The Design Guidelines Collaborative Framework: A Design for Multi-X Method for Product Development

The Design Guidelines Collaborative Framework Stefano Filippi · Ilaria Cristofolini The Design Guidelines Collaborat...

Author: Stefano Filippi | Ilaria Cristofolini

51 downloads 768 Views 5MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

The Design Guidelines Collaborative Framework

Stefano Filippi · Ilaria Cristofolini

The Design Guidelines Collaborative Framework A Design for Multi-X Method for Product Development

123

Prof. Stefano Filippi University of Udine DIEGM Department Viale delle Scienze, 208 33100 Udine Italy [email protected]

Dr. Ilaria Cristofolini University of Trento DIMS Department Via Mesiano, 77 38050 Trento Italy [email protected]

ISBN 978-1-84882-771-4 e-ISBN 978-1-84882-772-1 DOI 10.1007/978-1-978-1-84882-772-1 Springer Dordrecht Heidelberg London New York British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Control Number: 2009940605 c Springer-Verlag London Limited 2010 Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms of licenses issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the publishers. 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 laws and regulations and therefore free for general use. The publisher makes no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility or liability for any errors or omissions that may be made. Cover design: eStudioCalamar, Figueres/Berlin Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)

Preface

In the industrial design and engineering field, product lifecycle, product development, design process, Design for X, etc., constitute only a small sample of terms related to the generation of quality products. Current best practices cover widely different knowledge domains in trying to exploit them to the best advantage, individually and in synergy. Moreover, standards become increasingly more helpful in interfacing these domains and they are enlarging their coverage by going beyond the single domain boundary to connect closely different aspects of the product lifecycle. The degree of complexity of each domain makes impossible the presence of multipurpose competencies and skills; there is almost always the need for interacting and integrating people and resources in some effective way. These are the best conditions for the birth of theories, methodologies, models, architectures, systems, procedures, algorithms, software packages, etc., in order to help in some way the synergic work of all the actors involved in the product lifecycle. This brief introduction contains all the main themes developed in this book, starting from the analysis of the design and engineering scenarios to arrive at the development and adoption of a framework for product design and process reconfiguration. In fact, the core consists of the description of the Design GuideLines Collaborative Framework (DGLs-CF), a methodological approach that generates a collaborative environment where designers, manufacturers and inspectors can find the right and effective meeting point to share their knowledge and skills in order to contribute to the optimum generation of quality products. The DGLs-CF integrates several tools to achieve this goal: a method to evaluate the compatibility between the products and the processes adopted to manufacture and verify them, a language and a data structure to formalise the knowledge about products, processes and so on, some procedures to infer new information from the gathered knowledge and, finally, a usable way for information sharing among the different domains. As a result, the DGLs-CF gives the users a sort of to-do list for modifying the product model and/or the physical representation of it through the whole development process, in order to achieve the best compromise between product functionalities and the characteristics of the available technologies.

vi

Preface

The first chapter deals with the need to clarify the concepts and terms related to the product lifecycle, regarding the context where this project develops. The concepts like the product lifecycle itself, the engineering design process, the Design for X, Concurrent Engineering, the Standards, the design methods, etc., are related to each other like some sort of Chinese boxes. The state of the art of each of them is described and a wide range of references is added to help in detailing further understanding of the topics. The second chapter starts to describe the DGLs-CF by introducing the previous work related to it. The project started around the year 2002 and up to now four releases of design guidelines have been developed. The chapter describes in detail this course for many reasons. First, the DGLs-CF is so complex and articulated that there was the need for this sort of background information to get all the details of it at best. Second, the DGLs-CF development trail put to evidence time by time different aspects, sometimes unexpected, of the product lifecycle management; this has been really valuable in targeting and tuning the authors’ competencies and skills and, hopefully, could be of help to the reader for the same reasons. Third, the examples of the application of the different releases in the field could help in gradually understanding the adoption of this sort of tool in the everyday work of designers, manufacturers and inspectors. The third chapter describes the DGLs-CF in detail. A simple formalism, the IDEF0 — Integration DEFinition — diagrams, has been used to keep track of the development and adoption of the framework. This choice allowed one to describe in a really clear way all the activities required by the DGLs-CF, the interfaces between them in terms of input, output, controls — the standards, other guidelines and, in general, any references used in performing the activities — and, finally, mechanisms — the resources expressed in terms of humans, procedures, software packages, etc. The fourth chapter describes the experiences in applying the DGLs-CF in the field. Some industrial products, together with the available technologies in their design, manufacturing and verification domains, have been evaluated using the DGLs-CF. The redesign/reconfiguration packages, the lists of actions to perform on the product coming as a result from the DGLs-CF elaboration, have been considered and applied. The fifth chapter acts as the final discussion about the project. Now the DGLsCF presents a good level of maturity concerning the methodological aspects and the most of the procedures used in the activities have been detailed as well; nevertheless, much work still remains to be done, especially on implementation and usability issues. This chapter summarises the hints coming from the different releases of the design guidelines and describes if and how the DGLs-CF matches them. These considerations lead to a list of suggestions for future work.

Acknowledgments

The authors would like to thank Dr. Barbara Motyl for her contribution during the development of the case studies described in Chap. 4. Anthony Doyle and Claire Protherough from Springer are also gratefully acknowledged: their precious help and kind patience contributed to the overall production of the book. Thanks are also due to Nadja Kroke and her colleagues from Le-tex: they did an excellent job in the detailed task of typesetting the book. Stefano Filippi would like to thank Prof. Umberto Cugini, who contributed definitely to his profession since the beginning, and Prof. Camillo Bandera, who has been collaborating with him during these last years. Ilaria Cristofolini would like to thank Prof. Giorgio Wolf, who introduced her into the world of engineering design, and Prof. Alberto Molinari, her precious point of reference since the beginning of her studies. She also thanks Marco, Stefano and Elena for their continuous encouragement, patience and smiles.

Contents

1 State of the Art in the Field ............................................................................... 1 1.1 The Product Lifecycle.................................................................................. 2 1.2 The Engineering Design Process ................................................................. 3 1.2.1 Definitions and Concepts ..................................................................... 3 1.2.2 The Problem Solving Process............................................................... 5 1.3 Concurrent Engineering............................................................................... 7 1.4 Standards ..................................................................................................... 8 1.4.1 Fundamentals ....................................................................................... 8 1.4.2 Highlights on the ISO GPS Features.................................................. 10 1.5 Design Methods......................................................................................... 13 1.6 DfX Methods ............................................................................................. 15 1.6.1 Overview ............................................................................................ 15 1.6.2 Design for Manufacturing — DfM .................................................... 15 1.6.3 Design for Assembly — DfA............................................................. 18 1.6.4 Design for Test and Maintenance — DfTM — and Design for Verification — DfV .................................................................................... 19 1.6.5 Design for Reliability — DfR ............................................................ 20 1.6.6 Design for Environment — DfE ........................................................ 21 Summary.......................................................................................................... 22 References ....................................................................................................... 23 2 The DGLs-CF — Introduction and Background .......................................... 31 2.1 Overview of the History of the DGLs........................................................ 32 2.2 First Release of the DGLs. The Beginning................................................ 34 2.2.1 Conceptual Diagram........................................................................... 34 2.2.2 Knowledge Matrix ............................................................................. 35 2.2.3 Case Study.......................................................................................... 37 2.2.4 Discussion .......................................................................................... 40 2.3 Second Release of the DGLs. Improvements and the Synergy with the ISO GPS ............................................................................................ 41 2.3.1 Conceptual Diagram........................................................................... 41 2.3.2 Knowledge Matrix ............................................................................. 42

x

Contents

2.3.3 Case Study..........................................................................................44 2.3.4 Discussion ..........................................................................................45 2.4 Third Release of the DGLs. Thinking Big .................................................47 2.4.1 Conceptual Diagram...........................................................................47 2.4.2 Knowledge Matrix..............................................................................49 2.4.3 Roadmap for the Adoption of the DGLs ............................................49 2.4.4 Case Study..........................................................................................49 2.4.5 Discussion ..........................................................................................60 Summary..........................................................................................................63 References........................................................................................................64 3 Detailed Description of the DGLs-CF.............................................................65 3.1 Conceptual Diagram ..................................................................................66 3.2 IDEF0 Fundamentals .................................................................................66 3.3 Purpose, Viewpoint and the Node A-0 of the DGLs-CF IDEF0 Diagram.67 3.3.1 Purpose and Viewpoint ......................................................................68 3.3.2 The Node A-0 — TOP Level .............................................................68 3.4 The Node A0. The Main Phases and the Modules.....................................71 3.4.1 Overview of the Three Main Phases...................................................71 3.4.2 Overview of the Seven Modules ........................................................72 3.5 The Node A1. First Setup ..........................................................................74 3.5.1 Activities ............................................................................................76 3.5.2 Modules of Interest Here ....................................................................77 3.6 The Node A2. Technological Configuration..............................................91 3.6.1 Activities ............................................................................................91 3.6.2 Modules of Interest Here ....................................................................92 3.7 The Node A3. Redesign/Reconfiguration Package Generation .................93 3.7.1 Activities ............................................................................................95 3.7.2 Modules of Interest Here ....................................................................95 3.8 Discussion..................................................................................................99 Summary..........................................................................................................99 References......................................................................................................100 4 Adopting the DGLs-CF in the Field..............................................................101 4.1 Rapid Prototyping — RP — Technologies..............................................102 4.2 Coordinate Measuring Machines — CMMs ............................................102 4.3 First Case Study. A Mechanical Bracket Built Using FDM ....................104 4.3.1 FDM Fundamentals..........................................................................105 4.3.2 First Setup ........................................................................................106 4.3.3 Technological Configuration............................................................119 4.3.4 Redesign/Reconfiguration Package Generation ...............................120 4.4 Second Case Study. A Mould Insert Built Using SLA ............................125 4.4.1 SLA Fundamentals ...........................................................................126 4.4.2 First Setup ........................................................................................128 4.4.3 Technological Configuration............................................................142

Contents

xi

4.4.4 Redesign/Reconfiguration Package Generation ............................... 144 4.5 Third Case Study. An Angle-Shaped Connector Built Using SLS .......... 148 4.5.1 SLS Fundamentals ........................................................................... 149 4.5.2 First Setup ........................................................................................ 150 4.5.3 Technological Configuration............................................................ 163 4.5.4 Redesign/Reconfiguration Package Generation ............................... 164 Summary........................................................................................................ 169 References ..................................................................................................... 169 5 Discussion and Hints for Future Work......................................................... 173 5.1 Conceptual Diagram and Knowledge Organisation................................. 173 5.2 Knowledge Description and ISO GPS Adoption..................................... 174 5.2.1 Product Features and Technological Characteristics........................ 174 5.2.2 Rules, Compatibility Expressions and Actions ................................ 175 5.3 Costs ........................................................................................................ 176 5.4 Implementation/Automatisms.................................................................. 176 5.5 DGLs-CF Adoption Process .................................................................... 177 5.5.1 Activity Timing and Concurrencies ................................................. 177 5.5.2 DGLs-CF Usability .......................................................................... 179 Summary........................................................................................................ 181 References ..................................................................................................... 181 Appendix. Generation of Some Meaningful Compatibility Expressions...... 183

Abbreviations

ABS API ASME CAD CAD/CAM CE CMM D4V DBMS DfA DfE DfM DfMA DfR DfTM DfV DfX DGLs DGLs-CF DMLS DOE DPs DSM EDM FDM FEM FMEA FRs GPS ICOM IDEF

Acrylonitrile Butadiene Styrene Application Programming Interface American Society of Mechanical Engineers Computer Aided Design Computer Aided Design/Computer Aided Manufacturing Concurrent Engineering Coordinate Measuring Machine Design for Verification Data Base Management Systems Design for Assembly Design for Environment Design for Manufacturing Design for Manufacturing and Assembly Design for Reliability Design for Test and Maintenance Design for Verification Design for X Design Guidelines Design Guidelines Collaborative Framework Direct Metal Laser Sintering Design Of Experiments Design Parameters Design Structure Matrix Electrical Discharge Machining Fused Deposition Modeling Finite Element Method Failure Modes and Effects Analysis Functional Requirements Geometrical Product Specifications Inputs, Controls, Outputs, Mechanisms Integration DEFinition

xiv

Abbreviations

ISO TC ISO ISO/TR MET MTBF NC OpenADE PERT RP SADT SLA SLS STL TRIZ UML UV

International Organization for Standardization Technical Committee International Organization for Standardization International Organization for Standardization Technical Report Materials, Energy, Toxicity Mean Time Between Failure Numerical Control Open Assembly Design Environment Project Evaluation and Review Technique Rapid Prototyping Structured Analysis and Design Technique Stereolithography Selective Laser Sintering Stereolithography file format Teoriya Resheniya Izobretatelskikh Zadatch (Theory of Inventive Problem Solving) Unified Modelling Language UltraViolet

1 State of the Art in the Field

This chapter describes the scenario where the DGLs-CF project takes place. Fig. 1.1 shows schematically the main items involved in product development and acts as an index for the present chapter, pointing directly to the description paragraphs.

Product lifecycle (Par. 1.1) Engineering design process (Par. 1.2) Concurrent Standards Engineering (Par. 1.4) (Par. 1.3) Design methods (Par. 1.5) and DfX methods (Par. 1.6)

Fig. 1.1 Items of interest in the product development involved here

Specifically, the figure highlights the role of the engineering design process inside the product lifecycle and its co-existence with standard-related issues in a collaborative environment. All the pieces of information needed for an effective engineering design process must be formalised, elaborated and made usable, and this can be performed by many different design methods and tools, which establish links between the engineering design process and the other phases of the product lifecycle. The term Design for X — DfX — means here the way to place, in terms of time and space, the design methods inside the engineering design process; in other words, a design method becomes a DfX method when a goal and a placement are associated with it.

2

1 State of the Art in the Field

Inside this scenario, the DGLs-CF can be considered a DfX method and this is why this chapter is ultimately focused on describing the state of the art in the field of DfX methods. Many authors are considered here, whose work appeared interesting during the development of the DGLs-CF. Some of them are cited explicitly; however, the list does not claim to be exhaustive because of the extent of this research field.

