan,
Oakland University, USA
Workshops Chairperson: Dr. Pradeep K. Khosla,
carnegie Mellon University, USA
Video/Tech Display Chairperson: Dr. Addagatla
J. G. Babu, University of South Florida, USA
Student Session Chairperson: Dr. Hamid R. Parsaei,
University of Louisville, USA
Exhibits Chairpersons: Jon Keith Parmentier,
Tektronix Inc., USA; Forrest D.
Brummett, GM, USA
Receptions Chairperson: Umesh B. Rohatgi, Charles
S. Davis Associates Inc., USA;
Dr. Bhagwan D. Dashairya, Inventors Council of
Michigan, Ann Arbor, MI, USA
Administration Chairperson: Dr. Prakash C.
Shrivastava, GM, USA
Conference Directory: Dr. Yogi Anand, Consultant,
Rochester Hills, MI, USA


Committees'Roster

PROGRAM COMMITIEE

Dr. Hsin·VI LoI, North Carolina Ag. & Toch.
State Unlve"lty, USA


Dr, Kang G. Shin, University of Michigan,
USA

Dr. Paiya Uu, Siemens Corp., USA

Anthony R Skomra, AutomaUon
Technology Products, USA

Dr. Sudhlr Aggarwal, Bell
Communications Research, USA

Dr. Surosh M. Mangrulkar, Ford Motor
Co., USA

Dr, John S. Bara., University 01 Maryland,
USA

Dwight Morgan, GMF RoboUcs, USA

Dr. Marc Becque~ Unlverslt.' Ubro De
Bruulle., BELGIUM
Thomas H. CaiaU, EDS, USA
James P. Cal., GM, USA

Dr. Michael Mulder, UniverSity 01 Portland,
USA
Yasuo Nagai, institute 01 New Generation
Computer Technology, JAPAN

Micha.1 F. Carter, GM, USA


Dr. Shlgeo Nakagakl, Toshiba Fuchu
Works, JAPAN

Dr, M. Colsaltis, UGRA CEN.FAR. FRANCE

Dr. Los:lo Nemes, CSIRO, AUSTRAUA

J. P. Crestin, DDREET, FRANCE

Dr. Elstratios Nikolaldls, VPI & Stat.
University, USA

Kenneth A. Crow, Western Data Systems,
USA

Dr. A. F. D'Sou:a, liT, USA
Catherln. Foregon, DDREET, FRANCE

Dr. Michael D. Oliff, University 01 South
CarOlina, USA
Loster Ottinger, THP Inc., USA

Dr. William M. Spurgeon, University 01
Michigan Dearborn, USA
Dr. Raj S. Sodhi, New Ja... y Institute 01
Technology, USA
Rick Stapp, Auto Simulations Inc, USA

Dr. Rajan Surl, University 01 WisconSin,

USA
Dr. Bharat Thacker, Universal Computer
Applications, USA
Dr. Joe Torok, Rochester Institute 01
TeChnology, USA
Michael J. Tracy, Smith Hinchman &
Gryfls Associates Inc, USA

Dr. H. S. T:ou, University 01 Kentucky, USA

Michael J. Frelling, Tektronix Lob., USA

Dr. Rejshekar D. 0:0, University 01
Aorida, USA

Don H. Turner, Arthur Young & Co., USA

Dr. Ramana V. Grandhl, Wright Stat.
University, USA

Dr. Sudhakar Paldy, Rochester Institut. 01
Technology, USA

Dr. Hsu·Pin Ben Wang, University 01
Buffalo, USA

Jack B. Harrison, Th. Hands-on JIT inc.,
USA

Prof. V. M. Ponomaryov, Academy 01

Sciences, USSR

Dr. Peter Ward, SDRC Englneerlng
Services Ltd., UK

M.C. Portmann, INRIA-Lorralne, FRANCE

Dr. Ronald L Websler, Morton Thlokol
Inc., USA

Dr. Gary P. Herring, US Postal Service,
USA
Dr. Jack Horgan, Ari•• Technology, USA
Dr. Ming C. Huang, EDS, USA

J.M. Proth, INRIA·Lorraln., FRANCE

Prof. J. G. Postalro, Unlveralte' De Ulle,
FRANCE

Dr. Ichlro Inou., NEC Corp., JAPAN

Dr. Tim Pryor, Diffracto, Ud., CANADA

William B. Johnson, Rockwell
International, USA

Prof. J. Ragot, Unlverslte' De Nancy,
FRANCE


Dr. Senlay Joshi, Pennsylvania Stat.
University, USA

Arthur D. Rogers, Integrated Automation
Corp., USA

Richard B. Katnlk, GM, USA

Joseph D. Romano, A. T. Koamey, USA

Dr. Rakesh K. Kapanla. VPI & Stat.
University, USA

Dr. Anll Salgal, Tulia University, USA

Gerald A. Kasten, NlA T.ch Corp., USA

Dr. Sunil Selgal, Worcester Polylochnlc
institute, USA

Prof. F. Kimura, University 01 Tokyo,
JAPAN

Harshad Shah, Eagl. Technology Inc.,
USA

Dr. Andrew Kusiak, University of
Manitoba, CNlADA

Dr. Rom P. Sharma, Westem Michigan

University, USA

Donald A. Vincent, RIA. USA

Dr. Tony Woo, National Science
Foundation, USA
Dr. Wei Uang Xu, Beijing Instltule 01

