•~~

ETFA '97-

IEEE 6th IntemationalConference on

EmergingTechnologies and

Factory AutomationProceedings

September 9-12,1997veLA

Los Angeles, California

Sponsored by the IEEEIndustrial Electronics Society

In Technical Co-sponsorship withSociety of Instrument and ControlEngineers, Japan

In Cooperation with MicromechanicalSystems Panel of the ASMEDynamic Systems and ControlDivision

r

1997 IEEE 6th INTERNATIONAL CONFERENCE ON EMERGING TECHNOLOGIESAND FACTORY AUTOMATION PROCEEDINGS

Abstracting is permitted with credit to the source. Libraries are permitted to photocopy beyond the limitsofU.S. copyright law for private use ofpatrons those articles in this volume that carry acode at thebottom of the first page, provided the per-copy fee indicated in the code is paid through the CopyrightClearance Center, 222 Rosewood Drive, Danvers, MA 01923. For other copying, reprint, or republicationpermission, write to the IEEE Copyright Manager, IEEE Service Center, 445 Hoes Lane, Piscataway, NJ08855-1331. All rights reserved. Copyright © 1997 by The Institute of Electrical and ElectronicsEngineers, Inc.

IEEE Catalog Number:

Library of Congress Number :

ISBN Softbound:ISBN Microfiche:

97TH8314

97-80024

0-7803-4192-90-7803-4193-7

Additional copies of this publication are available from

IEEE Operations CenterP. O. Box 1331445 Hoes Lane

Piscataway, NJ 08855-1331 USA

1-800-701-IEEE1-732-981-0060

1-732-981-9667 (FAX)833-233 (Telex)email: customer.service@ieee.org

Track A: Advanced Manufacturing Systems

13:20 - 15:40, Wednesday, September 10, 1997Session Al: Innovations in Design and Integrated

ManufacturingChair: P. Ranky, New Jersey Institute ofTechnology

Co-Chair: Yushun Fan, Tsinghua University

Track B: Flexible and Intelligent Automation

13:20 -15:40, Wednesday, September 10, 1997Session Bl: Advances in Industrial

Automation

Chair: Madhukar Pandit - University of Kaiserslautern

Co-Chair: Kuang C. Wei - Ford Research Laboratory

MACIP: Solution for CIMS Implementation inManufacturing EnterprisesY. Fan and C.Wu - Tsinghua Univ.A Framework ofVirtual Design forCustomizationM. M. Tseng, J. Jiao and C. J. Su - The Hong KongUniv. ofScience & TechnologyIntelligent and Integrated Production ControlSystem for Manufacturing Enterprise,Y.S. Noh and 1. Favrel - INSA (Institut Nationaldes Sciences Application) de LyonAn Automatic CAPP System for RotationalPartsJ. C. Rico, C. Suárez, S. Mateos, E. Cuesta andA. Duarte - Universidad de OviedoThe Research, Development, Validation andApplication of an Advanced Concurrent(parallel) Engineering Research Toolset at RolIs­Royce Motor Cars Limited (I)P. T. Ashton - RoIls-Royce Motor Cars LimitedP. G. Ranky - New Jersey Institute ofTechnologyLocating Method Analysis Based Rapid FixtureConfiguration DesignY. Rong and X. Li - Southern IIIinois University atCarbondaleGertrude: 00 for BPRL. Benedicenti, P. Predonzani, andT. Vernazza - Universita di GenovaG. Succi - Universita di Trento

N/A - Papers are not available at time o[printing.(1) - Papers or sessions are invited.

1

7

15

19

24

27

33

xvi

WinFT: Using Off-the-Shelf Computerson Industrial EnvironmentsJ. Carreira, D. Costa and J. G. Silva­University of CoimbraSpeed Estimators for Indirect Field­Oriented Control of Induction MotorsL. U. G6kdere and C. W. Brice ­University of South CarolinaM. A. Simaan - University of PittsburghThe Development of PC-Based HighlyModel-Mixed Engine DispatchingSystemsC. H. Kuo and H. P. Huang - NationalTaiwan UniversityK. C. Wei - Ford Research LaboratoryS. S. H. Tang - Ford Lio-Ho Motor Co.Remote Wireless Measuring SystemsT. J. Su, C. H. Yang - NationalKaohsiung Institute ofTechnologyT. 1. Moir - University of PaisleyG. J. Jong - National Kaohsiung Inst. ofTech., and University ofPaisleyA Comparison of Passivity-Based andInput-Output Linearization Controllersfor Induction MotorsL. U. G6kdere, C. W. Brice - Universityof South Carolina andM. A. Simaan - University ofPittsburghLearning Control of Cyclic ProductionProcesses