1.1 The Product Lifecycle The product lifecycle may be described by different phases that could be grouped into four areas: product development, production and delivery, use, and end of life (Ullman 2002a). These areas are briefly summarised here, focusing on the need for considering them during the engineering design process to improve the quality of the products, the efficiency of the processes, and to minimise the related costs. Product Development It traditionally concerns what is generally named engineering design process, which is detailed in the next paragraph. Only some particular aspects are mentioned here, to highlight the many different items that could affect this phase. Product development is fundamental in determining the success of a product, and the decisions made in this phase affect profoundly both the product performance (Osteras et al. 2006) and the whole product lifecycle performance (Borg et al. 2000). In trying to define the efficiency and effectiveness of the product development, some concepts were explored by Duffy and O’Donnell (1997) to establish correct metrics to obtain performance values. Nevertheless, it is difficult to define the performance of the product development in a general way, because many different items must be considered and properly modelled, and their relationships must be analysed (Yassine et al. 2003). For example, issues that seem really important in the development of family of products (Alizon et al. 2007; Zhang et al. 2006) may be less decisive when developing flexible systems (Olewnik and Lewis 2006). This implies that sometimes the most reliable models for product development are not general but they are tailored to the development of specific classes of products. Interesting model-based methods to organise the different tasks in product development were proposed by Eppinger et al. (1994). Production and Delivery They concern the phases before the use of the product, from manufacturing to installation. Manufacturing processes actually generate the product and they may profoundly affect its characteristics, so that materials, processes and technologies must be defined both considering product functionality and process efficiency (Chan et al. 1998). Often, in fact, different alternatives must be compared in order to improve manufacturability (Gupta and Nau 1995; Xue 1997), never neglecting the problem of the related costs (Grewal and Choi 2005; Sharma and Gao 2007; Shehab and Abdalla 2001). Alike the manufacturing processes, assembly also determines product characteristics and costs and again it

1.2 The Engineering Design Process

3

must be approached systematically (Mantripragada and Whitney 1998), also defining rules for the generation of the assembly sequence (Lin et al. 2007). Techniques called Design for Manufacturing — DfM — and Design for Assembly — DfA — were developed and, because of their importance here, they will be detailed in a specific paragraph. Use The use of a product is, or should be, what mainly determines its characteristics, so that it is pointless underlining that this phase must be considered during the engineering design process (Pahl et al. 2007). Moreover, usability issues should be accounted too during the first stage of the product lifecycle (Nielsen 1993). It must also be considered that a product must guarantee some reliability and thus it may require some maintenance. Some research has been focused on developing reliability predictive models and on assessing reliability tests (Crowe and Feinberg 2001), particularly in the field of electric and electronic engineering (Mehlitz and Penix 2003). Nevertheless, the problems related to reliability now start to be considered in relationship to the whole product lifecycle (Yang 2007). Again, specific techniques have been developed: Design for Reliability — DfR — and Design for Test and Maintenance — DfTM. They will be described later, together with DfM and DfA. End of Life The end of life of a product implies that it is retired, disassembled, reused and/or recycled. The need for considering product disassembly, as well as the possibility of reuse, often implies some modifications in the product characteristics. All these issues are becoming really important, considering the increasing attention of society to the environment. The problem was first approached in terms of recycling and remanufacturing (Hundal 2000), then by developing the wider concept of eco-design (Houe and Grabot 2007). Again, specific Design for Environment — DfE — techniques were developed, also considering the evolving specific legislation.

1.2 The Engineering Design Process

1.2.1 Definitions and Concepts Focusing on the engineering design process and collecting the main characteristics in a basic definition, it may be described as a thoughtful process based on the laws and insights of science which generates the prerequisites for the realisation of products, processes, or systems, with specified constraints (Cross 2000; Dym and Little 2003; Pahl et al. 2007). As it is easy to imagine, the argument is extremely wide; an exhaustive survey on engineering design research was given by Horvàth (2004); here only a brief description of the aspects of interest is given.

4

1 State of the Art in the Field

In the literature there are several models to describe the engineering design process. These models have been classified as descriptive models, prescriptive models and integrative models (Birmingham et al. 1997; Cross 2000; Dym and Little 2003). Descriptive models, more or less detailed, describe only the sequence of the activities occurring in the design process, while prescriptive models propose algorithmic procedures to provide design methodologies. Integrative models focus on the iterative nature of the engineering design process, where the understanding of the problem and the development of the solution are achieved together. In order to define other effective models to describe the engineering design process, many authors focused attention on its different aspects, thus developing sequential design models, cyclical models, and, later, hybrid ones (Dwarakanath and Blessing 1996; Haik 2003; Hubka and Eder 1992; Otto and Wood 2000; Pahl et al. 2007; Pugh 1990; Reymen et al. 2006; Ullman 2002a; Ulrich and Eppinger 2003). Many studies have been focused on an empirical validation of these models (Baumgartner and Blessing 2001; Pahl et al. 1999), and some of them found that the engineering design models are applied rarely as prescribed (Maffin 1998). The engineering design process may also be described using process-based, task-based, and parameter-based models. Process-based models represent a flexible approach, providing a good top-level view of the design process and highlighting clearly the design goals. On the other hand, they often appear just philosophical and difficult to put into practice. Conversely, task-based models may precisely represent a specific design process, but they are not versatile. For each task, the specific goals must be defined in terms of optimisation of functions, minimisation of costs, ergonomic criteria, etc., while the identification of the critical goals is particularly important, especially in case of conflicts. Finally, parameter-based models tend to integrate the advantages of the previous ones, collecting the design tasks characterised by their input parameters. In this way, some flexibility in a task-based model is achieved; the limitation is that the parameters are often unknown at the time of the model development. The engineering design process may be further specified to highlight and differentiate better the activities that implement it. First of all, the origin; in case of a marketing analysis of the customer requirements rather than a specific customer order, the development will be very different. The same happens if an original design must be developed instead of an adaptive design or a variant one, for batch production or mass production, etc. (Eckert et al. 2004; Ullman 2002b). To characterise the engineering design process, the different branches involved — mechanical engineering, chemical engineering, transport engineering, software development, and so on — must be separated, as well as the complexity of the elements to be designed, being they plants, machines, assemblies or parts. The analysis of concepts and the development of conceptual design may prove to be helpful in organising and sharing this information (Alisantoso et al. 2005; Cartwright 1997; Kurakawa 2004). A lot of representation techniques were developed in order to structure the knowledge and enhance the implementation of the models describing the engineering design process. Some of them are PERT — Project Evaluation and

1.2 The Engineering Design Process

5

Review Technique (Meredith et al. 1985; Wiest and Levy 1977), Signal flow graphs (Eppinger et al. 1997; Isaksson et al. 2000), SADT/IDEF0 — Structured Analysis and Design Technique/Integration DEFinition 0 (Ross 1977; Qureshi et al. 1997), DSM — Design Structure Matrix (Steward 1981; Gebala and Eppinger 1991; Smith and Eppinger 1997), Petri Nets (Diaz and Azema 1985), SemanticNet Models (Blanchard and Fabrycy 1981) and Process Specification Languages (Qureshi et al. 1997).

1.2.2 The Problem Solving Process Now that some aspects concerning the models used to describe the engineering design process have been mentioned, attention may be focused on the main activity of the engineering design process, problem solving. To design means solving problems. Engineering design problems may be very different from each other; they have in common just a goal, some constraints and some criteria, which lead the definition of successful solutions. Engineering design problems are generally ill-structured and open-ended; there is no definitive formulation of the problem and any formulation of it may embody inconsistencies. Moreover, formulations of the problem are solution-dependent and the specified constraints are often conflicting with each other (Eide 2001; Howell 2001; Lumsdaine et al. 1999). Anyway, some basic actions in problem solving may be identified (Ullman 2002a), as shown in Fig. 1.2.

Fig. 1.2 The basic actions of problem solving

6

1 State of the Art in the Field

First, the problem to solve is recognised and a plan to solve it is established. The development of requirements and the proposal of existing solutions for similar problems help with understanding it. Alternative solutions are generated and evaluated, in terms of compliance to the requirements, and compared with each other. Decisions are made on acceptable solutions. It is interesting to underline that Fig. 1.2 shows the basic, linear process, but these actions are not necessarily taken consequentially, due to the iterative nature of the engineering design process. Throughout this process, it is very important that the results are communicated and discussed. As reported in Ullman’s work (Ullman 2002a), problem solving, i.e. making decisions, necessarily means managing a lot of different pieces of information distributed on different levels, and relating, interpreting and evaluating them. At the bottom level, corresponding to the simplest form of information, there are raw data, which means parameters or values for variables. Models, a form of information of increased value, are derived by the development of relationships among these data, and placed in the next level. Models may be just qualitative or developed by means of equations or empirical relationships, but in any case they are static. If the behaviour of the models is analysed and interpreted, then knowledge is achieved and this corresponds to the third level content. On the basis of knowledge, decisions can be made by judgment, corresponding to the highest level of information value. Decision analysis may be determined by different criteria, which can be considered to establish an integrated approach as by Belton and Stewart (2002), Bennett (1985) and Olewnik and Lewis (2005, 2006). Different types of knowledge are included in the chunks of information elaborated to make decisions: general knowledge, basic, not regarding a specific domain; domain specific knowledge, related to the form and function of specific objects or classes of objects; and procedural knowledge, based on general knowledge and/or on domain specific knowledge and leading to making decisions. Again, many methodologies and systems have been developed to formalise, elaborate and make knowledge useful and usable. They are based on different techniques and mathematical tools and may be focused on general or specific problems. Some researchers focused their attention on new approaches to manage knowledge in the engineering design process, for example by defining a method where both the product and the process models are modified synchronously, coupling relationships between them (Huang and Gu 2006). Others defined methods to express the knowledge in the design process by means of the flexible knowledge concept (Koh et al. 2005), while some proposed systems for an integrated engineering data management (Burr et al. 2005), or information systems to provide new design approaches (Houssin et al. 2006). The knowledge management related to specific manufacturing technologies was also studied, as in the case of hot forging (Kulon et al. 2006; Xuewen et al. 2005), die design (Tor et al. 2005), arc welding processes (Wang et al. 2005), injection moulding (Shehab and Abdalla 2002), powder metallurgy (Smith 2003), etc.

1.3 Concurrent Engineering

7

1.3 Concurrent Engineering The description of the engineering design process shows that several elements could heavily influence product functionalities and performance. Designers must define the characteristics of a product considering the different phases of its lifecycle, in order to generate a quality product, well suitable for the manufacturing process, easy to measure and assemble, etc. Thus, the contribution of the knowledge coming from marketing managers, manufacturers, inspectors, materials specialists, assembly managers, etc., is needed at the design stage, because designers’ decisions will affect the whole product lifecycle. Therefore, it is necessary to establish relationships among all these people from the beginning of the engineering design process, in order to share their knowledge in a collaborative environment. In the last few decades, the increasing complexity and variety of products, the many available technologies, and the specific characteristics of the market has led to profound changes in the structure of the design process. The classic Over-thewall process model cited in Ullman (2002a), also known as Waterfall design model in software engineering (Royce 1970), had a linear way, and the different competencies were kept separate. The designers received the communication of the perceived market requirements from the marketing people; after that, they developed the specifications of the product interpreting the request and communicated the result of their work to the production in terms of drawings, bills of materials, verification and assembly instructions, and so on. The information proceeded one-way, thrown over-the-wall, separating marketing from designers, designers from manufacturers, etc., as shown in Fig. 1.3. In this process model, the designer’s interpretation of the market requirements could be insufficient, and/or the choice of materials and technologies inadequate, in terms of availability or costs, thus resulting in poor quality products. This risk increased with the complexity and variety of market requirements, products, technologies, environments, etc., so that in the late 1970s Simultaneous Engineering concepts began to develop. They emphasised the need for considering the manufacturing process simultaneously with the development of the product, so that the communication proceeded from design to production and vice versa. Later on, this concept broadened, becoming Concurrent Engineering — CE, thus implying that an effective product development could be obtained only if the designers and the members owning the knowledge of the different phases of the product lifecycle interact, performing their activities together. In this way, the problems of the Over-the-wall design method became limited and the available resources were exploited as well as possible. The key points of CE may thus be identified in focusing on the whole product lifecycle, establishing engineering design teams, and developing and communicating information about the product and the related processes. The basic principles of Design Coordination may be found in Duffy et al. (1993, 1999), in a

8

1 State of the Art in the Field

framework where different models concerning product development, plans, resources, tasks, etc., are used, and their interactions studied.

Fig. 1.3 The Over-the-wall design method

The link between CE and product lifecycle was studied focusing on different aspects; for example, Fleischer and Liker (1997) evaluated the effectiveness of CE in integrating product development across the organisation, while Seif (1998) defined some problem-solution schemes to optimise product design based on a CE approach. A collaborative environment also favours the evolution of the knowledge in the design teams (Wu and Duffy 2002), enhancing the development of systems for data integration (Kleiner et al. 2003) and knowledge sharing (Vergeest and Horváth 1999). Specifically, a CE approach was used to model products using multidisciplinary collaborative design (Liang and Guodong 2006), or to integrate product and process modelling (Nahm and Ishikawa 2004). Others analysed the problem of modelling dependencies in CE processes involving numerous tasks (Park and Cutkosky 1999), or evaluated the design alternatives in the collaborative development of products (Hung et al. 2007). Some other works developed specific systems for CE environments to support dynamic design reasoning (Chiang et al. 2006), or defining a working situation model, where the working situation, its characterising elements, and their relative concepts are defined and integrated (Houssin et al. 2006). Collaborative systems to approach the conceptual design of industrial systems (Veeke et al. 2006) or for selecting materials, processes, and apparatuses (Chan et al. 1998) were developed too.