Aeronautics & Astronautics. PRC

Dr. Y. F. Zheng, Clemson University, USA
Dr. William J. Zd.bllck, Molcut Res.
Associates, USA
Dr. John S. Zuk, Brooklyn PolytechniC
University, USA


x
ADVISORY COMMITTEE
Tony Affuso, EDS, USA

Dr. Carloa A. er.bbla, W....x In.tltut. of
Technology, UK
Dr. M. A. I""gham, The Open Unlv.rsity,
UK
Dr. G. G. Dodd, GM Research
laboratori ••, USA
Rudi Gem, EDS, USA
W. C. Hamann, Ford Motor Company, USA


Dr. Pierro Har.n, Intelligence Loglcl.lle,
FRANCE
Russ.1I F. Henke, Automation T.chnology
Products, USA
Prof. K. Iwata, Kobe University, Japan

Dr. Henry H. Fong, MARC Analysis
Research Corporation, USA

Dr. Muke.h Gandhi, Michigan State
University, USA

D. Galara, EDF/DEFVSEP, FRANCE

Dr. Kalyan Ghosh, Eoole Polytechnlque,
CANADA

Dr. Dan G. Gam., Davidson Research, USA
John E. Gotz, Fru.hauf Corporation, USA

Dr. Abid Ghuman, Chrysl.r Molors, USA
GI.nn R. Gramling, Hewl.tt.Packard
Company, USA
Jam.s D. Hock, GM, USA

Dr. ANn Jain, BP Am.rlca Inc., USA
Dr. Hiroshi .Kawanishl, NEC Corporation,
JAPAN

Dr. R. T. Hallka, VPI and Stale University,

USA
Dr. C. B. Jennings, South Bank
Polytechnic, UK
Dr. R. T, Johnson, University of Missouri,
USA
Dr. Steven H. Kim, Massachusetts Institute
of Technology, USA

Dr. Kant Kothawala, EMRC, USA

Dr. Robert D. Loghcer, Massachusetts
Institute ofTechnology, USA

Dr. Marshall M. Uh, National Science
Foundation, USA

Dr. Virendra Kumar, General E1.ctric
Company, USA

Dr. Arch W. Naylor, University of Michigan,
USA

Dr. M. E. Merchant, Metcut Research
Associates, Inc., USA

Dr. Peter A. Marks, Automation
Technology Products, USA

Dr. Wall.r D. Pilk.y, Unlv.rslty of Virginia,
USA


Dr. Sanjay Mittal, Xerox. USA

Dr. Paul G. Ranky, Unlv.rslty of Surr.y, UK

Dr. Munlr M. Kamal, GM Research
laboratorl.s, USA

Dr. Howard Morall, National ScI.nce
Foundation, USA
Georg. E. Munson, University of
Califomia Santa Barbara, USA

Dr. Jay Nathan, Unlv.rslty of Scranton,
USA

Wallace M. Murray, Morton Thlokollnc.,
USA
laJos Imre Nagy, Ford Molor Company,
USA
Ram G. Narula, Bachtel Corporation, USA

Dr. G. J. OIling, Chrysl.r Motor., USA

Dr. Frank Plonka, Chrysler Motors, USA

Dr. A. P.t.rs, rNWA. FRG

Donald L Smith, Ford Motor Company,
USA


Kar.n L Resmussen, GM, USA
Robert B. Schwar1l, Fru.hauf Corp., USA
Oonnls E. Wisnosky, Wizdom Systems
Inc., USA

Dr. Gerald A. Thompson, Hughe. Aircraft
Co., USA

UNIVERSITY COMMITTEE

Dr. S. S. Rao, Purdu. University, USA
Dr. Eugene I. Rivin, Wayne Stat.
University, USA
Dr. Rak.sh Sagar, South Bank
Polyt.chnlc, UK
Dr. Har.sh C. Shah, Stanford University,
USA
Dr. Nanua Singh, Unlv.rsity of Windsor,
CANADA
Dr. Ouwuru Sriram, Massachusetts
Institut. of Technology, USA
Dr. K.s. Tararnan, lawr.nce Institut. of
T.chnology, USA

INDUSTRIAL COMMITTEE

Dr. David Ardayfoo, Wayne Stata
University, USA


W. Robert Bu.II, Ford Motor Comparry,
USA

Dr. V.S. Chadd&, University of Ootrolt, USA

Dr. wayne W. Walt.r, Rochest.r Institute
of Technology, USA

Edward J. Carl, m, USA

Dr. John B. Cheatham, Jr~ Ace
University, USA

Dr. Marek B. Zaremba, Universlt.' Cu
auebec.' HuD, CANADA

Dr. Robert G. Cubensky, Chrysl.r Motors,
USA

Dr. Rollin C. Oix, Ilinolsinstitut. of
Technology, USA

Dr. Macl.J Zgorz.lskl, GMI Eng. &
Management Inst., USA

Etim Sam Ekong, Unlsys Corp., USA

Dr. Joe G. Elsley, Unlvarsity of Michigan,
USA


EdwIn J. Fablszak, Jr., MSC/CAD COMP
Inc., USA

Dr. Nar.n R. Vir., Howard University, USA


XI
INTERNATIONAL ORGANIZING
COMMIITEE
BELGIUM
M. Bec:quet (Bru••els)