M. Pandit and S. Baque - UniversityKaiserslauternOptimal Model Matching WithDisturbance Rejection for the Non­Minimum Phase PlantsX. Wu, X. Xie and Y. Zhong - TsinghuaUniversity

39

45

49

55

60

64

71

An Automatic CAPP System for Rotational Parts

Carlos Rico, Carlos Suirez, Sabin0 Mateos, Eduardo Cuesta, Angel Duarte Dept. of Manufacturing Engineering

Oviedo University Campus de Viesques, 33204 Gijon

Spain

Abstruct - This paper describes a turning process planning of rotational parts. It also reports general aspects to develop an automatic planning system, from the part geometrical identification to the NC program generation. The minimum number of grippings (based on the part shape), the blank type, the boring operations and the geometrical tolerance presence are established in order to determine the machining operations. The proposed method used to recognize the part geometry is based on a combination between an automatic identification of 2D profiles and a feature-based method. By this way, the developed CAPP system could use geometrical information given by any CAD system.

1. INTRODUCTION

In the Manufacturing Engineering field, the process planning can be seen as a bridge between the design phase and the manufacturing phase (Fig. 1). It can be said that the input to a Computer Aided Process Planning System (CAPP) is the description of the product and the outputs are manufacturing instructions. An automatic process planning has to take into account:

- Interpretation of the part geometry - Decomposition of the process in basic machining

operations - Selection of tools and grips

- Calculation of the working conditions - Generation of NC programs - Generation of technical documents (worksheets)

The traditional forms used to make this planning are based on variant or generative approaches. On the variant planning systems, parts are codified and classified in families according to design and manufacturing similarities [3][7]. The planning process to each family of parts are loaded in the systems in order to be easily recovered and edited during the application phase of the program. Generative approach systems include logical decisions and algorithms to obtain the process planning automatically from the CAD model of the part [ 111.

Nevertheless to all efforts made in order to develop a planning system based on computer, only a few systems are now using in industry. The main reasons for the slowness of this process of assimilation are the complexity of planning tasks and the difficulty in adapting a general system to the particular requirements of each factory and sector.

11. TURNING PROCESS PLANNING

The present work tries to make a previous study of all the activities related to process planning of turning operations,

Constraints

Manufacturing Instructions

Product Description

Q Specifications Geometry Recognition

Operations Definition

Selection of Fixtures

Selection of Cutting Tools

Seledion of Cutting Conditions

NC Programs Generation

Worksheets Generation ~~~ ~ ~ ~

Fig. 1 The procesS Planning links design and manufacturing

0-7803-4192-9/97/$10.00 0 1997 IEEE 19

PART

SHAPE

Decreasing Increasing-Decreasing I Decreasing-Increasing Fig. 2 Basic classitlcation of rotational parts

that allows the development of an automatic planning system using a computer [2] [3]. The system will be able to assist the engineer to take any decision (from the part identification to its final manufacturing) which are aimed to the elaboration of the process planning. This involves the knowledge of all the variants and parameters associated to the turning process and the relations between them and also the elaboration of algorithms and strategies in order to take decisions automatically [lo].

The attainment of this general objective implies the solution of several particular problems, which will be explained next.

A.- Recognition of the part geometry

Nowadays it is accepted that the geometrical models need high level of information associated to them in order to adjust to the requirements of an integration between the design and manufacturing systems.

At present, this integration is one of the most important activities in the research field. The main problem is the determination of a schema of product representation. This schema can operate as carrier of information. The aim is to obtain a unique representation for both CAD systems and CAM systems, improving the integration between them.