1.4 Standards

1.4.1 Fundamentals All the phases of the product lifecycle should comply with standards, from those defining the rules for technical drawings to those regulating the recycling and

1.4 Standards

9

reuse. In the last years of the past century, a technical committee was constituted, the ISO TC 213 — International Organization for Standardization Technical Committee 213 — to develop and harmonise standards in the scope of Geometrical Product Specifications — GPS. What was particularly interesting and new in the ISO TC 213 vision was the effort to consider and develop the standards as an “improved engineering tool”. In detail, this vision that summarises the goal of ISO TC 213 runs as follows: “This integrated GPS system for specification and verification of workpiece geometry is an improved engineering tool for product development and manufacturing. This GPS system is necessary, as companies are rapidly moving ahead with new technologies, new manufacturing processes, new materials and visionary products in an environment of international outsourcing” (ISO/TC 213 N355 Annex 1). Even if now the attention of the ISO GPS is focused on establishing links between the specification and verification phases, there is a will to consider the phase of manufacturing too, in the near future, to improve the efficiency and effectiveness of the whole process. This philosophy fits at best what has been described up to now, and it has been a basic item in the development of the DGLsCF, as explained in the next chapter. As described in the ISO Technical Report ISO/TR 14638 — GPS Masterplan, all ISO GPS standards are collected in the General GPS Matrix, consisting of four groups of standards: Fundamental GPS standards, Global GPS standards, General GPS standards and Complementary GPS standards (ISO/TR 14638:1995). Fundamental GPS standards establish the fundamental rules and procedures for the characterisation of products; Global GPS standards cover or influence several or all of the so-called chains of standards, described in the following; General GPS standards are the main body of GPS standards, establishing both rules for drawing indications/definitions and verifications principles for different types of geometrical characteristics. Finally, Complementary GPS standards establish complementary rules for specialised categories of features or elements. In the General GPS Matrix, standards are organised into chains of standards (Bennich 1994). Each chain comprehends all the standards related to the geometrical characteristics of a feature, from those dealing with the drawing indication of the product characteristics to those concerning the functional requirements of the measurement equipment used to verify them. Table 1.1 shows an extract of the General GPS Matrix, where each row represents a geometrical characteristic of a feature with the related chain of standards. The concept of chain of standards demonstrates well the linking of all the standards for the geometrical specification and verification of a feature. Features, in fact, define the product geometry and their characteristics are described in some Global GPS standards, mainly ISO 14660-1, ISO 14660-2, ISO 17450-1, and ISO 17450-2. The importance of the ISO GPS features in the scope of this research requires a specific paragraph as follows.

10

1 State of the Art in the Field

Table 1.1 Extract from the General GPS Matrix

1.4.2 Highlights on the ISO GPS Features As described in ISO 14660-1, geometrical features exist in three worlds: the world of specification, the world of workpiece — physical world, and the world of inspection. The same feature can differ substantially in the three worlds; the manufacturing processes, environmental conditions and other aspects can affect the geometry of physical workpieces, so that they may become different from the nominal components specified at the design stage. Designers are conscious of this, and they must consider the manufacturing and environmental conditions when they dimension the component, to guarantee the product functionalities. To do this they use dimensional and geometrical tolerances. Thus, the physical workpiece must be measured to verify the conformity to the specified tolerances, and the verification process gives a true representation of the physical workpiece to be compared with the designed one. Coherently with the ISO TC 213 vision, ISO GPS standards aim at determining univocally the different representations of the geometrical features in the three worlds and at establishing relationships among them. The example shown in Fig. 1.4, where a simple cylinder is considered, can help with the comprehension of these concepts. In the world of specification, the cylinder is described by a nominal integral feature — its theoretically exact surface — and a nominal derived feature — its theoretically exact axis.

1.4 Standards

11

Fig. 1.4 Example of relationships between geometrical features

In the world of workpiece — the physical world, the workpiece is represented by its real feature, a quasi-cylinder, different from the nominal one. In the world of inspection, the acquirement of a finite number of points from the real surface generates an approximated representation of the real feature, called extracted integral feature. A sort of axis of the real feature representation — an extracted derived feature — may be related to this approximated representation. The final step consists of associating the extracted integral feature with an integral feature of perfect form. This approach allows one to recognise univocally the features in the different worlds and to determine their relationships.

Fig. 1.5 Procedures and operations to establish biunique relationships among the features in the different worlds

12

1 State of the Art in the Field

The biunique relationships among the features in the different worlds are ensured by the use of the same operations in the world of specification and in the world of inspection. The ISO/TS 17450-2 defines an operation as a “specific tool required to obtain features or values of characteristics, their nominal value and their limits”. Figure 1.5 shows this correspondence. For example, the operation “Partition” is used to identify bounded features, both in the nominal model and in the real workpiece. Figure 1.6 shows this concept graphically.

Fig. 1.6 Partition operation in the nominal model — specification, and in the real workpiece — verification

Now, if the same operations are recognisable in the specification and in the verification procedures, the correspondence called the duality principle is realised. The designers define the functional requirements taking into consideration that they must be verified; for this reason, they communicate precisely these requirements to the inspectors. In this way, the inspectors can measure exactly what the designers want, and how they want it to be measured. In this context, datums and datum systems are defined (ISO 5459:1981); they are the elements used to define clearly and measure the relative position and orientation of the features in the parts. The concept of feature of size is also introduced here because of its importance for the DGLs-CF project. ISO 14660-1 defines the feature of size as a “geometrical shape defined by a linear or angular dimension which is a size”, and note 1 clarifies “the features of size can be a cylinder, a sphere, two parallel opposite surfaces, a cone or a wedge”. Features of size are important because they are implicitly associated to dimensions so that they can be immediately related to the dimensional characteristics of parts. Therefore, features of size may characterise products both from the geometrical and the dimensional point of view. Nonetheless, it might be worthwhile to underline that the use of features of size in the standards ASME — American Society of Mechanical Engineers — dates quite far back in time — the Taylor principle dated 1901 is based on them, while their use was less common in ISO standards before the ISO GPS. Finally, the topological relationships among ISO GPS features are defined using situation features — “feature of type point, straight line, plane or helix which allow defining the location and/or the orientation of a feature” — and situation characteristics — “characteristics defining the relative location or

1.5 Design Methods

13

orientation between two situation features” (ISO 14660-1:1999; ISO 146602:1999). This description of the ISO GPS features reports only the basic concepts, while many other important issues exist, such as the specification of the conventions on how to obtain an extracted feature from the real one, the way to associate the integral feature of perfect form to the extracted one, the problems related to the measurement procedures, the filtration of the acquired data, etc. Some of them are defined in other standards and will be recalled when necessary; some others are still objects of study. Even if this topic is gaining considerable interest, for now the main researchers working on it are those involved in the development of the standards selves. In particular, the work of Choley et al. (2007), Désenfant and Priel (2006), Dovmark (2001), Nielsen (2003, 2006) and Srinivasan (2001, 2008) may be of great help in understanding better the ISO GPS principles.

1.5 Design Methods Considering the increasing complexity of products, processes and systems, and in order to improve the effectiveness of the engineering design process, many design methods have been developed, focused on one or more specific problem solving activities. Some general characteristics of design methods may be pointed out; the description in the following reports those considered of interest in the DGLs-CF project. Regarding the main characteristics required to design methods, they must be helpful in determining optimum solutions, inciting creativity and understanding; they must be compatible with concepts and methods deriving from other disciplines; they must tend to exploit existing solutions; they must allow the electronic formalisation and elaboration of information; and, finally, they must provide an effective communication of the results. Although in accordance with the aforementioned characteristics, design methods may be very different from each other. They may consist in completely new rational procedures, adaptation of procedures deriving from other disciplines, or formalisations of informal ways of working; they may have different goals and be related to different stages of the engineering design process. As a broad classification, design methods may be divided into creative methods and rational methods. Creative methods, as one would expect, tend to stimulate creativity by freely sharing ideas coming from different backgrounds. Rational methods, those most commonly intended as design methods, tend to systematise the approach to the design, not necessarily restraining creativity. Some rational methods try to clarify objectives by identifying sub-objectives and relationships; others focus on establishing functions, on specifying performances, on determining characteristics to satisfy customer requirements, on generating and evaluating alternatives, on improving details, etc. (Cross 2000; Pahl et al. 2007).

14

1 State of the Art in the Field

Many design methods have been developed, characterised by a different approach to the several aspects of the engineering design process; among them, some are particularly widespread and are briefly described hereafter. At the conceptual design stage of the engineering design process, the focus is on the general and essential aspects of the products and the approach must necessarily be abstract. This approach is called functional decomposition modelling. The overall function of the product is identified on the basis of the main requirements, and is expressed in terms of flow of energy, materials and information. The overall function can then be defined by sub-functions of lower complexity, again expressed in terms of flows, properly structured and related to each other. In this way, the problem is formulated to an abstract, general level, which allows identifying the optimum solutions far from the influence of conventional or predefined ideas (Cross 2000; Pahl et al. 2007; Sturges et al. 1993; Ullman 2002a). Functional decomposition modelling has also been studied, focusing on some particular aspects of design processes; among them, there are design reuse (Khadilkar and Stauffer 1996; Xu et al. 2006), redesign (Hirtz et al. 2002), and product family design (Zhang et al. 2006). Functions and sub-functions are controlled by parameters; they must be identified and put into relationship. These parameters represent all the factors on which the product and the processes depend and contribute defining the product/process characteristics in terms of geometry, materials properties, etc. Parameter management has also been approached by defining different models for the engineering design process, as a parameter-driven task-based model (Clarkson and Hamilton 2000), or a model for the refinement of the design parameters (Lee and Lee 2004). Many authors have studied the definition of parameters in the component modelling, for example in parallel with the development of CAD — Computer Aided Design — systems. Among them, Abdel-Malek et al. (1999), Hosaka (1992), McMahon and Browne (1998), Salomons et al. (1993) and Shah and Mäntylä (1995). The axiomatic design (Suh 1990, 1998, 1999, 2001, 2005) describes the design process by establishing relationships between the Functional Requirements — FRs — in the functional domain, and the Design Parameters — DPs — in the physical domain. FRs and DPs are mathematically represented by vectors, which are related through matrices. The whole theory develops axioms, corollaries and theorems to describe these relationships. The axiomatic design can also be combined with a creative theory named TRIZ (Kim and Cochrani 2000) or with methods such as Design for Six-Sigma (El-Haik 2005) and Design of Experiments (Engelhardt 2000). TRIZ (Altshuller 1988) is the Russian acronym for the Theory of Inventive Problem Solving (Teoriya Resheniya Izobretatelskikh Zadatch). This theory starts from the consideration that many engineering problems have already been solved, partially or totally, even if often in completely different contexts. Altshuller’s method is based on contradictions and inventive principles. Contradictions occur when the fulfilling of a requirement is disadvantageous to other requirements; TRIZ helps with finding the major contradictions and generates the ideas for

1.6 DfX Methods

15

overcoming them, using 40 inventive principles (Altshuller et al. 2005) derived from the analysis of more than 40,000 patents. Design for Six Sigma, also known as Robust Design or Taguchi method (Taguchi 1999, 2004), concerns one of the most important tasks in the engineering design process, the research of the optimal values to be associated with the parameters in order to optimise the whole product/process characteristics and performances (Park 1996; Phadke 1989; Ross 1996; Wu and Wu 2000). In fact, parameter values are often set without considering tolerances and noises, so that in the detail design phase tight tolerances must be specified to obtain the requested quality. Robustness strategies substantially define parameter values so that they result in less sensitivity to the causes of variation To obtain reliable and robust parameter values, experiments must be precisely planned and defined, as in the design method named Design of Experiments — DOE (Antony 2003; Barrentine 1999; Montgomery 2004; Roy 2001). DOE techniques define the control variables and how they must be collected and analysed to obtain the most reliable robust parameter values. These techniques may be applied to different aspects of the design processes, for example relating the robust design parameters to the customer satisfaction (Jain and Sobek 2006). Optimal value setting is generally considered in the step of detailed design, but interesting studies propose methods to evaluate robust parameter values already at the conceptual design phase (Chen et al. 1996).

1.6 DfX Methods

1.6.1 Overview As said at the beginning of this chapter, design methods become DfX methods when they are focused on particular aspects of the product lifecycle. The role of the DfX methods is schematically represented in Fig. 1.7. The information coming from the several phases of the product lifecycle is used by the DfX methods in the engineering design process. Several DfX methods have been developed, focusing on different phases of the product lifecycle. Some of them are particularly relevant in the DGLs-CF project, so they are detailed hereafter.

1.6.2 Design for Manufacturing — DfM Design for Manufacturing — DfM — defines the characteristics of a product to allow an efficient, high-quality manufacture. The goal also implies the

16

1 State of the Art in the Field

minimisation of costs and times. In principle, many manufacturing processes may be chosen for any product, and each manufacturing process shows different characteristics which must be considered during the design (Gupta et al. 1997). DfM methods have been extensively studied by many authors; among them, some analysed DfM methods in their completeness (Andersen 2004; Bralla 1998; Geng 2004; Poli 2001; Shah and Wright 2000), while others focused particularly on specific aspects, such as, for example, the representation and modelling of the manufacturing knowledge (Kulon et al. 2006; La Trobe-Bateman and Wild 2003; Zhao and Shah 2004).

Fig. 1.7 Role of the DfX methods in the product lifecycle

DfM methods and tools may be very different, depending on what aspects of product development they focus on. The type of product, the small or large batch, the continuous or discrete manufacturing process, the kind and amount of related information, and many other issues determine this difference. If the focus is on the product, then shape and geometry constraints are predominantly evaluated, while if the focus is on the process, process complexity, reliability, and tools accessibility are considered. However, some common principles may be identified; for example, the maximisation of the use of standard components, tools and materials (while also considering new processes and materials), the minimisation of the geometrical complexity, loosing tolerances, etc. The methods and tools developed for DfM generally include an evaluation in terms of technological feasibility, and/or economic feasibility, and/or trade-off optimisation (Shah and Wright 2000). The different approaches for determining technological feasibility may be grouped into two classes: variant approaches and generative approaches. Briefly, variant approaches are based on past history, which means that historical data

1.6 DfX Methods

17

concerning production processes and product characteristics are grouped and organised, so that it is easy to find the closest matching production process for a new component (Hyde 1981; Maher et al. 1995). Generative approaches are based on first principles, concerning for example the geometry of the products or physical, kinematic and mechanic issues of the process. Generative approaches may be process-based if analytical models of manufacturing process are developed (Gupta et al. 1997), or feature-based when features are defined as types with which a set of feasible processes is associated (Zhao and Shah 2004). In determining economic feasibility, different models have been developed for estimating manufacturing costs; these models may be more or less specific for different manufacturing processes (Hundal 1993). The cost of a new component can be evaluated in comparison with that of a similar component present in a database, and then adjusted considering materials, batch size, geometric complexity, manufacturing processes, and so on (Grewal and Choi 2005). Otherwise, specific manufacturing processes with known costs can be targeted, so that the cost estimation can be very accurate but less flexible (Chang et al. 1999). Domain independent analyses were also proposed by Zhao and Shah (2002). Finally, in determining an optimisation based on tradeoffs, design is not considered as fixed; design performance and manufacturing cost are studied in order to increase the value of the product. Multi-disciplinary optimisation (Capasso and Périaux 2005), Genetic algorithms (Goldberg 1989), Decision analysis (Figuera et al. 2005; Edwards and Von Winterfeldt 2007), and Quality Function Deployment (Akao 1990) are some of the methods within this area. In order to highlight the importance of the DfM methods in the context of the DGLs-CF project described in this book, it can be pointed out that the first release of the Design GuideLines — DGLs, the methodological approach to the generation of design guidelines that has been the starting point for the research, was developed as a DfM, focused on determining the technological feasibility of products with respect to Rapid Prototyping technologies. The approach was essentially a variant one, aimed at defining design rules on the basis of a knowledge base containing the characteristics of components and technologies. A rough evaluation of costs was also proposed. Instead, the top-down approach of the DGLs-CF, it is now more similar to a generative one, because it allows determining the technological feasibility of the products exploiting the definition of classes of features. Different DfM methods may be applied in different contexts, and often manufacturing issues are considered together with assembly ones (Boothroyd et al. 2002; Molloy et al. 1998). In Design for Manufacturing and Assembly — DfMA — the manufacturability of components is evaluated once they are defined by Design for Assembly — DfA — methods, as shown in Fig. 1.8. For this reason, some basic elements of DfA are discussed in the following.