R. Gobin (leINen)

J. Peter. (leINen)
R. SnCO)" (LelNOn)

PEOPLES REPUBLIC OF CHINA
a.lan Zhang (Beijing)

POlAND

A. Morecld (Warsaw)
REPUBLIC OF CHINA

H. Van Brussel (Lewen)

Shul-Shong Lu (T.Ipei)

P Vanherck (lewen)


ROMmIA

BRAZIL

Voleu N. Chloroanu (Slghetu Marmatiei)

Maria Emilia Camargo (Santa Marla)

Mircea Ivaneseu (Cralova)

Edger Pereira (porto Alegre)

SPAIN

CANADA

R. Core. (Madrid)

B. Manas Das (Calgary)

THAlLmD

Mark B. Zaremba (Hull)

R. Sadananda (Bangkok)

DENMARK

UNITED KINGDOM


Finn Fabricius (lyngby)

John Billingsley (portsmouth)

FRANCE

Carlos A. Brebbla (Southampton)

BourJault Alain (Besancon)

M. A. Dorgham (Milton Keynes)

Phlilipe Pract (Besancon)

David G. Hughes (Plymouth)

Marcel Staroswleckl (Vilienewe-D'Ascq)

David Paul Stoten (BrIstol)

Oaude Viebet (Evry)
INDIA
C. Amarnath (Bombay)
P.

C. Pandey (Roorkee)

Rakesh Sagar (Delhi)
V. Singh (IIaranasi)

N. Viswanadham (Bangalore)
ITALY
Ario RomlU (Torino)

JAPm
Yoshiakllchlkawa (HitachI)
T_ Yamashita (Tobata)

NETHERLmDS

J. A. M. Willenborg (Utrecht)


Letter from the President, ISPE

Dear Participants and Guests;
1987-1988 was the best and the most fruitful year in the history of ISPE. With your
continued suppon and co-operation, ISPE has seen considerable growth and popularity.
You will agree that our focus is very much mainstream and activities are clearly aimed
towards bringing all the peninent issues found in technological, business, socio-economic,
and organizational horizons for discussion and resolution.
After successful sponsorship of three conferences in the USA, ISPE is now
sponsoring the Fourth Intemational Conference at 1.1.T. Delhi, India during December
19-22, 1989. I hope, with your active participation and suppon, the fourth conference is
bound to be a success.
We would like you to know that your continued technical input, written to share
constructive ideas and innovative development strategies have been our backbone. your
involvement has been the key to our success but our continued growth requires more
efforts. The society is constantly in need of creative ideas and experienced hands. So far,
we have been carrying out the responsibilities with sustained contributions from a limited

number of members. Now, we are requesting your cooperation and help.
With this letter, I extend a personal invitation to each of you to come up with fresh
ideas and new ways of thinking - a pannership that can strengthen ISPE technical and
financial foundations so that we could be more aggressive in promoting yours interests and
improving the quality of life to which ISPE stands.
With good wishes,

Dr. Suren N. Dwivedi .
West Virginia University
Morgantown, West Virginia
USA


ISPE Conference Mission
ISPE was founded in 1984 with the goal to accelerate the international exchange of ideas and scientific knowledge with absolutely no
barriers of disciplines or fields of technological applications.

The

main objective of ISPE is to foster cross-fertilization of technology,
strategy and 4M resources (manpower, machine, money and management) to
enhance productivity - to increase profitability and competitiveness,
and thereby improve the quality of life on land, sea, air and space.
One of the aims of the society is to provide opportunities for contact
between members through national and international conferences, seminars, training courses and workshops.

The Society also aims to create

a channel of communication between academic researchers, entrepreneurs, industrial users and corporate managers.
ISPE embraces both the traditional and non-traditional fields of

engineering, manufacturing and plant automation, all areas of computer
technologies, strategic planning, business and control.

Equal empha-

sis is being placed on the cross-fertilization of emerging technologies and effective utilization of the above 4M resources.


Acknowledgements
The Third International Conference on CAD/CAM, Robotics and Factories of the Future (CARS & FOF '88) was hosted by the International
society for Productivity Enhancement (ISPE) and was endorsed by more
than 18 societies, associations and international organizations. The
conference was held in southfield, Michigan at Southfield Hilton Hotel
during August 14-17, 1988. Over 450 people from 12 foreign countries
attended.
People from industries, universities, and government were
all represented.
Over 250 technical presentations organized into 11
forums (panels), 61 specialty sessions, 3 plenary sessions and 4
workshops were conducted during the four days program. Six major
symposia were concurrently held.
I wish to acknowledge with many thanks the contributions of all the
authors who presented their work at the conference and submitted the
manuscripts for pUblication. It is also my pleasure to acknowledge
the role of keynote, banquet, and plenary sessions speakers whose
contributions added greatly to the success of the conference. My
sincere thanks to all sessions chairmen and sessions organizers. I
believe that the series of the International Conferences on CAD/CAM,
Robotics and Factories of the Future which emphasizes on crossfertilization of technology, strategy and 4M resources (manpower,
machine, money and management) will have a major impact on the correct