In order to solve this problem, feature-based model systems arise yet. These features have geometrical information associated to a particular part and also attributes about some (or almost all) activities for its manufacturing.

When working with features, feature recognition systems [l] and feature-based systems can be found [6] [7]. In the present work a combined system has been chosen, This system can make an automatic recognition of the part from

the 2D profile, and can also adds modifications using feature pattems such as chamfer, fillet, grooves, etc. A 2D representation has been chosen because it is the most usual for rotational parts.

Most of the rotational parts manufactured on a lathe contain geometrical entities given by only lines an arcs entities. For this reason the recognition process has to take into account the identification of these two entity types.

The usual method to draw a 2D profile of a rotational part consists of drawing a basic profile and then adding modifications such as chamfer, fillet, thread, etc. These types of modifications have been included as features to accelerate the part design. By adding these features the system can also recognize special manufacturing areas such as threads and knurlings. However the user can also introduce directly these modifications (chamfers, grooves, etc.) during the profile drawing.

B. Elaboration of machining plans

The generation of machining plans requires the previous recognition of the part, from both a geometrical and technological point of view. This information constitutes the base for the selection of elementary machining operations and all possible machining plans [SI.

In the case of turning rotational parts, the gripping system and its location in relation with the part surfaces must be taken into account. In this sense, the number of possible solutions can be large. This implies to make an ordering to choose the best plan. This selection can be based on maximum production rate criterion or on the minimum cost per component criterion. For this reason the previous knowledge of the cutting conditions are needed. So, the best machining plan has to be done at the end of the process.

20

INITIAL SHAPE: BLANK, FINAL SHAPE: DECREASE TYPE LORING )perations

.EFT AND !lGHT

TOLERANCES OF FORM AND LOCATION

Left, ext. Left, ext. Left, ext. Right, ext. Left, int. Right, mt. Left, int.

I I

<m Bar from Bar from Bar I extchucking 1 extxhudcing 1 ext.chucking I int.chudting 1 int chucking 1 int chucking

2 grips I I 2grips

2grips 1 2grips

Special

Left I Right 1

Fig. 3 Classification matrix for decreasing shape rotational parts with boring operations on their left and right faces

BORING Operations

LEFT

5:J

INITIAL SHAPE: BLANK. FINAL SHAPE: INCREASE-DECREASE TOLERANCES OF FORM AND LOCATION

Left, ext. NO I Right, ext.

Left, ext. Right, int.

P 2 grips

2 grips

.----_

P from Ear

\E3 from Bar

\ I fromBar

from Bar

Left, ext. Left, int. Right, int. Right, int.

Special surface

NO

Lek

Right

All

Fig. 4 Classification matrix for increashg-decreashg shape rotational parts with boring operations on their left or right face

21

With the purpose of establishing the minimum number of grips for a given part automatically, a classification of the part based on its general shape is proposed. This previous classification is very important when clamping lathe phases are considered. In this sense, three basic shapes had been considered (Fig. 2). Besides, they had been subdivided depending on the blank type, whether there are boring operations or not (in one or two faces) and whether there are form and positions tolerances or not (Fig. 3.and 4).

On the identified profile entities, the geometrical information (extreme point coordinates, orientation, length, radius, etc.) can define the different material areas for each grip. These areas have to be removed taking into account the initial and final profile of the part (Fig. 5). The strategy applied here is based on the reconstruction of the initial shape from the final shape.

The material areas defined by this method allow to determine the main roughing areas. Only the concaves areas need to be analized in order to check whether they can be removed in one operation or not. The identification of threaded and knurled areas is immediate because they had been designed using features. Depending on the generation method applied, grooves, chamfers and fillets can be identified through their dimensions (lengths, radius) or by features.