18

1 State of the Art in the Field

Fig. 1.8 Main steps of Design for Manufacturing and Assembly — DfMA

1.6.3 Design for Assembly — DfA Design for Assembly — DfA — methods concern the ease of connecting parts, components or subsystems, to constitute the final product. Quality and costs are strongly affected by the assembly process; this is why its characteristics must be considered in the early design process stages. Clearly, the number of components to assemble increases the assembly times and costs, as well as the number and kind of fasteners; for this reason, the first step in DfA is always the minimisation of the number of components. Each pair of interfacing components is evaluated to establish whether it can be substituted with a single component (Boothroyd et al. 2002; Ullman 2002a). However, the time needed for the assembly process does not depend only on the number of components, so that the efficiency of all the steps in the assembly process must be studied and optimised. The assembly process consists in retrieving components from storage, handling them to establish their relative orientation, and mating them. The geometry of the single parts profoundly affects the efficiency of all these steps and specific rules have been developed to help the design of components, considering characteristics as symmetry, sharpness, thickness, roundness, flexibility, etc. (Andreasen 1988; Boothroyd et al. 2002; Molloy et al. 1998). Data structures have been established to define the functions representing the linked geometry of parts, in order to create an assembly design environment for top-down design, from function to geometry (Gui and Mäntylä 1994; Shah and Rogers 1993).

1.6 DfX Methods

19

Another fundamental aspect determining the efficiency of the assembly process is the assembly sequence. Different theories have been developed to define it, from establishing constraint-consistent assembly sequences through connective data models (Mantripragada and Whitney 1998; Whitney et al. 1999), to the use of contact relation matrices (Lin et al. 2007) or a variant of Petri nets (Wu and O’Grady 1999). A lot of work has been done to integrate assembly information and concepts in CAD systems, as for example in OpenADE — Open Assembly Design Environment (Lyons et al. 1999). Regarding the exploitation of the DfA concepts in the DGLs-CF project, assembly issues have not been considered until now, but the way for organising and deriving the knowledge is clear, well defined and open, so that a further implementation in this sense could be easily approached, as recalled in Chap. 5. The formalisation of features, for example, which is now focused on the technological feasibility and measurability, could consider assembly issues as well. The need for the accessibility to the components should be considered both for assembly and for maintenance and disassembly, as will be pointed out in the following paragraphs regarding the DfTM.

1.6.4 Design for Test and Maintenance — DfTM — and Design for Verification — DfV Design for Test and Maintenance — DfTM, in some cases also called Design for Verification — DfV, deals with the possibility of measuring the performance of the critical functions of a product and maintaining them under given conditions. In DfTM it is necessary first to identify the components associated to critical functions because they must be easily accessed and maintained. This implies that measurement tools should be considered and defined at the beginning of the design phase, in order to provide adequate accessibility. It is also necessary to define preventive maintenance procedures establishing the frequency and modes of check, and provide an adequate stock of critical parts, so as to minimise downtime in case of failure. Design for Verification was developed particularly in the electric and electronic engineering field, where it is often called D4V (Gamma et al. 1995; Lam 2005; Riel 1996; Smith 1996). Plenty of work has been done developing models for test and maintenance of software; for example, models for checking programs were defined (Visser et al. 2000), while semantic models for object-oriented software were generated to organise all the information during the software development lifecycle, with particular regard to testing and maintenance (Deng et al. 2004; Meyer 1997). The attention was also focused specifically on the reliability of components in software development (Fischer 2000; Mehlitz and Penix 2003).

20

1 State of the Art in the Field

In the context of the research described in this book, verification issues are approached in a quite different way. First of all, there is particular concern with the physical representation of products, so that the verification concepts are analysed regarding the measurement methods and tools used to evaluate the conformity of the product to the requirements. Moreover, the integration of the ISO GPS principles gives new and particular characteristics to the proposed approach to the development of design guidelines regarding the verification issues. Maintainability and reparability concepts are also related to product reliability, which is analysed by DfR. However, they are also part of disassembly issues, which are analysed by DfE, together with the aspects related to reuse or recycling of products at the end of their lifecycle. All these topics are presented in the following paragraphs.

1.6.5 Design for Reliability — DfR Design for Reliability — DfR — considers how the quality of a product, in terms of performance under stated conditions of use and maintenance, is maintained over time. Reliability is ensured if the parameters that could determine a failure are identified and monitored. In this sense, DfR is close to DfTM; the maintainability concept, in fact, can be associated with the probability of repairing a component before it fails, if appropriate procedures are defined and performed. Exhaustive studies may be found in the work of Crowe and Feinberg (2001), Dhillon (1999), Wassermann (2002), Crespo Márquez (2007), or Ebeling (1997), and Thompson (1999), where reliability and maintainability are considered together in the design process. Reliability is defined by means of failures, which are intended as unsatisfactory performances, and precisely by means of the Mean Time Between Failures — MTBF — indicator, or by its inverse function, the failure rate (Finkelstein 2008). The most widespread technique for identifying potential failures is called Failure Modes and Effects Analysis — FMEA, which may be of great help in DfR (Stamatis 1995). During the product development process, the possibility of failure must be evaluated for each function, also identifying the possible effects on the parts of the product. The cause of the failure must then be identified — design or manufacturing errors, changes due to operational conditions, etc. — so that corrective actions may be defined. These could be redesign actions, manufacturing process changes and operational conditions changes, aimed at minimising the effect of the failure. It is very important to understand how the different parts of a product affect reliability. If parts may be considered as working in series, the failure of a part will mean the failure of the whole product. In this case, the reliability of the product will be given by the product of the reliabilities of its parts. In redundant

1.6 DfX Methods

21

systems, the parts may be considered as working in parallel, so that the failure of a part does not imply the failure of the whole product. For critical parts, redundancy is often considered a good way to improve the reliability; however, this increases costs and might add other failure modes. In any case, a good estimation of the reliability implies necessarily a deep knowledge of failure modes, which may be obtained through experiments, analysis of historical data, use of models of physical phenomena, and information sharing among experienced specialists in the different fields in a collaborative environment. The DGLs-CF project has not considered reliability issues until now, but this information could be further introduced, for example by developing procedures to highlight the most critical product features and/or process characteristics, as discussed again in Chap. 5.

1.6.6 Design for Environment — DfE Design for Environment — DfE — concerns all the issues of the product lifecycle which may affect the environment, such as product manufacturing, distribution, use, retire, reuse or recycle (Bras 2006; Graedel and Allenby 1996). For each of them, the related information must be effectively managed and organised (Bras 1997). Different approaches have been developed to reduce the environmental impact, and these are known as sustainable development, industrial ecology, green engineering, eco-design, etc. (Brezet and Van Hemel 1997; Coulter et al. 1995). First of all, it is necessary to evaluate the impact on the environment of the materials used, both in terms of raw materials and of the energy needed to produce, recycle, ignite and waste them. Different parameters have been defined to evaluate the environmental risk; one of them is the MET — Materials, Energy, Toxicity — score. About materials, the exhaustion of scarce materials is evaluated, while greenhouse effect, acidification, smog, and eutrophication are considered with respect to the energy associated with all the processes in the product lifecycle. These processes are also evaluated in terms of human toxicity, ecotoxicity and ozone depletion. An interesting activity-based method was proposed to analyse and trace the cost due to energy consumption and waste generation in the lifecycle design (Emblemsvåg and Bras 1997). Products should be designed in order to ease disassembly, so that components with different characteristics can be separated at the end of the product lifecycle (Dowie 1994; Hundal 2000). It could also be highlighted that disassembly and material separation techniques may considerably differ, depending on the kind of products, be they vehicles (Coulter et al. 1996), electronic equipment (Campbell and Asad 2003), or others. The cost of disassembly must be considered too. Components must be designed so that they can be mostly reused or recycled; when they cannot, their degradability must be evaluated. To optimise the end of life value of a product, an

22

1 State of the Art in the Field

algorithmic approach for DfE was proposed by Pnueli and Zussman (1997); it identifies weak spots in the product design and proposes possible solutions. The need for considering how a product may impact on the environment while still at the beginning of the design process is underlined by the development of a specific set of standards in the ISO 14000 series. These standards concern the reduction in use of raw materials, resources and energy, the improvement in the efficiency of processes, the reduction in the quantity of waste and the enhancement in the use of recycling. Specifically, ISO 14062 — Design for environment — aims at the improvement of environmental performance of products (ISO/TR 14062:2002). Standards, however, express the requirements in a textual form, and they are often difficult to integrate in the engineering design process. A tool was developed to translate these textual requirements into constraints, in order to verify the compliance of products with the standards (Houe and Grabot 2007). It is not difficult to foresee that DfE concepts could be integrated in the research described in the rest of this book. Thanks to the open structure already mentioned when considering the possible introduction of DfA or DfR concepts, the environmental issues could be introduced as well. For example, product feature formalisation could contain some information for an easy disassembly, and the definition of technological characteristics could include a sort of MET score, as discussed again in Chap. 5. All the DfX methods described here aim at linking the product development process with the whole product lifecycle. Actually, the ideal product should be designed considering all the DfX methods, thus linking the development process with all the phases of the product lifecycle, in a real implementation of the concurrent engineering vision. This is not easy, nor always possible; for example, Design for Assembly and Design for Environment, in its disassembly issues, could imply conflicting activities. Anyway, this is an interesting challenge, leading to the development of design for multi-X methods and tools, where the knowledge of the different domains is integrated and organized (Borg et al. 2000). The next chapters will highlight that the result of the DGLs-CF project may be considered as a design for multi-X method, given that it develops the design process considering both the manufacturing and verification processes; moreover, its structure looks particularly suited to enlarge further its application to the other phases of the product lifecycle.

Summary This chapter has described the state of the art of the scenario where the DGLs-CF project takes place. It opened highlighting that the engineering design process must be considered inside the product lifecycle since its first beginning. Moreover, the effectiveness of the engineering design process is enhanced if it is developed in a collaborative environment, also considering the related standards. In this

References

23

context, the focus is mainly on the design methods, which are developed to formalise, elaborate and make usable all the pieces of information generated and used in this process. Since the DGLs-CF can be regarded as a design for multi-X method, the main characteristics of the DfX methods have been described, so that it will be easier to highlight the specific characteristics of the DGLs-CF in the next chapters.

References Abdel-Malek K, Zou H-L, Wang J-Y, Othman S (1999) Automated Design and Parametrization of Mechanical Part Geometry. Res Eng Des 11:206-217 Akao Y (1990) Quality Function Deployment: Integrating Customer Requirements into Product Design. Productivity Press Alisantoso D, Khoo LP, Ivan Lee BH, Fok SC (2005) A rough set approach to design concept analysis in a design chain. Int J Adv Manuf Technol 26:427-435 Alizon F, Khadke K, Thevenot HJ, Gershenson JK, Marion TJ, Shooter SB, Simpson TW (2007) Frameworks for Product Family Design and Development. Concurr Eng 15(2):187-199 Altshuller GS (1988) Creativity as an Exact Science. Gordon and Breach, New York Altshuller GS, Clarke DW, Shulyak L, Lerner L (2005) 40 Principles: TRIZ Keys to Innovation [Extended Edition]. Technical Innovation Center, Inc. Andersen DM (2004) Design for manufacturability. CIM Press Andreasen M (1988) Design for Assembly. Springer Verlag, New York Antony J (2003) Design of Experiments for Engineers and Scientists. Butterworth-Heinemann Barrentine LB (1999) An Introduction to Design of Experiments: A Simplified Approach. ASQ Quality Press Baumgartner C and Blessing L (2001) Establishing causal links between success characteristics and project outcome. In: Proceedings of the international conference on engineering design, Glasgow Belton V and Stewart TJ (2002) Multiple criteria decision analysis - an integrated approach. Kluwer Academic Publishers Bennett PG (1985) On Linking Approaches to Decision-Aiding: Issues and Prospects. J Oper Res Soc 36(8):659-669 Bennich P (1994) Chains of Standards - A New Concept in GPS Standards. Manuf Rev 7(1):2938 Birmingham R, Cleland G, Driver R, Maffin D (1997) Understanding engineering design. Prentice-Hall, London Blanchard GS, Fabrycky WJ (1981) Systems Engineering and Analysis. Prentice-Hall, New York Boothroyd G, Dewhurst P, Knight WA (2002) Product design for Manufacture & Assembly. Marcel Dekker Borg JC, Yan X, Juster NP (2000) Exploring decisions’ influence on life-cycle performance to aid “design for Multi-X”. AIEDAM 14:91-113 Bralla JG (1998) Design for Manufacturability Handbook. Mc Graw- Hill Bras B (1997) Incorporating Environmental Issues in Product Design and Realization. Ind Environ 20:1-2 Bras B (2006) Design for Environment, Design for Recycling, Part 1-2. Georgia Institute of Technology, Lecture Notes. http://www.srl.gatech.edu/education/ME4171/. Accessed 24 April 2009