use of productivity means - to increase profitability and competitiveness, and thereby improve the quality of life on land, sea, air and
space.
I acknowledge with gratitude the help and the guidance received from
the various organizing committees. I also wish to extend my gratitude
to the sponsoring organizations. Grateful appreciations are due to
stUdent volunteers from Oakland University, Wayne State University,
University of Detroit and University of Michigan for their enthusiastic participation and help in organizing this conference. Thanks are
also due to all my colleagues, friends, and family members who extended their help in organizing this conference and making it a success.
In particular, I acknowledge the help and cooperation extended by
Electronic Data Systems (EDS) without which this would not have been
possible.
I would like to appreciate the excellent work done by SpringerVerlag in publishing this proceedings.
B. Prasad
Conference Chairman and Chief Editor


Conference Proceedings
The papers included in this volume were presented at the Third
International Conference on CAD/CAM, Robotics and Factories of the
Future (CARS & FOF '88) held in southfield, Michigan, USA during
August 14-17, 1988.
CARS & FOF '88 featured 11 panels, 6 symposia and 4 workshops. The
symposia covered six specific themes of productivity tracks (representing foundations of connectivity) in "The Look of the Future in
Automated Factories" •. Under each symposium, several key sessions were
planned, focussing both on the opportunities and challenges of new or
emerging technologies and the applications. Over 250 papers from over
12 countries covering a wide spectrum of topics were presented in the
following six symposia:
Symposium
Symposium


I: CAED - Product & Process Design
II: CIM & Manufacturing Automation

Symposium III: Design/Build Automation
Symposium
Symposium
Symposium

IV: AI & Knowledge Automation
V: Robotics & Machine Automation
VI: Plant Automation & FOF

The conference proceedings are published in three bound volumes by
Springer-Verlag. The three Volumes are:
Volume
Volume

I: Integration of Design, Analysis and Manufacturing
II: Automation.of Design, Analysis and Manufacturing

Volume III: Robotics and Plant Automation
Volume I includes papers from Symposia I and II, Volume II includes
papers from symposia III and IV, and Volume III includes papers from
Symposia V and VI. The papers presented in the panel sessions and
plenary sessions are distributed to the Volumes based upon the subject
matters. The complete list of papers for all volumes are included at
the end of each Volume.



Preface
This volume is about automation - automation in design, automation
in manufacturing, and automation in production. Automation is essential for increased productivity of quality products at reduced costs.
That even partial or piecemeal automation of a production facility can
deliver dramatic improvements in productivity has been amply demonstrated in many a real-life situation. Hence, currently, great efforts are being devoted to research and development of general as well
special methodologies of and tools for automation. This volume reports on some of these methodologies and tools.
In general terms, methodologies for automation can be divided into
two groups. There are situations where a process, whether open-loop
or closed-loop, is fairly clearly understood. In such a situation, it
is possible to create a mathematical model and to prescribe a mathematical procedure to optimize the output. If such mathematical models
and procedures are computationally tractable, we call the corresponding automation - algorithmic or parametric programming.
There is, however, a second set of situations which include processes that are not well understood and the available mathematical models
are only approximate and discrete. While there are others for which
mathematical procedures are so complex and disjoint that they are
computationally intractable.
These are the situations for which
heuristics are quite suitable for automation. We choose to call such
automation, knowledge-based automation or heuristic programming.
The papers in this volume range from highly theoretical to specialized treatment of very practical problems. The techniques borrowed
from artificial intelligence have to do with the use of knowledge
bases, the art of reasoning, and the application of the concept of
expert systems. These papers, more or less, divide themselves into
the following four chapters:
Chapter
I: Computer-Aided Design
Chapter II: Automation in Manufacturing
Chapter III: Applications of Artificial Intelligence
Chapter IV: Expert Systems
The works reported in the first two chapters of this volume deal
with algorithmic/parametric programming. The rest of the volume deals

with heuristic programming.


Contents
CHAPTER I: Computer-Aided Design
Introduction ....•.•••......••...•••..••••..••.••••••.•••••.•.•
I.l. Shape Optimization ....•....•..•................•••.••••••...•
A Geometry-Based 2-Dimensional Shape Optimization
Methodology and a Software System with Applications
V. Kumar, M.D. German, and S.-J. Lee .•••....••.•••.••...•..
optimum Design of continuum Structures with SHAPE
E. Atrek, and R. Kodali .••...•••........••..••....••...•..•
The Velocity Field Matrix in Shape optimal Design
A.D. Belegundu, and S.D. Rajan •.••••..•••••••••..•...••..••
Implementation Issues in Variational Geometry and
Constraint Management
J.C.H. Chung, J.W. Klahs, R.L. Cook, and T. Sluiter ••••••.•
I.2. Probabilistic Design Optimization .•.••...••....•••..••.••••••
Probabilistic Vibration Analysis of Nearly Periodic Structures
K.F. Studebaker, and E. Nikolaidis ••..........••.....••....
Experience Gained From First-Order Reliability Methods (FORM)
in Structural Analyses
D. Diamantidis ................••...••..•••.•.•..••..•...•..
Reliability Analysis of Layered Cylindrical Structures
under Combined Mechanical and Thermal Loads
S. Thangjitham, and R.A. Heller ....•••...••....•.....•.....
Design Reliability optimization Using Probabilistic Design
Approach and Taguchi Methods
M: Afzal,. and K.C: Ka~ur •..••.......•..••..•.•..••.•...•...
I.3. Opt~mum Des~gn Appl~cat~ons ••..•.....••••..•.••.•••.•.•••••••