Fig. 5 Roughmg machining areas for a decreasing shape rotational parts

C. Selection of tools and cutting conditions

Besides the workpiece (profile, material, machining directions, finishing conditions, etc.) and the machine tool (gripping system, workholding size) the selection of the tools is linked to the cutting conditions due to the great relation between them [4][9][12]. So, in a turning process, the geometry of the cutting inserts will be determined by the tool application area. This application area will be determined by the range of cutting depths and the feedrate of each operation [5 ] . It also have to be taken into account the possibility of machining several surfaces of the part with the same tool in order to minimize the amount of tools on the lathe turret. Besides, priority criteria must be

established to choose among several possible tools.

related to materials, tools and machines. This phase of the planning process needs a data base

D. Toolpaths generation

In order to obtain an integrated planning system in operations with numerical control machine tools, the process have to finish with the generation of toolpaths and the corresponding NC programs. The possibility of automatic identification of the geometry and technological data of the workpiece provides the generation of different NC programs and also reduces the user intervention. With the generated information till now, machining times and costs can be obtained.

E. Generation of worksheets

As a summary, the CAPP system will make worksheets of the Werent planning tasks. Worksheets include the characteristics of machine tool, the different machining operations, the sequential among them, the tools, the cutting conditions, the NC program and the approximately time and cost of each operation.

III. CONCLUDING REMARKS

Before the establishment of the machining operations, the development of an automatic system for turning process planning obliges to take into account special functions aimed to the identification of the part geometry. The recognition of the part profile must allow not only the machining operation establishment but also the gripping number. To define the order of different machining operations, the part clamping system has to be taken into account. In the present work, a initial classification of rotational parts has been proposed. This classification is based on the final and initial shape part, whether it has boring operations or not and whether it has tolerances or not. Following this method the minimal gripping number can be known. To determine the machining operations, the selected method is based on the reconstruction of the initial shape from the final shape.

IV. ACKNOWLEDGEMENTS

This paper belongs to the initial stage of a Research Project financed by CICYT (Spanish State Commission for Science and Technology) (Ref.-TAP95-0681) aimed at the design of CAPP systems.

V. REFERENCES

[l] S. Alind, G.R. Graves, C.M. Parks and L.A. Mann,

22

“A methodology for recognizing features in two- dimensional cylindrical part designs”, Int. J. of Prod. Res., vol. 28, no. 8, 1990, pp. 1401-1416. D.S. Domazet and S.C.-Y. Lu, “Concurrent design and process planning of rotational parts”, Annuls of the CIRP, vol. 41, no. 1, 1992, pp. 181-184. F. Giusti, M. Santochi and G. Dini, ”KAPLAN: a knowlwdge-based approach to Process Planning of Rotational Parts”, Annuls of CIRP, vol. 38, no. 1,

S. Hinduja and G. Barrow, “SITS: a semi-intelligent tool selection system for turned components”, Annuls of CIRP, vol. 42, no. 1, 1993, pp. 535-539. S. Hinduja, D.J. Petty, M. Tester and G. Barrow, “Calculation of optimum cutting conditions for turning operations”, Proc. Instn. Mech. Engrs., Part

A.H. Juri, A. Saia and A. de Pennington, “Reasoning about machining operations using features-based models”, Int. J. Prod. Res., vol 28, no. 1, 1990, pp.

R.-K. Li, “A part-feature recognition system for rotational parts”, Int. J. of Prod. Res., vol. 26, 1988,

[2]

[3]

1989, pp. 481-484. [4]

[5]

B, vol. 199, 1985, 81-92. [6]

153-171. [7]

pp. 1451-1475. [8] C.-H. Lin and H.-P. Wang, “Optimal operation

planning and sequencing: minimization of tool changeovers”, Int. J. Prod. Res., vol. 31, no. 2, 1993,

[9] J.C.S. Plummer and R.G. Hannam, “Design for manufacturing using a CAD/CAM system: a methodology for turned parts”, Proc. Instn. Mech. Engrs. vol. 197, 1993, pp. 184-195. S. Rajiv and R. Shivakumar, “Metex: an expert

system for machining planning”, Int. J. Prod. Res., vol. 30, no. 7, 1992, pp. 1501-1516. S.H. Yeo, Y.S. Wong and M. Rahman, “Integrated

knowledge-based machining system for rotational parts”, Int. J. Prod. Res., vol. 29, no. 7, 1991, pp.

J.H. Zhang and S. Hinduja, “Determination of the optimum tool set for a given batch of turned components”, Annals of the CIRP, vol. 44, no. 1,

pp. 311-324.

[lo]

[ll]

1325-1337. [12]

1995, pp. 445-450.

23