24

1 State of the Art in the Field

Brezet H and Van Hemel C (1997) Ecodesign - A Promising Approach to Sustainable Production and Consumption. United Nations Publication. Burr H, Vielhaber M, Deubel T, Weber C, Haasis S (2005) CAx/engineering data management integration: enabler for methodical benefits in the design process. J Eng Des 16(4):385-398 Campbell MI and Asad H (2003) Design Evaluation Method for the Disassembly of Electronic Equipment. In: Proceedings International Conference on Engineering Design - ICED 03 Capasso and Périaux (ed) (2005) Multidisciplinary Methods for Analysis Optimization and Control of Complex Systems. Springer, Berlin Heidelberg Cartwright AJ (1997) Interactive Prototyping - a Challenge for Computer Based Design. Res Eng Des 7:10-19 Chan K, King C, Wright P (1998) COMPASS: Computer Oriented Materials, Processes, and Apparatus Selection System. J Manuf Syst 17(4):275-286 Chang K-H-O, Silva J, Bryant I (1999) Concurrent Design and Manufacturing for Mechanical Systems. Concurr Eng Res Appl 7(4):290-308 Chen W, Allen JK, Mavris DN, Mistree F (1996) A concept exploration method for determining robust top-level specifications. Eng Optim 26(2):137-158 Chiang TA, Trappey AJC, Ku CC (2006) Using a knowledge-based intelligent system to support dynamic design reasoning for a collaborative design community. Int J Adv Manuf Technol 31(5-6):421-433 Choley JY, Riviere A, Clement A, Bourdet P (2007) A New Variational Association Process for the Verification of Geometrical Specifications. Transactions of the ASME J Comp Inf Sci Eng 7:66-71 Clarkson PJ and Hamilton JR (2000) ‘Signposting’, A Parameter-driven Task-based Model of the Design Process. Res Eng Des 12:18-38 Coulter S, Bras B, Foley C (1995) A lexicon of green engineering terms, In: Proceedings ICED’95 Coulter S, Bras R, Winslow G, Yester S (1996) Designing for material separation: lessons from automotive recycling. In: Proceedings of DETC’96, 96-DETC/DFM-1270 Crespo Márquez A (2007) The Maintenance Management Framework Models and Methods for Complex Systems Maintenance. Springer Series in Reliability Engineering Cross N (2000) Engineering Design Methods: Strategies for Product Design. Wiley & Sons Crowe D and Feinberg A (ed) (2001) Design for Reliability. CRC press Deng D, Sheu PC-Y, Wang T, Onoma AK (2004) Model-based Testing and Maintenance. In: Proceedings of the IEEE Sixth International Symposium on Multimedia Software Engineering (ISMSE’04):1-8 Désenfant M, Priel M (2006) Road map for measurement uncertainty evaluation. Meas 39:841848 Dhillon BS (1999) Design Reliability: Fundamentals and Applications. CRC press Diaz M, Azema P (1985) Petri net based models for the specification and validation of protocols. In: Lecture Notes in Computer Science 188, Springer-Verlag, New York Dovmark J (2001) New interesting concepts from ISO/TC 213. In: Proceedings of INTERSEC Dowie T (1994) Design for Disassembly. Technical Report, Manchester Metropolitan University. Duffy AHB and O’Donnell FJ (1997) A Model of Product Development Performance. In: Proceedings Darmstadt Symposium on Designers - The Key to Successful Product Development, Darmstadt, Germany, ISBN 1-85233-031-7: 269-283 Duffy AHB, Andreasen MM, MacCallum KJ, Reijers NL (1993) Design coordination for concurrent engineering. Int J Eng Des 4:251-265 Duffy AHB, Andreasen MM, O’Donnell FJ (1999) Design Co-ordination. In: Proceedings International Conference on Engineering Design (ICED'99), Munich, Germany, (1):113-118 Dwarakanath S and Blessing L (1996) Ingredients of the Design Process. In: Cross N, Christiaans H and Dorst K (ed) Analysing Design Activity, Wiley, Chichester Dym CL and Little P (2003) Engineering Design: A Project-Based Introduction. Wiley and Sons

References

25

Ebeling CE (1997) An Introduction to Reliability and Maintainability Engineering. McGraw-Hill Companies Inc, Boston Eckert C, Clarkson PJ, Zanker W (2004) Change and customisation in complex engineering domains. Res Eng Des 15:1-21 Edwards W and Von Winterefeldt D (ed) (2007) Advances in Decision Analysis: From Foundations to Applications. Cambridge University Press Eide AR (2001) Engineering fundamentals and problem solving. Mac Graw Hill El-Haik B (2005) Axiomatic Quality: Integrating Axiomatic Design with Six-Sigma, Reliability, and Quality Engineering. Wiley-Interscience Emblemsvåg J and Bras B (1997) An activity-based life-cycle assessment method. In: Proceedings of DETC’97, DFM-97-119 Engelhardt F (2000) Improving Systems by Combining Axiomatic Design, Quality Control Tools and Designed Experiments. Res Eng Des 12:204-219 Eppinger SD, Whitney DE, Smith RP, Gebala DA (1994) A model-based method for organizing tasks in product development. Res Eng Des 6(1):1-13 Eppinger S, Nukala M, Whitney D (1997) Generalized models of design iteration using signal flow graphs. Res Eng Des 9:112-123 Figuera J, Greco S, Matthias E (ed) (2005) Multiple Criteria Decision Analysis: State of the Art Surveys. Springer Finkelstein M (2008) Failure Rate Modeling for Reliability and Risk. Springer Series in Reliability Engineering Fischer B (2000) Specification-Based Browsing of Software Component Libraries. J Autom Softw Eng 7(2):179-200 Fleischer M and Liker JK (1997) Concurrent Engineering Effectiveness: Integrating Product Development Across the Organizations. Hanser Gardner, Cincinnati, OH Gamma E, Helm R, Johnson R, Vlissides J (1995) Design Patterns Elements of Reusable ObjectOriented Software. Addison Wesley Gebala DA and Eppinger S (1991) Methods for analyzing design procedures. In: L. Stauffers LA (ed) Design Theory and Methodology. ASME, Miami Geng H (2004) Manufacturing Engineering Handbook. Mc Graw Hill Handbooks Goldberg DE (1989) Genetic Algorithms in Search, Optimization, and Machine Learning. Addison-Wesley Professional Graedel TE and Allenby BR (1996) Design for Environment. Prentice Hall Grewal S and Choi CK (2005) An Integrated Approach to Manufacturing Process Design and Costing. Concurr Eng 13(3):199-207 Gui J and Mäntylä M (1994) Functional understanding of assembly modelling. CAD 26:435-451 Gupta SK, Nau DS (1995) A systematic approach for analyzing the manufacturability of machined parts. CAD 27(5):342-343 Gupta SK, Regli WC, Das D, Nau, DS (1997) Automated Manufacturability Analysis: A Survey. Res Eng Des 9:168-190 Haik Y (2003) Engineering design process. Brooks/Cole Publishing, Pacific Grove Hirtz J, Stone RB, McAdams DA, Szykman S, Wood KL (2002) A functional basis for engineering design: Reconciling and evolving previous efforts. Res Eng Des 13:65-82 Horváth I (2004) A treatise on order in engineering design research. Res Eng Des 15:155-181 Hosaka M (1992) Modeling of curves and surfaces in CAD/CAM. Springer, Berlin Heidelberg New York Houe R and Grabot B (2007) Knowledge Modeling for Eco-design. Concurr Eng 15(1):7-20 Houssin R, Bernard A, Martin P, Ris G, Cherrier F (2006) Information system based on a working situation model for a new design approach in concurrent engineering. J Eng Des 17(1):35-54 Howell SK (2001) Engineering design and problem solving. Prentice Hall Huang H-Z and Gu Y-K (2006) Development Mode Based on Integration of Product Models and Process Models. Concurr Eng 14(1):27-34

26

1 State of the Art in the Field

Hubka V and Eder WE (1992) Engineering Design: General Procedural Model of Engineering Design. Heurista, Zurich Hundal MS (1993) Rules and Models for Low-Cost Design. In: Proceedings Design for Manufacturing Conference, ASME Hundal M (2000) Design for Recycling and Remanufacturing. In: Proceedings International Design Conference -DESIGN 2000, Dubrovnik Hung H, Kao H, Ku K (2007) Evaluation of design alternatives in collaborative development and production of modular products. Int J Adv Manuf Technol 33:1065-1076 Hyde W (1981) Improving Productivity by classification coding and Database Standardization. Marcel Dekker, New York Isaksson O, Keski-Seppälä S, Eppinger S (2000) Evaluation of design process alternatives using signal flow graphs. J Eng Des 11(3):211-224 ISO 14660-1:1999, Geometrical Product Specification (GPS) - Geometrical features - Part 1: General terms and definitions ISO 14660-2:1999, Geometrical Product Specification (GPS) - Geometrical features - Part 2: Extracted median line of a cylinder and a cone, extracted median surface, local size of an extracted feature ISO 5459:1981, Technical drawings - Geometrical Tolerancing - Datums and Datum Systems for geometrical tolerances ISO/TC 213 N355 Annex 1 (2000), Next generation of the Geometrical Product Specifications (GPS) language - The vision for an improved engineering tool ISO/TR 14638:1995, Geometrical product specification (GPS) - Masterplan ISO/TR 14062:2002, Environmental management -- Integrating environmental aspects into product design and development ISO/TS 17450-1:2000, Geometrical Product Specifications (GPS) - General Concepts - Part 1: Model for geometrical specification and verification ISO/TS 17450-2:2002, Geometrical Product Specifications (GPS) - General concepts - Part 2: Basic tenets, specifications, operators and uncertainties Jain VK and Sobek DK II (2006) Linking design process to customer satisfaction through virtual design of experiments. Res Eng Des 17:59-71 Khadilkar DV and Stauffer LA (1996) An experimental evaluation of design information reuse during conceptual design. J Eng Des 7(4):331-339 Kim Y-S, Cochrani DS (2000) Reviewing TRIZ from the perspective of axiomatic design. J Eng Des 11(1):79-94 Kleiner S, Anderl R, Gräb R (2003) A collaborative design system for product data integration. J Eng Des 14(4):421-428 Koh H, Ha S, Kim T, Lee S (2005) A method of accumulation and adaptation of design knowledge. Int J Adv Manuf Technol 26:943-949 Kulon J, Broomhead P, Mynors DJ (2006) Applying knowledge-based engineering to traditional manufacturing design. Int J Adv Manuf Technol 30:945-951 Kurakawa K (2004) A scenario-driven conceptual design information model and its formation. Res Eng Des 15(2):122-137 La Trobe-Bateman J and Wild D (2003) Design for manufacturing: use of a spreadsheet model of manufacturability to optimize product design and development. Res Eng Des 14(2):107-117 Lam WK (2005) Hardware Design Verification: Simulation and Formal Method-Based Approaches. Prentice Hall Lee K-S and Lee K (2004) Evolutionary design and re-design using design parameters and goals. J Eng Des 15(2):155-176 Liang C and Guodong J (2006) Product modeling for multidisciplinary collaborative design. Int J Adv Manuf Technol 30:589-600 Lin M-C, Tai Y-Y, Chen M-S, Chang CA (2007) A Rule Based Assembly Sequence Generation Method for Product Design. Concurr Eng 15(3):291-308 Lumsdaine E, Lumsdaine M, Shelnutt JW (1999) Creative Problem Solving and Engineering Design. Mc Graw-Hill

References

27

Lyons K et al. (1999) The Open Assembly Design Environment: An Architecture for Design Agent Interoperability. In: Proceedings ASME Design Engineering Technical Conf., ASME, New York, (CD-ROM) Maffin D (1998) Engineering design models: context, theory, and practice. J Eng Des 9(4):315327 Maher M, Balachandran B, Zhang D (1995) Case based reasoning in design. Lawrence Erlbaum Publisher Mantripragada R and Whitney DE (1998) The Datum Flow Chain: A Systematic Approach to Assembly Design and Modeling. Res Eng Des 10:150-165 McMahon C and Browne J (1998) CAD/CAM-principles, practice and manufacturing management, 2nd edn. Addison-Wesley, Harlow Mehlitz P and Penix J (2003) Design for Verification - Using Design Patterns to Build Reliable Systems. In: Proceedings of 6th ICSE workshop on component based software engineering, Portland OR Meredith DD, Wong KW, Woodhead RW, Wortman RH (1985) Design Planning of Engineering Systems. Prentice-Hall, Englewood Cliffs, NJ Meyer B (1997) Object Oriented Software Construction. Prentice Hall Molloy O, Warman EA, Tilley S (1998) Design for Manufacturing and Assembly: Concepts, architectures and implementation. Springer Montgomery D.C (2004) Design and Analysis of Experiments, 6th edn. Wiley Nahm Y-E and Ishikawa H (2004) Integrated Product and Process Modeling for Collaborative Design Environment. Concurr Eng 12(1):5-23 Nielsen J (1993) Usability Engineering. Academic Press, Cambridge, MA Nielsen HS (2003) Specifications, operators and uncertainties. In: Proceedings 8th CIRP International Seminar on Computer Aided Tolerancing Nielsen HS (2006) New concepts in specifications, operators and uncertainties and their impact on measurement and instrumentation. Meas Sci Technol 17:541-544 Olewnik A and Lewis K (2005) On Validating Engineering Design Decision Support Tools. Concurr Eng 13(2):111-122 Olewnik A and Lewis K (2006) A decision support framework for flexible system design. J Eng Des 17(1):75-97 Osteras T, Murthy DNP, Rausand M (2006) Product performance and specification in new product development. J Eng Des 17(2):177-192 Otto K and Wood K (2000) Product Design. Prentice Hall Pahl G, Badke-Schaub P, Frankenberger E (1999) Resume of 12 years of interdisciplinary empirical studies of engineering design in Germany. Des Stud 20(5):481-494 Pahl G, Beitz W, Feldhusen J, Grote KH (2007) Engineering Design: A systematic approach, 3rd edn. Springer Park SH (1996) Robust Design and Analysis for Quality Engineering. Springer Park H and Cutkosky MR (1999) Framework for Modeling Dependencies in Collaborative Engineering Processes. Res Eng Des 11:84-102 Phadke MS (1989) Quality Engineering Using Robust Design. Prentice Hall PTR, US edn Pnueli Y and Zussman E (1997) Evaluating the end-of-life value of a product and improving it by redesign. Int J Prod Res 35(4):921-942 Poli C (2001) Design for Manufacturing: A Structured Approach. Butterworth Heinemann Pugh S (1990) Total design. Addison-Wesley, Wokingham Qureshi SM, Shah JJ, Urban SD et al. (1997) Integration model to support archival of design history in databases. In: Proceedings ASME Design Theory and Methodology Conference, DETC97/DTM-3876 Reymen IM, Hammer DK, Kroes PA, van Aken JE, Dorst CH, Bax MFT, Basten T (2006) A domain-independent descriptive design model and its application to structured reflection on design processes. Res Eng Des 16(4):147-173 Riel AJ (1996) Object-Oriented Design Heuristics. Addison-Wesley