optimization of Frame structures with Thin Walled
sections of Generic Shape
S. Belsare, M. Haririan, and J.K. Paeng ....•••....••••.•.•.
Optimal Design of Box Beams with Coupled Bending and Torsion
Using Multiple Frequency Constraints
R.V. Grandhi, and J.K. Moradmand •.•••..•••••••••••.••..••••
Experiences on Analysis and Optimal Design of Pyramidal
Truss Panels
M.A. Wiseman, J.W. HOu, and T.A. Houlihan .•.•••••••••.••..•
A Computational Procedure for Automated Flutter Analysis
D.V. Murthy, and K.R.V. Kaza ••...•.•.•..•••••...••.•..•...•

1
3
5
11
16
22
29
31
36
41
46
53
55
60
65
71



XVIII

I.4. Design Methodologies ••••••••••••••••••••••••••••••••••••••••• 77
Axisymmetric Boundary Element Design sensitivity Analysis
J.T. Borggaard, and S. Saigal ••••••••••••••••••••••••••••• 79
Simultaneous Computation of Multiple sensitivities by a
Boundary Element Structural Analysis Formulation
J.H. Kane, and M. Stabinsky •••••••••••••••••••••••••••••••• 84
Lagrangian Interpretation of Nonlinear Design sensitivity
Analysis with continuum Formulation
J.B. Cardoso, and J.S. Arora ••••••••••••••••••••••••••••••• 90
A New Reanalysis Technique Suitable of Being used in Design
Automation and Opeimization
M. No, and s. Lopez-Linares •••••••••••••••••••••••••••••••• 95
Calculating Functionals for Arbitrary Geometries
A. Tristan-Lopez ••••••.•••••.••••••.••••••••.•••••••••••••• 100
I.5. CAD/CAM Automation ••••••••••••••.•••••••••.•••••••••••••••••• 105
A Graphics User Interface for Interactive Three Dimensional
Free-form Design
P.J. stewart, and K.-P. Beier •••••••••••••••••••••••••••••• 107
XCAD: A CAD Object-oriented Virtual Solid Modeler for an
Expert System Shell
B. Trousse ••••••••••••••••••••••••••••••••••••••••••••••••• 112
Chapter II: Automation in Manufacturing
Introduction ••••••••.•••••••••••••••••••••••••••••••••••••••• 117
II.1. Planning and Control •••.•••••••••••••••••••••••••••••••••••• 119
Decentralization of Planning and Control in CIM
S.K. Taneja, S.P. Rana, and N. Singh •••••••••••••••••••••• 121
An Intelligent Tactical Planning System: The Integration of
Manufacturing Planning Islands Using Knowledge Based

Technology
M.D. Oliff, J. Davis, L. Vicens ••••••••••••••••••••••••••• 126
Automated Process Planning for Mechanical Assembly Operations
J. Yung, and H. - P • Wang ••••••••••••••••••••••••••••••••••• 131
An Unorthodox Approach to Job-Scheduling
H. Bera .••........••.••...•.•••.•••••••••••••.••.••••••.•• 136


XIX

II.2. Group Technology ..•••••••••..•••.••••••••••.•••.•..••••••••. 143
Development of a Group Technology Workstation
R.M. Mackowiak, P.H. Cohen, R.A. Wysk, and C. Goss ••••.••• 145
A Comparison of Hierarchical Clustering Techniques for
Part/Machine Families Formulation
C.-H. Chu, and P. Pan ••.•..•.•.••••••••••.•••...•••••••••. 150
An Application of Fuzzy Mathematics in the Formation of
Group Technology Part Family
H. Xu, and H.-P. Wang ••••••••••.......•••••••.•..•••••.••• 155
Automatic Generation of Production Drawings and Part
Routings for Valve Spools
s.P. Pequignot, and A. Soom .•.•.....••..•..........•..•••. 160
Chapter III: Applications of Artificial Intelligence
Introduction .•••••••........•.•••...••...•.•.•.•.••••....... 165
III.l. AI Tools •..•.•.••••.••.••••......•••.....•..•••••.••••.•... 167
THINK: A C Library for Artificial Intelligence Tasks
M.E. Grost .........••.......••...,••.•..••....•.........•• 169
Using Artificial Intelligence Paradigms in Solving
Manufacturing Problems Demonstrated in the CPC
Stacking/Des tacking Expert System