28

1 State of the Art in the Field

Ross DT (1977) Structured Analysis (SA): A language for communicating ideas. IEEE Trans Softw Eng 3(1):16-34 Ross PJ (1996) Taguchi Techniques for Quality Engineering. Mac Graw Hill professional Roy RK (2001) Design of Experiments Using The Taguchi Approach: 16 Steps to Product and Process Improvement. Wiley-Interscience, Har/Cdr edition Royce W (1970) Managing the Development of Large Software Systems. In: Proceedings of IEEE WESCON 26:1-9 Salomons OW, van Houten FJ, Kals HJ (1993) Review of research in feature-based design. J Manuf Syst 12(2):113-132 Seif MA (1998) A Concurrent Engineering Approach for Product Design Optimization. Concurr Eng 6(2):101-110 Shah J and Mäntylä M (1995) Parametric and Feature-Based CAD/CAM: Concepts, Techniques, and Applications. Wiley Interscience Shah JJ and Rogers M (1993) Assembly modeling as an extension of feature-based design. Res Eng Des 5:218-237 Shah J and Wright P (2000) Developing theoretical foundations of DfM. In: Proceedings ASME Design for Manufacturing Conference, (CD ROM Paper#DETC2000/DFM-14015) Sharma R and Gao JX (2007) A knowledge-based manufacturing and cost evaluation system for product design/re-design. Int J Adv Manuf Technol 33:856-865 Shehab EM, Abdalla HS (2001) Manufacturing cost modeling for concurrent product development. Robot Comput-Integr Manuf 17(4):341-353 Shehab EM and Abdalla HS (2002) An Intelligent Knowledge-Based System for Product Cost Modelling. Int J Adv Manuf Technol 19:49-65 Smith DR (1996) Toward a Classification Approach to Design. In: Proceedings of the Fifth International Conference on Algebraic Methodology and Software Technology, AMAST'96, LNCS 1101, Springer Verlag Smith LN (2003) A Knowledge-based System for Powder metallurgy Technology. Professional and Engineering Publishing, London & Bury St. Edmunds, UK Smith R and Eppinger S (1997) A predictive model of sequential iteration in engineering design. Manage Sci 43(8):1104-1120 Srinivasan V (2001) An Integrated View of Geometrical Product Specification and Verification. In: Proceedings 7th CIRP International Seminar on Computer Aided Tolerancing Srinivasan V (2008) Standardizing the specification, verification, and exchange of product geometry: Research, status and trends. CAD 40:738-749 Stamatis DH (1995) Failure Mode and Effect Analysis: FMEA from Theory to Execution. American Society for Quality Steward DV (1981a) The design structure system - a method for managing the design of complex systems. IEEE T Eng Manage 28(3):71-74 Steward DV (1981b) Systems Analysis and Management: Structure, Strategy and Design. Petrocelli Books, New York Sturges RH, O’Shaughnessy K, Reed G (1993) A systematic approach to conceptual design. Concurr Eng: Res Appl 1:93-105 Suh NP (1990) The Principles of Design. Oxford University Press Suh NP (1998) Axiomatic Design Theory for Systems. Res Eng Des 10:189-209 Suh NP (1999) A Theory of Complexity, Periodicity and the Design Axioms. Res Eng Des 11:116-131 Suh NP (2001) Axiomatic Design: Advances and Applications. Oxford University Press, USA Suh NP (2005) Complexity: Theory and Applications. Mit-Pappalardo Series in Mechanical Engineering, Oxford University Press, USA Taguchi G (1999) Robust Engineering: Learn How to Boost Quality While Reducing Costs & Time to Market. McGraw-Hill Professional Taguchi G (2004) Taguchi's Quality Engineering Handbook. Wiley-Interscience Thompson G (1999) Improving Maintainability and Reliability through Design. Wiley

References

29

Tor SB, Britton GA, Zhang WY (2005) A knowledge-based blackboard framework for stamping process planning in progressive die design. Int J Adv Manuf Technol 26:774-783 Ullman DG (2002a) The mechanical design process. McGraw-Hill Science/Engineering/Math Ullman DG (2002b) Toward the ideal mechanical engineering design support system. Res Eng Des13:55-64 Ulrich KT and Eppinger SD (2003) Product Design and Development. Mc Graw Hill Veeke HPM, Lodewijks G, Ottjes JA (2006) Conceptual design of industrial systems: an approach to support collaboration. Res Eng Des 17:85-101 Vergeest JSM and Horváth I (1999) Design Model Sharing in Concurrent Engineering: Theory and Practice. Concurr Eng 7(2):105-113 Visser W, Havelund K, Brat G, Park S (2000) Model Checking Programs. In: Proceedings of the 15th International Conference on Automated Software Engineering (ASE), Grenoble, France Wang B, Chen SB, Wang JJ (2005) Rough set based knowledge modeling for the aluminium alloy pulsed GTAW process. Int J Adv Manuf Technol 25:902-908 Wasserman G (2002) Reliability Verification, Testing, and Analysis in Engineering Design. CRC Whitney DE, Mantripragada R, Adams JD, Rhee SJ (1999) Designing Assemblies. Res Eng Des11:229-253 Wiest JD, Levy FK (1977) A Management Guide to PERT/CPM. Prentice-Hall, Englewood Cliffs, NJ Wu Z and Duffy A (2002) Mutual Knowledge Evolution in Team Design. In: Proceedings of 7th International Conference on Artificial Intelligence in Design (AID '02), Cambridge, UK Wu T and O’Grady P (1999) A Concurrent Engineering Approach to Design for Assembly. Concurr Eng 7(3):231-243 Wu Y and Wu A (2000) Taguchi Methods for Robust Design. ASME books Xu QL, Ong SK, Nee AYC (2006) Function-based design synthesis approach to design reuse. Res Eng Des 17:27-44 Xue D (1997) A Multi-level Optimization Approach Considering Product Realization Process Alternatives and Parameters for Improving Manufacturability. J Manuf Syst 16(5):337-351 Xuewen C, Siyu Z, Jun C, Xueyu R (2005) Research of knowledge-based hammer forging design support system. Int J Adv Manuf Technol 27:25-32 Yang G (2007) Life Cycle Reliability Engineering. Wiley Yassine A, Whitney D, Daleiden S, Lavine J (2003) Connectivity maps: modeling and analysing relationships in product development processes. J Eng Des 14(3):377-394 Zhang WY, Tor SY, Britton GA (2006) Managing modularity in product family design with functional modelling. Int J Adv Manuf Technol 30:579-588 Zhao Z and Shah J (2002) A Normative Framework for DfM based on Benefit-Cost Analysi In: Proceedings ASME Design for Manufacturing Conference, (CD ROM Paper#DETC2002/DFM-34176) Zhao Z and Shah J (2004) Modeling and representation of manufacturing knowledge for DFM systems. In: Proceedings ASME Computers and Information in Engineering Conference, (CD ROM Paper#DETC2004/57724)

2 The DGLs-CF — Introduction and Background

The DGLs-CF is a methodological approach for product design and process reconfiguration, aimed at effectively helping and leading the activities of designers, manufacturers and inspectors. The initial consideration is that designers are not necessarily experts in manufacturing and verification processes; likewise, manufacturers and inspectors might not be experts in design. As said in the previous chapter, the birth and evolution of the DGLs-CF took place in a scenario where tools and methods for product development have been analysed from a concurrent engineering point of view, and where the adoption of standards is a key point. Considering the terminology explained in the previous chapter, the DGLs-CF may be considered essentially a design for multi-X method. At the beginning, the former DGLs were developed as a simple DfM method, particularly focused on establishing guidelines for product design considering the characteristics of some manufacturing processes. Subsequent improvements highlighted other interesting peculiarities that started to differentiate the DGLs from the classic DfX methods. According to the ISO GPS vision, it was also considered that the verification process affects the design phase, so that verification aspects have been evaluated together with the manufacturing ones. A sort of Design for Manufacturing and Verification method was the result, to link design, manufacturing and verification in a new way. The study and development of the DGLs-CF as a design for multi-X method, strictly connected with the ISO GPS standards, can be considered the most interesting issue of the whole research. Figure 2.1 shows the role of the DGLs-CF in the product development process. The full-line boxes and arrows represent the knowledge and its flow corresponding to the structure of the former DGLs, while the dashed elements correspond to the new aspects introduced by the DGLs-CF. The contribution of the DGLs-CF, actually representing a differentiation from the classic DfX methods, stands in the fact that knowledge evaluation not only generates guidelines for product development, but also for process reconfiguration. Not only guidelines to redesign the product according with the

32

2 The DGLs-CF – Introduction and Background

manufacturing and verification characteristics are given, but also guidelines to reconfigure the manufacturing/verification processes specifically for the redesigned product. Anyway, the attention is always focused on the product; the process reconfiguration must be intended as a process customisation where the different representations of the product — digital model or physical part — are managed and modified. For now, there is no redefinition of the general parameters of the process. As said before, all of this is shown in Fig. 2.1, where this additional knowledge generation and flow is depicted by the dashed box and arrows.

Fig. 2.1 Role of the DGLs-CF as a design for multi-X method in the product lifecycle

From this new point of view, the specific product is generated by exploiting the specific process characteristics effectively and adequately, since the product is tailored to manufacturing and verification processes, and vice versa.

2.1 Overview of the History of the DGLs Detailed descriptions of the several releases of the DGLs were given by Bandera et al. (2004, 2005, 2006), Cristofolini et al. (2006), Filippi et al. (2001) and Filippi and Cristofolini (2007). Several fundamental aspects are recalled here to make the understanding of what follows easier. The DGLs development started from two main considerations. The redesign activities must remain the responsibility of the designers. Quite often today, manufacturers carry out redesign activities in order to improve the feasibility of products with particular technologies, but this is potentially dangerous; the operators may have little knowledge of the product domain and

2.1 Overview of the History of the DGLs

33

may not take the best decision in the case of multiple choices. Even worse, sometimes they may prejudice the model functions (Kumaran and Chittaro 1998). Moreover, designers need specific knowledge of the manufacturing processes to succeed in redesign (Haffey and Duffy 2000). Design rules and handbooks existing in the literature now represent the only help for designers in modifying the product. Even if constituting a good starting point, they show some limitations that severely reduce their usability (Nielsen 1993) and the consequent effectiveness during the design phase. In fact, they do not have a structure that is helpful in generating, managing and using the rules; there is a lack of application criteria to guide the redesign activities; overall, they impose a scarce autonomy due to the fact that they have not been specifically produced for designers. Three releases of the DGLs exist, corresponding to three important milestones during their development. They are briefly introduced hereafter, and the following paragraphs describe them in detail. The DGLs-CF, representing both the target of the research about a new concept of DfX method and the starting point for further theoretical and applicative research, will be described and applied in the field in the next two chapters of this book. In their first release, the DGLs were based on a set of design rules that completed the existing set in literature and increased the possibilities of their use as an effective guide for the modification of a product during the design phase. Modifications guaranteed not only that the model could be built with particular technology, but also that the building process was advantageous when compared to other technologies, thus exploiting its capabilities and minimising time and costs. This concept is the basis for the definition of the so-called positive rules that will be recalled more than once throughout the book. In the second release of the DGLs, the attention was on the verification phase, following the idea that the product quality could be improved when considering at the design phase not only the manufacturing process, but also the verification one. In this way, the DGLs enlarged their field of application in realising links between design and verification, according to the ISO GPS concepts. Finally, in the third release of the DGLs the impact in terms of process reconfiguration was evaluated in depth; in this way the knowledge sharing among designers, manufacturers and verification experts became biunique. In a true concurrent engineering environment, the DGLs suggested the way to redesign the product and to reconfigure the processes to obtain and maintain the required functionalities. This chapter goes ahead with the description of these three releases of the DGLs. They have been considered in this book because the development of each of them generated new and heterogeneous knowledge about the problem — the product redesign and process reconfiguration — and about the method to solve it — the DGLs-CF. The DGLs-CF answers almost all the questions and hints coming from these previous releases. Moreover, the evolution of the whole development process could be considered for similar situations in different application domains.

34

2 The DGLs-CF – Introduction and Background

Each description shows quite the same structure, an introductory paragraph followed by three elements: the conceptual diagram representing the process for gathering and deriving the knowledge, the knowledge matrix that describes the data structure used to encode this knowledge, and the pros and cons at the end of each description that weigh the importance of the results reached and list the directions for the next release. The description of the third release of the DGLs diverges slightly from this structure; the complexity of this release required a sort of roadmap for its application in the field and this new element has been used as a tool to explain the release. Moreover, the descriptions of the three releases contain some case studies to clarify the effects of the DGLs adoption in some real industrial domains.

2.2 First Release of the DGLs. The Beginning The first release of the DGLs was derived directly from the authors’ experience in the field as mechanical designers and Rapid Prototyping — RP — experts. The characteristics of RP technology named Direct Metal Laser Sintering — DMLS — were analysed first, in terms of advantages and drawbacks; then, how these characteristics could affect the product redesign has been investigated (Filippi et al. 2001). The DMLS process was chosen because this is one of the most promising RP technologies currently available. It can build metal objects using the same material as that of the final product. This characteristic widens the field of application of this technology, which can also be used in rapid tooling, i.e. for the generation of inserts for plastic injection moulding (Kuzman et al. 2001; Nelson 2002), and in rapid manufacturing, to build small series of complex mechanical parts (Gatto and Iuliano 1998; Jacobs 1995). On the other hand, the DMLS process shows some critical aspects mainly due to the behaviour of metal powders, such as complex sintering dynamics, residual stress, thermal deformation, etc. (Agarwala et al. 1995), so that more study is needed to solve or avoid current limitations even in early activities, for example during the design phase (Otto and Wood 2000). At the beginning of the development of this release, the attention was mainly focused on collecting knowledge in a spreadsheet as in the work of La TrobeBateman and Wild (2003), and in generating rules to link the design and the manufacturing domains; after that, some software development and concerns about usability issues took place.

2.2.1 Conceptual Diagram The conceptual diagram in Fig. 2.2 shows that this release of the DGLs has been heavily influenced by the considerations about the manufacturing technologies.

2.2 First Release of the DGLs. The Beginning

35

Fig. 2.2 Conceptual diagram of knowledge generation in the first release of the DGLs

Two domains were present, design and manufacturing, and, starting from the manufacturing requirements, a set of design rules was derived. Then the product was characterised in terms of attributes. The match between design rules and attributes gave a configuration, expressed in terms of activities to be performed on the product to make it compatible with the manufacturing technology.

2.2.2 Knowledge Matrix Table 2.1 shows the knowledge matrix of this release of the DGLs. This matrix represents the knowledge structure and contains some examples of pieces of information, expressed in terms of manufacturing requirements, design rules, attributes and redesign solutions. Each row represents a different requirement with an associated design rule. For each rule there is one attribute that allows the characterisation of the product. For each requirement, redesign solutions are suggested to be used when the product does not match the rule. After the definition of the data structure and the collection of some pieces of information, a very simple software prototype was generated using Microsoft Access. This implementation is of interest here because it allowed applying some refinements to the knowledge content of the DGLs. For example, the consequences of the violation of each rule and a coarse evaluation of the required post-process manufacturing cost — using a qualitative low-medium-high ranking — were defined; moreover, each hint — redesign solution — had its redesign cost associated. Finally, a picture explained the meaning of the rule. Figure 2.3 shows the form containing the description of a rule, with some comments about the meaning of the fields used.

36

2 The DGLs-CF – Introduction and Background

Table 2.1 Knowledge matrix of the first release of the DGLs Manufacturing requirements Design rules Assure compatibility between workpiece dimensions and workspace of the sintering machine (workspace is 250×250×185 mm) Minimise building time (building time increases with the height) Assure proper building conditions (the recoater may cause bending of parts with high form ratios)

Attributes Descriptions

Keep maximum Maximum Maximum size size below 185 size of the model mm Minimise the height

Height

Redesign solutions Split the model

Height of the model

Split the model Change orientation Minimise the Form ratio Ratio between Change form ratio the height and orientation the horizontal Add reinforcing cross-section of structures the model Minimise post-processing Avoid lower Lower Surfaces Add overhangs operations (lower surfaces surfaces surfaces oriented Change require external support which downward with orientation has to be removed an angle <30° subsequently) from the horizontal Minimise post-processing Avoid surfaces Surface Roughness None operations (average roughness with a required texture value of DMLS parts is 15 m) roughness < 15 m

Fig. 2.3 Description of a rule using a form of the Microsoft Access database developed for the first release of the DGLs

2.2 First Release of the DGLs. The Beginning

37

2.2.3 Case Study A coffee machine to be manufactured with the DMLS process was used as a case study to illustrate the adoption of this release of the DGLs in the field. This adoption consisted of a three-step procedure: product description, compatibility check, and configuration of the redesign choices. The description of this case study is based on Fig. 2.4, showing the model of the product under evaluation, and on Fig. 2.5, where the forms of the software prototypes corresponding to the three steps of the procedure are depicted.

Fig. 2.4 Starting product model - section - of the body of the coffee machine used in the case study

Step 1. Product Description Figure 2.4 shows the section of the starting model of the coffee machine, while Fig. 2.5a reports an example of the attributes used to describe it. These attributes represented the functional components of the product, and they could be related to the following classes: geometry, dimensions, orientation, finishing, precision, mechanical behaviour and physical properties. Given this description, a set of rules for the current product was selected automatically from the database of rules related to the DMLS technology. Step 2. Compatibility Check For each attribute of the coffee machine, the database showed the list of the rules that could be related to it, as shown in Fig. 2.5b. Then the user checked manually whether these rules were really pertinent to the current situation and, if yes, if they appeared violated or not. Step 3. Configuration of the Redesign Choices For each critical situation highlighted before, the user made some choices that led the redesign activities. In this very important step, the software prototype generated a report containing all the information required for the next tasks. The left side of the form in Fig. 2.5c

38

2 The DGLs-CF – Introduction and Background

lists all the attributes; for each of them, the right side contains the knowledge related to the incompatibilities, expressed by the hints to solve the problems and the related costs.

Fig. 2.5 Forms representing the three-step procedure of the DGLs adoption

In other words, the users of this software prototype defined a new product configuration, expressed as a collection of redesign activities. This happened because they were making some choices for each critical situation. The software prototype gave the total cost for the final configuration. The user could obviously generate more than one configuration, and compare them in the end. In the case of the coffee machine, two different configurations were considered, presenting different characteristics; in the first one, for example, the surfaces with a slope less than 30° were maintained unchanged, so some external supports would have been required; in the second, the product was changed in order to avoid supports. The

2.2 First Release of the DGLs. The Beginning

39

comparison of the costs of these two configurations led to the choice of the second one. The chosen configuration contained also other suggestions for the reconfiguration of the model of the coffee machine. Figure 2.6 shows the result of the application of this configuration, while in Fig. 2.7 there is the physical prototype after the manufacturing activities.

Fig. 2.6 Result of the redesign activities based on the DGLs adoption

Fig. 2.7 The physical body of the coffee machine built using DMLS

The quality of the prototype confirmed the goodness and effectiveness of the guidelines. The coffee machine was successfully manufactured; it had quite good mechanical properties, given that it showed very few flaws. Moreover, void channels shaped like a helix were generated without problems and they represent a good example of the positive rules used to highlight and exploit the possibilities offered by the specific technology. The number of external supports was reduced as expected, to optimise the manufacturing process and minimise the postprocessing activities. Nevertheless, the prototype presented some deficiencies: small parts were poorly manufactured in terms of precision and quality; some

40

2 The DGLs-CF – Introduction and Background

supports were generated corresponding to thin walls and for this reason the model could be damaged during the removal of them; moreover, somewhere it was not so clear where the supports finished and the prototype started. Finally, some supports were attached to non-flat surfaces, making support removal a difficult task. All of this means that the DGLs showed some deficits regarding the knowledge content, but this could also partly depend on the incomplete information coming from the manuals and literature about DMLS technology.

2.2.4 Discussion Even if this first release of the DGLs allowed one to focus the main issues of the redesign process and generated some interesting results as in the case of the coffee machine, some problems, mistakes and open issues arose during the research. They are listed in the following, using a classification that will be exploited throughout the whole book. Conceptual Diagram and Knowledge Organization • The conceptual diagram was too simple and did not represent knowledge generation correctly as occurs in real application domains. • The redesign column of the knowledge matrix did not have real corresponding elements in the conceptual diagram. • The location of the domains and their characterization were wrong. • The knowledge matrix structure could lead to considerable redundancy, for example when the same technological requirement required more rules, each one with several attributes. • The distinction among the different domains present in the conceptual diagram was lost in the knowledge matrix. As will be clear in the following, this is a serious drawback. Knowledge Description • The descriptions of technological requirements, rules, attributes, etc. were always too general, with no quantification. • Product characterisation is one of the most important issues of the whole research. It is an intrinsically difficult task, given the large number of degrees of freedom involved. In this first release of the DGLs, this characterisation was extremely simple, merely to test the complete process, from product characterisation to product redesign.

2.3 Second Release of the DGLs. Improvements and the Synergy with the ISO GPS 41

Costs • Costs were managed in an extremely coarse way. They were considered just to set a future role for them and to check some minimal related functionality. Implementation/Automatisms • There was nothing automatic or, at least, semi-automatic. This release of the DGLs was a simple database with an interface. There was no data processing at all and users must do everything themselves.

2.3 Second Release of the DGLs. Improvements and the Synergy with the ISO GPS The remarks in the previous paragraphs were considered as the starting point for the next release of the DGLs. The attempt to eliminate problems and mistakes was coupled with an effort to widen the view, considering other aspects of the product lifecycle. A synergy started with some experts in product verification, focusing on the dimensional, micro- and macro-geometrical issues (Bandera et al. 2004, 2005). All of this led to the adoption of the ISO GPS principles. In this context, the analysis of the possible interaction/integration between the DGLs and the ISO GPS seemed very interesting and convenient for many reasons. The closeness between the DGLs and the ISO GPS was immediately clear, from a conceptual point of view, because both of them are basically tools to help the designer in specifying, refining and communicating the product characteristics. As the DGLs were a design tool related to a specific manufacturing process, the possibility and opportunity of revising their structure and contents to agree with ISO GPS concepts seemed of great interest. This could signify establishing the possibility of making a real link between design, manufacturing and verification, and enlarging in this way the DGLs application domain. As explained below, ISO GPS principles have been actively adopted for updating the generation and the formalisation of knowledge within the DGLs. The most important new elements in this release of the DGLs were the introduction of a new domain — the verification domain, a classification of requirements and rules, and a new knowledge matrix layout.

2.3.1 Conceptual Diagram Figure 2.8 shows the conceptual diagram of this release of the DGLs. There were three domains, and the manufacturing and verification ones determined two different sets of requirements. Two sets of rules arose directly from them: the DfM

42

2 The DGLs-CF – Introduction and Background

rules, and the DfV rules. The attributes used for describing the product came from both these sets. The product configuration, once defined in terms of attribute values, determined the generation of the last two sets of derived rules: manufacturing rules and verification rules. The meaning of all these elements should be clear considering the content of the knowledge matrix. It can be seen that here the new aspects of the DGLs start to become evident; manufacturing rules and verification rules and their flow of information, represented with dashed arrows and boxes, correspond quite well to the dashed elements shown in Fig. 2.1.

Fig. 2.8 Conceptual diagram of the second release of the DGLs

2.3.2 Knowledge Matrix As regards the knowledge matrix, there are many new elements and issues to consider here. First of all, the layout of the matrix shown in Table 2.2 did not match the conceptual diagram. This is because the idea was to bring the knowledge structure closer to the final user than to the system implementation. The attributes, the most important elements from the users’ point of view, appeared in the left side of the matrix. Then there were the DfM and DfV rules with the different attributes. To the right there were the manufacturing and verification requirements that, in a sense, justify the presence of the rules and of the attributes. For each attribute, in the lower part of the knowledge matrix there were the manufacturing rules and the verification rules selected during the definition of the product configuration and represented explicitly. These rules were kept separated from the others because they were related directly to the manufacturing and verification domains; they were not redesign rules.

2.3 Second Release of the DGLs. Improvements and the Synergy with the ISO GPS 43 Table 2.2 Knowledge matrix of the second release of the DGLs Attribute Description DfM and DfV rules Height

Height of Minimise the height of the model the model (linear dimension) On the basis of functional requirements and related dimensional tolerances, establish the verification method and indicate it on the drawing with proper symbols (for example, LP-two-point size; SNMinimum statistical size; SX-Maximum statistical size; SAAverage statistical size) a

Attribute Description DfM and DfV rules Surface texture

a

Manufacturing requirements Minimise building time (building time depends on the height) -

Verification requirements -

Manufacturing rules Use thicker layers

Verification rules Measure the height according to the indication established on the drawing

Manufacturing requirements Minimise post-processing Roughness Avoid surfaces with a value required roughness < 15 operations (average roughness of DMLS parts μm is 15 μm) Avoid locating surfaces with strict roughness specifications in deep and narrow cavities. Manufacturing rules If fine surfaces are required, use 20 μm metal powder and thin layers

The verification methods and tools have to be chosen considering the characteristics of the features to measure

Verification requirements -

Assure enough space for standard measure instruments Verification rules If the roughness verification is difficult to accomplish with standard instruments, evaluate the possibility of using non-contact surface measurement

LP may be recommended for dimensions which are not critical, while for critical dimensions in assembly, statistical measurements could be preferred. Particularly, SN for external features, which have to guarantee a fitting with interference, SX for external features, which have to guarantee a fitting with clearance and SA for individual features.

44

2 The DGLs-CF – Introduction and Background

2.3.3 Case Study Also for this release of the DGLs a simple software package was developed and adopted in the field for the redesign of a mould insert for the production of a plastic part used in the textile industry. The problem to solve was a high cycle time and a poor quality of the moulded part due to an incorrect cooling of the mould. Figure 2.9 shows the plastic part with highlights corresponding to uncooled regions.

Fig. 2.9 The plastic part used in the case study of the second release of the DGLs

To solve the problem, the idea was to redesign the mould insert used to generate the cavities of the plastic part. Its feasibility with the DMLS technology was evaluated and improved with this second release of the DGLs. Design solutions were considered to avoid mechanical problems of the insert under moulding loads and to exploit at best the potentialities of the DMLS technology, by using again the positive rules. Figure 2.10 shows the evolution of the model of the insert due to the DGLs adoption. The main differences between the starting and the final configuration stand in a different number of dangling elements — three instead of four — in a different shape of cooling channels, and in the addition of some post-processed elements to improve the strength of the insert during the moulding. Moreover, given that it was important to ensure a good surface finish, as well as the coaxiality relationship between the inner hole and the external tapered surfaces, the attention to the verification issues was important as well.

Fig. 2.10 Redesign of the mould insert thanks to the DGLs adoption

2.3 Second Release of the DGLs. Improvements and the Synergy with the ISO GPS 45

Figure 2.11 shows the mould where the redesigned insert was placed and the final plastic part generated using it.

Fig. 2.11 Mould with the redesigned insert (left) and final moulded part (right)

2.3.4 Discussion This paragraph, unlike the corresponding one in the description of the first release of the DGLs, opens with some final considerations about the positive outcomes of this release of the DGLs and about the problems solved. After that, the list of the remaining problems, mistakes, and open issues, is given as before. Positive Outcomes The ISO GPS adoption helped the development of this DGLs release in several aspects. Here product formalisation was much better and the consideration of the verification phase of the product lifecycle enhanced the coverage of this knowledge based tool, using a conceptual diagram closer to the real scenarios. The correspondence between the conceptual diagram for knowledge gathering and the knowledge matrix appeared enhanced as well. Moreover, regarding the quantitative aspects, some attributes with corresponding values were introduced. Relating to the categories used to classify problems and open issues of the previous release, here something positive was done in the Conceptual diagram and knowledge organisation, and in Knowledge description. However, the experience described above gave some impressions on the presence of criticisms even on this revised version of the DGLs. They appear in the following. Conceptual Diagram and Knowledge Organisation • The knowledge gathering process was still wrong. The different pieces of information were misplaced; the derivation rules were rough. Sometimes there were coarse mistakes in the interpretation of the meaning of the data. For example, what were called here manufacturing or verification rules would have been better intended as a collection of hints to consider or actions that must be

46

2 The DGLs-CF – Introduction and Background

performed. This is one of the main elements taken into consideration during the development of the next release of the DGLs. • The structure of the knowledge matrix was unclear and ambiguous. Things were simpler in the first release, where the knowledge matrix was a plain table. Here the knowledge matrix was again a sort of table but the manufacturing and verification rules broke the table concept, introducing some asymmetry in the structure. At the same time, it was not a bi-dimensional matrix, as it was impossible to identify the two independent dimensions to label the rows and the columns. • Redesign hints disappeared in attempting to make the knowledge matrix consistent with the conceptual diagram. Apparently this solved the problem of the previous release but this is untrue because the redesign hints were still present and, even worse, in a misunderstood way, in the form of manufacturing and verification rules. • It should have been the time to adopt a clear formalism to label the items and the entities involved in the process. Knowledge Description • There was still a lack of quantification. Some attributes allowed consideration of numerical values — dimensions of the product, surface finishing values, etc. — but the approach was not systematic and these values were used to get an over-simple evaluation of the product compatibility with the manufacturing and verification activities. • The possibilities offered by the ISO GPS to define a clear way to characterise the product to redesign have not been exploited enough. Costs • Costs completely disappeared in this release of the DGLs, mainly because the focus was on the ISO GPS adoption; it was quite clear that a serious and rigorous cost evaluation was impracticable at that moment. Implementation/Automatisms • Again, there were no automatisms in this second release of the DGLs, but, even worse, there wasn’t any feeling about some serious identification of procedures and about the implementation of the corresponding software modules. ISO GPS Adoption • Even if the adoption of the ISO GPS introduced a high added-value to the DGLs, too few ISO GPS concepts have been exploited. For example, the duality between specification and verification, one of the most interesting ISO

2.4 Third Release of the DGLs. Thinking Big

47

GPS concepts, found scarce correspondence in the structure and content of the DGLs knowledge. DGLs Adoption Process • In some ways, the moment of the generation of the manufacturing and verification rules was wrong; there was no reason to wait until the determination of the product configuration. The set of hints to consider or actions to perform could be defined at the beginning, during the DGLs setup, independently from the specific product under analysis. Once defined, the configuration of the specific product, the meaningful hints or actions, could be simply selected from the set. An important outcome of these considerations is that it was the first time that the need for a clear distinction between the setup phase of the DGLs and their adoption became clear. • Given that the DGLs were going to become a large project, the need for an overall architecture and for some sort of roadmap for their adoption arose.