T. Jaeger •••••••..••....•••••..•••••.••.•••••••.•••..•••. 174
A LISP-Based Environment for Simulation of Robot
Applications
M.C. Leu, and D.K. Pai ..•.....••••••.••••••.••.•••.•.•••. 179
III. 2. AI Methodologies •••...••••••••..•••••........•••••.••.•.••• 185
An Interactive Refutation Learning Approach for Skill
Acquisition in Knowledge-Based CAD System
Q. Zhu •••••••••••••.•••••...••...••••.•••••.•••••.••••••• 187
KBSS: A Knowledge-Based System for Scheduling in
Automated Manufacturing
A. Kusiak, and M. Chen •.••••.•••••.••••••••••••••••.••••• 192
Action, Reflective Possibility, and the Frame Problem
F.M. Brown, and S.S. Park ..•••..•.••.•••.••.•.•••..••.••• 197
Development of AI-Based Automated Process Planning Systems
G. S. Kumar •••••••••••••.•..•••.••.•••.••••••••••••••••••. 202
Automated Fixture Selection for Rotational Parts
P.H. Cohen, and B. Bidanda ••.•.•.•••••••••••••••.•••.•••• 207


xx
III.3. Decision Support Systems ••••••••••••••••••••••••••••••••••• 213
A Frame-Based User Enquiry Method for Supporting strategic
Operations Planning
o. B. Arinze •••••••••••••••••••••••••••••••••••••••••••••• 215
Construction of a Knowledge Base for the Detection of
Decision Errors
F. Mili, D. Shi, and P. Zajko •••••••••••••••••••••••••••• 220
On Representing Human Heuristic Reasoning
F. Mili, and A. Noui-Mehidiidi ••••••••••••••••••••••••••• 225
Chapter IV: Expert Systems

Introduction ••••.••••••••••••••••••••••••••••••••••••••••••• 231
Expert Systems for Diagnostics ••••••••••••••••••••••••••••• 233
An Expert System to Diagnose Failures in Industrial Robots
S.R. Vishnubhotla •••••••••••••••••••••••••••••••••••••••• 235
An Operations Analysis Expert System for Fiberglass
Manufacturing
G. Biswas, and M.D. Oliff •••••••••••••••••••••••••••••••• 240
Failure Detection and Diagnosis - Application to a
Grinding-Classification Circuit
T. Cecchin, J. Ragot, D. sauter, and M. Darouach ••••••••. 245
IV.2. Expert Systems for Design and Production ••••••••••••••••••• 251
Expert System for Specifying of CAD Software Systems
K. Ghosh, L. Villeneuve, and N.D. Tai •••••••••••••••••••• 253
An Expert System for IC Factory Design
P.K. Ramaswamy, and T.-L. Wong ••••••••••••••••••••••••••• 258
Towards an Expert System Architecture for Routine Design
- Focusing on Constraint Representation and an
Application Mechanism for Mechanical Design
Y. Nagai ••••••••••••••••••••••••••••••••••••••••••••••••• 263
Knowledge-Based Design Aid for Axisymmetric casting Parts
I.C. You, C.N. Chu, and R.L. Kashyap ••••••••••••••••••••• 268
IV.3. Expert Systems for Scheduling, Assembly, and Planning •••••• 275
Expert System Supervision of Robots During a
Vision-Assisted Assembly Task
J.B. Cheatham, C.K. wu, Y.C. Chen, and T.F. Cleghorn ••••• 277
Intelligent Lot-Size Advisor for MRP Systems
C.H. Dagli ••••••••••••••••••••••••••••••••••••••••••••••• 282
Intelligent Scheduling Systems for Parallel Machines with
Different Capability
G. Leininger ••••••••••••••••••••••••••••••••••••••••••••• 287

Expert System-Based Finite Scheduler
K. Barber, K. Burridge, and D. Osterfeld ••••••••••••••••• 291
IV.1.

Contents of Volume I •••••••••••••••••••••••••••••••••••••••••••••• 297
Contents of Volume III •••••••••••••••••••••••••••••••••••••••••••• 302
Author Index (Volume II) ••••••••••••••••••••••••••••••••••••••••• 307


Invited Lectures
Keynote Speech:
Eric Mittelstadt,
President and Chief Executive Officer, GMF Robotics
Auburn Hills, MI, USA
Banquet Speech:
Senator Carl Levin,
Chairman, Senate Small Business Sub Committee on
Innovation, Technology and Productivity,
US Senate, Washington, DC, USA
Plenary Sessions:
A Case for Computer Integrated Manufacturing
J. Tracy O'Rourke,
President and Chief Executive Officer,
Allen Bradley Co., Rockwell International, pittsburgh, PA, USA
Future Trends in AI/Robotics - A Pragmatic view
Randall P. Shumaker,
Director, Navy Center for Applied Research in AI,
Washington, DC, USA
Future of Engineering Design Practice
Kenneth M. Ragsdell,

Director, Design Productivity Center,
University of Missouri, Columbia, MO, USA
A New Departure in Programmable Robotic Design
G.N. Sandor,
Research Professor and Director, M.E. Design and
Rotordynamics Labs, University of Florida,
Gainesville, FL, USA
Cost Management as the criterion for Integrated Design and
Manufacturing
Ali Seireg,
Mechanical Engineering Department, university of wisconsin,
Madison, WI, USA
Earth observing Satellite System
Gerald A. Soffen,
Director, NASA Program Planning, Goddard Space Flight Center,
Greenbelt, MD, USA
Rapid Response to competition
Raj Reddy,
University Professor of Computer Science and Director
Robotics Institute, Carnegie Mellon University,
pittsburgh, PA, USA
Engineering Research Centers - A Vision for the 90's
Howard Moraff,
Program Director, Cross-Disciplinary Research,
National Science Foundation,
washington, DC, USA
Robots Beyond the Factory
W.L. Whittaker,
Robotic Institute, Carnegie Mellon University,
pittsburgh, PA, USA



CHAPTER I:
Computer-Aided Design
Introduction
One of the areas where algorithmic or parametric programming has
made its biggest contribution is in Computer Aided Design.
The
traditional CAD/CAM programs simply offer a STATIC visual aid to users
for the documentation of a preconceived part or assembly.
No provision exists to determine the effects of desired changes on performance.
Parametric programming is a concept of automating the product
design-development cycle by capturing its knowledge in terms of
parameters.
It maintains the real-world relationships between model
elements, their physical characteristics and the environments.
The
generic modeling, analysis, and optimization are used as integral
parts of the design.
In this way the parametric system "knows" the
identity and behavior of the individual part as well as the environment in which it fits or is subjected to, with all information residing symbolically in a unified data base.
This goes beyond the conventional CAD method of capturing geometry
in terms of points, lines and surfaces in a typical CAD/CAM system.
The algorithmic or parametric programming is based upon the exploitation of basic characteristics of the products' life cycle, which are
"generic" in nature.
The idea is similar to that of creating an
"expert system" except that the knowledge is derived largely from
algorithmic sources. Heuristic plays a smaller role.
The first section of this chapter reports on the advances that have
been made in developing techniques for shape optimization.

The
parametric methodologies which are employed for probabilistic
reliability analysis, and optimization are the subject of the second
section.
The third section provides some practical examples of
optimum design application.
The fourth section deals with the
elements of design methodologies, which are relevant to algorithmic or
parametric programming, while the final section reports on the advances in CAD/CAM automation areas.
The papers of this chapter are divided into the following sections:
I.1. Shape optimization
I.2. Probabilistic Design optimization
I.3. optimum Design Applications
I.4. Design Methodologies
I.S. CAD/CAM Automation


Shape Optimization


A Geometry-Based 2-Dimensional Shape
Optimization Methodology and a Software System
with Applications
V. Kumar, M. D. German and S. -J. Lee
Corporate Research and Development
General Electric Company
Schenectady, New York 12301

Summary
A geometry-based shape optimization methodology and a software system is presented for design

optimization of 2-D solids. Geometric modeling techniques are used for shape description and
for formulation of the optimization problem. An automatic mesh generation method is employed
for creating the finite element model initially and during the optimization iterations. The design
optimization of a turbine disc is discussed as an illustrative example.
Introduction
There has been a tremendous interest in recent years in using the numerical optimization technology for structural and mechanical design for a variety of reasons. From a technical viewpoint, it
provides a quantitative, systematic and computer-automatable interface between engineering and
design. From a business point of view, on the other hand, it offers a procedure for achieving an
optimal or the best possible design with several potential payoffs: weight (and therefore cost)
reduction, improved performance and increased engineering productivity. Shape optimization is
one of the most important topics in structural optimization, and it refers to design of two- and
three-dimensional structural components in which the geometry or topology varies during optimization iterations and, therefore, constitutes design parameters. The pioneering work of Bennett
and Botkin [1-3] on this subject has created interest in both academia and industries, and as a
result several papers and reports have been published during the past few years [4-6].

This paper presents a two-dimensional (2-D) shape optimization methodology and an associated
software package, SHAPE-OPT, with applications to practical design problems. The overall
technical approach is based on the integration of concepts of geometric modeling, automatic mesh
generation, numerical optimization, finite element methods and pre- and post-processing. The
geometric modeling techniques are used for shape description in terms of boundary points (fixed
as well as design variables) and geometric entities like lines, circular arcs and splines. The structural optimization formulation is also carried out at the geometry level in that the stress and other
design constraints are specified in terms of boundary points, geometric entities and domains
rather than individual finite elements or mesh points. Automatic mesh generation is employed for
creating the initial finite element model and also for automatic remeshing as the shape changes
during optimization. The issues of mesh updating between two successive remeshing and for
design sensitivity calculations are also addressed together with a shape control procedure. The
commercial fInite element code ADINA [7] is employed for structural analysis, and a publicdomain software package COPES/ADS [8] is used for numerical optimization. The postprocessing software packages MOVIE.BYU, SUPERTAB and PLOTlO are utilized for


6


displaying the shape, finite element mode~ stress contours, objective function, design constraints
and other pertinent information at various optimization steps. The approach/software developed
is demonstrated on a 2-D real-life industrial shape optimization problem. Several topics for
future developments are also briefly discussed.
Integration of Finite Element Analysis and Numerical Optimization
Essential elements of integrating a finite element software with a numerical optimization code are
the design sensitivity analysis and an interface program between analysis and optimization programs. In the present work, both the finite difference and the semi-analytical (or implicit
differentiation) approaches were implemented in the ADINA code for design sensitivity computations. The finite difference method was implemented external to ADINA, whereas the semianalytical approach required substantial internal finite element enhancements. Both size and
shape optimization problems were considered, for static as well as dynamic cases and encompassing a wide range of element types (truss, beam, plate and 2-D continuum). Centrifugal and thermal loadings were also included for 2-D solid elements. The technical issues involved and their
ADINA implementation, a comparison of the two approaches in terms of computational
efficiency, solution accuracy and the ease of software implementation, and other related topics
will be discussed at length in a forthcoming article [9].
Subsequent to the development of the ADINA design sensitivity analysis procedures as described
above, a number of interface programs were developed between ADINA and the optimization
software ADS. First, an optimizer to analyzer processor OPT-AN was developed which automatically updates an ADINA input file to incorporate shape and/or size design changes that occur
during various optimization iterations. Similarly, an analyzer to optimizer processor AN-OPT
was also developed which, through an intermediate binary output file .BOF, takes the ADINA
output file as the input, computes objective functions and design constraints as specified by the
user and transmits this data to the optimizer. The AN-OPT processor and the .BOF file were
also interfaced with a number of post-processing software packages like MOVIE.BYU and
SUPERTAB so that the user can display the structural shape, stress contours, iteration histories
of objective functions and constraints and other analysis/design quantities of interest. These
developments will be described in detail in reference [9].
Geometry-Based Shape Description, Attribute Specification and Problem Formulation
An approach was developed for shape description and contro~ attributes or boundary conditions
specification, and optimization problem formulation at the geometry level rather than the finite
element level by using the geometric modeling techniques. Specifically, an in-house geometric
modeler BZGEOM [10] was used, but the concepts developed are generic and can be readily
applied with most of the commercially available geometric modeling software packages. In this

approach, the shape is described in terms of boundary points and boundary curves (lines, circular
arcs and cubic splines) to form simple- or multiple-connected regions. Boundary points and
curves which are permitted to vary during optimization are termed design points and curves,
respectively. Design variables are specified in terms of Cartesian coordinates of design points. A
concept of shape design variable linking was evolved that allows the user to specify different
design models (i.e., number and distribution of design variables, and number and types of design
curves, etc.) at different optimization stages during the input file preparation stage without having
to restart new batch jobs. Similarly, shape control procedures were introduced to eliminate shape


7

irregularities during optimization iterations, for example, by including constraints on slopes and
curvatures at certain boundary points. These developments are not elaborated upon here any
further because of text limitations, but the relevant details can be found in reference [11].
The attribute specification for traction/displacement boundary conditions and the optimization
problem formulation, i.e., objective function and constraints, is also carried out at the geometry
level in terms of boundary points, boundary curves and zones rather than at the level of finite elements and associated node points. This procedure provides an effective treatment of deaJing with
different number of nodes and elements that arise when an automatic mesh generator, to be discussed in a subsequent section, is used to create new finite element models for updated shapes at
various optimization iterations. If the design constraints were tied to elements and nodes, the
number of constraints will change when the shape is remeshed using automatic mesh generator,
and this in turn will cause several fundamentally technical as well as software development related
problems. For similar reasons, many difficulties would also arise if the traction and displacement
boundary conditions were specified in terms of elements and nodes. In the present work, an inhouse software MAP_LOADS [10] was utilized for specifying attributes at the geometry level in
an interactive manner via the use of the geometric modeler BZGEOM. It allows linear and
splined distributions of pressure and displacement along a line, arc, or spline, and some enhancements were also made for specifying fixed displacement, concentrated forces and prescribed temperature distributions. A number of interface programs were developed for integrating
BZGEOM, MAP_LOADS and an automatic mesh generator which is described next.
Integration With Automatic Mesh Generation
Shape changes resulting from optimization iterations require updating of the mesh used in the
finite element analysis. When these changes are small, the mesh can be updated by relocating the

nodes, i.e., by utilizing the r-method of mesh refinement. For moderate or large shape variations,
however, it becomes necessary to modify the mesh topology itself, requiring thereby an altogether
new fInite element model. The present study employs an in-house geometry-based, fully
automatic, 2-D continuum, finite element mesh generator, QUADTREE [12,13], for creating the
initial finite element model and also for automatic remeshing as the shape changes during optimization iterations. Using the shape description file from the geometric modeler BZGEOM as the
input, the QUADTREE software develops through a number of file format translators, the finite
element connectivity and nodal data as required by the ADINA input file in a fully automatic
manner without the user's intervention. Complete remeshing is not required at each optimization
iteration; it is performed only when shape changes are significant to warrant an entirely new mesh
according to a user supplied criterion. The mesh is otherwise updated by simply relocating the
boundary nodes followed by the Laplacian method of interior smoothing.
The shape design sensitivity analysis involves perturbing a shape design variable by an
infInitesimal amount, performing the finite element analysis of the perturbed geometry, and then
using the finite difference method (or the semi-analytical approach) to calculate the required gradients. Automatic remeshing or the use of QUADTREE is not necessary for creating the finite
element mesh of the perturbed geometry. It is obtained from the mesh associated with the unperturbed shape by moving the boundary nodes to the perturbed boundary and employing a
geometric modeling utility software. Since the shape perturbations for design sensitivity calculations are rather small and only one design variable is changed at one time, the interior smoothing
is generally not required. It is important to remark that the application of QUADTREE or some