2.4 Third Release of the DGLs. Thinking Big The development of the third release of the DGLs started from the suggestions summarised in the last paragraph of the previous section. In this case, the main goals were to introduce an overall architecture of the project, to take care of knowledge formalisation and correctness, to introduce a real management of product configuration in terms of parameter quantification, to update the knowledge content, to take care of terminology correctness, and to keep under consideration the customisation/flexibility of the system (Filippi and Cristofolini 2007).

2.4.1 Conceptual Diagram In this release of the DGLs the conceptual diagram became more complex and articulated. This arose from a deep investigation of knowledge generation, the cause-effect paradigm, the relationships between the various domains and the different pieces of information involved. A multi-storey structure was generated, the DGLs building, with five floors: Compatibility floor, Rules floor, Design domain floor, Manufacturing domain floor, and Verification domain floor, as shown in Fig. 2.12.

48

2 The DGLs-CF – Introduction and Background

Fig. 2.12 The five-storey DGLs building

The meaning of these information loci is reported hereafter. • Compatibility Floor This floor was used to evaluate the compatibility of the product with the characteristics of the manufacturing and verification processes. Characteristic was the new term used instead of requirement or others. For each product feature — this was the new term used instead of attribute or others — this floor contained all the information needed to calculate a numerical value representing this compatibility. • Rules Floor This floor contained all the rules determined by relating the technological characteristics of manufacturing and verification to the features used to characterise the product. Each rule had an optional brief explanation that justified its presence. This information represented added value for the user who could acquire knowledge of the technologies and use it in future product development. This floor was filled during the setup of the DGLs, before their use. Only during the use of the DGLs, by determining the configuration of a specific product, did the knowledge activation take place; in other words, the selection of the meaningful pieces of information from a general database happened time by time, case by case. • Design Domain Floor This floor contained all the actions related to the design phase of the product. Actions were the suggested way to redesign the product in order to respect the rules. According to the ISO GPS concept of establishing links among design and manufacturing/verification, some actions in the diagram were represented as coupled. For example, there could be one action in the Design domain floor, “Split the model”, and another, linked to it, in the Manufacturing or Verification domain floor, “Merge the split parts”. Moreover,

2.4 Third Release of the DGLs. Thinking Big

49

an estimated value was associated with each action, representing its cost. These values were used during the evaluation of the product compatibility as it will be clear in the following. • Manufacturing Domain Floor The content and meaning of this floor was the same as the previous one but the actions were related to the manufacturing domain. • Verification Domain Floor Same as before, but related to the verification domain.

2.4.2 Knowledge Matrix In this case, the knowledge matrix reflected exactly the conceptual diagram. There were three-dimensional matrices where each row was related to a feature and each column to a technological characteristic. The cell contents referred to the five floors. The complete structure with the full content of the knowledge matrix is described in the case study.

2.4.3 Roadmap for the Adoption of the DGLs It was possible to distinguish three separate phases in adopting this release of the DGLs: a Setup phase where the DGLs were addressed to a particular class of manufacturing and verification technologies but not to specific brands and models; a Configuration phase where the DGLs were customised using the data of the available technology; and a Usage phase, the only one depending on the product analysed. Given the articulated structure of the DGLs and the presence of many components with very different meanings and operations, the so-called DGLs roadmap was developed, in order to organise and sequence all the required activities. The case study described hereafter strictly followed this roadmap and the description itself is DGLs roadmap-based, so that the entire sequence should be clear.

2.4.4 Case Study The spacer shown in drawing format in Fig. 2.13 allowed the testing in the field of this third version of the DGLs. This component, adequately positioned on a base by means of the narrow slide and the overhangs, allows the connection of an upper part using pins inserted in the holes. The goal was to redesign this spacer to optimise its manufacturing with the RP technology named Fused Deposition Modelling — FDM — and its verification with a Coordinate Measuring Machine

50

2 The DGLs-CF – Introduction and Background

— CMM. FDM and CMM technologies will be described in detail later in the case studies related to the adoption of the DGLs-CF; anyway, to go into details with some meaningful examples related to FDM technologies see Jacobs (1995), and for CMM see Bosch (1995).

Fig. 2.13 Mechanical drawing of a spacer, the test model for the third release of the DGLs

Setup Phase Identification of the Technological Characteristics This step analysed the classes of available manufacturing and verification technologies — FDM and CMM — and described them in terms of technological characteristics and related, meaningful parameters. The result is shown in Table 2.3. Table 2.3 Technological characteristics with related parameters Label Characteristic Parameters M1 Manufacturing workspace xMmax, yMmax, zMmax (max dimensions) M2 Supports xSmin, ySmin, zSmin, αSmin, αSmax (min dimensions, angles between vertical and walls) V1 Verification workspace xVmax, yVmax, zVmax (max dimensions) V2 Probes φPmin, lPmax (min diameter, max length)

Identification of the Features to Characterize the Product Here the class of the products to be evaluated was analysed and a list of features was generated to

2.4 Third Release of the DGLs. Thinking Big

51

characterise them best, with the related parameters and a brief description. The result is shown in Table 2.4. Table 2.4 Product features with related parameters Label Feature Parameters F1 Dimensions Xmin, Ymin, Zmin, Xmax, Ymax, Zmax (minimum and maximum dimensions) F2 Overhangs α (overhangs angle) F3 Cavities xCmin, yCmin, dCmax (minimum dimensions and depth)

Description (optional) This feature represents the overall dimensions of the product -

Generation of Rules Each characteristic/feature pair could generate the situation where one or more rules were inserted in the Rules floor of the knowledge matrix. Rules allowed the development of the actions in the upper floors. In this release of the DGLs the attention was focused on overcoming the critical characteristics of the process, while the positive rules were not considered. Table 2.5 shows the rules used in this case study. Table 2.5 Rules generated from technological characteristics and product features Label Origin Rule R1 M1 vs F1 Maximum dimensions of the product must be minor than maximum dimensions of the building room R2 M2 vs F1 Minimum dimensions of the product must be greater than minimum dimensions related to the presence of supports R3 M2 vs F2 The presence of overhangs must be evaluated considering the need for supports R4 M2 vs F3 The dimensions and depth of cavities must be compatible with the need of supports R5 V1 vs F1 Maximum dimensions of the product must be minor than maximum dimensions of the measuring volume R6 V2 vs F3 Dimensions and depth of cavities must be compatible with the characteristics of the probes

Generation of Expressions and Overall Judgment Criteria to Evaluate Compatibility In correspondence with the entries where rules have been inserted, compatibility values between the product and the technologies were collected. This step generated the expressions to calculate these values. The expressions populated the Compatibility floor of the knowledge matrix. Normalised compatibility values (interval [0..1]) were used. Of course, the quantity/quality of the expressions inserted in the knowledge matrix determined the accuracy of the compatibility evaluation. Together with the expressions, this step set the criterion to express an overall judgment on the product/processes compatibility. Sometimes the sum of all the compatibility values might fit; in other cases the average might be better, and so on. This judgment was particularly important when the user needed to compare different products or different configurations of the same

52

2 The DGLs-CF – Introduction and Background

product. In this case study, the overall judgment criterion was set to the average. Table 2.6 shows the expressions used to evaluate the compatibility values in the case study. There is an appendix at the end of the book, showing some details about a couple of them. Table 2.6 Expressions to compute compatibility values Label Origin Expression E1 M1 vs F1 E1=1 IF ZmaxzMmax OR Xmax>xMmax OR Ymax>yMmax E2 M2 vs F1 E2=1 IF Zmin>zSmin AND Xmin>xSmin AND Ymin>ySmin ELSE E2=0 IF Zmin3•MIN(xSmin, ySmin) ELSE E4=0.5 IF MIN(xSmin, ySmin)<MIN(Xmin, Xmin)<3•MIN(xSmin, ySmin) ELSE E4=0 IF MIN(Xmin, Ymin)<MIN(xSmin, ySmin) E5 V1 vs F1 E5=1 IF ZmaxzVmax OR Xmax>xVmax OR Ymax>yVmax E6a V2 vs F3 E6=0 IF dCmax>lPmax ELSE E6=1 IF MIN(xCmin, yCmin)≥5•φPmin ELSE E6=(MIN(xCmin, yCmin)2•φPmin)/3•φPmin IF 2•φPmin <MIN(xCmin, yCmin) <5•φPmin ELSE E6=0 IF MIN(xCmin, yCmin) ≤2•φPmin a See Appendix for details

Meaning Compatibility between model maximum dimensions and Workspace dimensions Compatibility between model minimum dimensions and need for supports Compatibility between model overhangs and need for supports Compatibilities between cavities and need for supports

Compatibility between model maximum dimensions and measuring volume Compatibility between cavities and probes characteristics

At this point, the content of the knowledge matrix related to the Compatibility floor and to the Rules floor was completed; Tables 2.7 and 2.8 show it respectively.

2.4 Third Release of the DGLs. Thinking Big

53

Table 2.7 Part of the knowledge matrix corresponding to the Compatibility floor COMPATIBILITY FLOOR

Characteristics Manufacturing M1: M2: Supports Manufacturing workspace xSmin, ySmin, zSmin, αSmin, αSmax xMmax, yMmax, zMmax

Features F1: E1=1 IF Dimensions Zmax zMmax OR Xmax> xMmax OR Ymax> yMmax F2: Overhangs α

F3: Cavities MIN(xCmin, yCmin), dCmax

E2=1 IF Zmin>zSmin AND Xmin>xSmin AND Ymin>ySmin ELSE E2=0 IF Zmin
Verification V1: Verification workspace

V2: Probes φPmin, lPmax

xVmax, yVmax, zVmax E5=1 IF ZmaxzVmax OR Xmax>xVmax OR Ymax>yVmax

E3=1 IF α≥αSmax ELSE E3=1(1/(1+(((α αSmin)/(αSmax αSmin))/0.5)4) IF αSmin< α<αSmax ELSE E3=0 IF α<αSmin E4=1 IF MIN(Xmin, Ymin)>3•MIN(xSmin, ySmin) ELSE E4=0.5 IF MIN(xSmin, ySmin)<MIN(Xmin, Xmin)<3•MIN(xSmin, ySmin) ELSE E4=0 IF MIN(Xmin, Ymin)<MIN(xSmin, ySmin)

-

E6=0 IF dCmax>lPmax ELSE E6=1 IF MIN(xCmin, yCmin)≥5•φPmin ELSE E6=(MIN(xCmin, yCmin)2•φPmin)/3•φPmin IF 2•φPmin <MIN(xCmin, yCmin) <5•φPmin ELSE E6=0 IF MIN(xCmin, yCmin) ≤2•φPmin

54

2 The DGLs-CF – Introduction and Background

Table 2.8 Part of the knowledge matrix corresponding to the Rules floor RULES FLOOR

Characteristics Manufacturing M1 Features F1 R1: Maximum dimensions of the product must be minor than maximum dimensions of the building room F2 -

F3

M2 R2: Minimum dimensions of the product must be greater than minimum dimensions related to the presence of supports R3: The presence of overhangs must be evaluated considering the need for supports R4: The dimensions and depth of cavities must be compatible with the need of supports

Verification V1 R5: Maximum dimensions of the product must be minor than maximum dimensions of the measuring volume -

-

V2 -

-

R6: Dimensions and depth of cavities must be compatible with the characteristics of the probes

Generation of Actions and Related Costs Each rule could suggest actions, referred to design, manufacturing or verification. These actions were labelled as must or hints, as follows. Must represented activities that had to be performed to generate or measure the product; hints were suggestions that the designer, the manufacturer or the inspector could follow to improve the compatibility between the product and the processes. Each must or hint had a qualitative estimated cost associated, expressed in the interval [0..10]. Table 2.9 reports the actions generated by analysing the rules. Each action was classified by domain — design, manufacturing, or verification — and the cost and the link to a coupled action if present were added. Table 2.9 Actions generated by analysing the rules Label Origin Action A1 R1 Split the model to make dimensions compatible with the workspace A2 R1 Post-process to merge the split part and eventually finish the resulting surface A3 R1 Scale the model making it smaller if A6 has not been performed yet A4 R2 Over-dimension thin parts A5 R2 Post-process to make over-dimensioned parts thinner A6 R2 Scale the model making it bigger if A3 has not been performed yet A7 R3 Change the orientation of the product in the workspace to minimise the quantity of required supports A8 R3 Over-dimension the part considering overall dimensions instead of overhangs

Domain Design

Cost Link 8 A2

Manufact. 5 Design

A1

5

Design 5 Manufact. 10 Design 5

A5 A4

Manufact. 1 Design

5

A9

2.4 Third Release of the DGLs. Thinking Big

55

Table 2.9 (continued) A9 A10

R3 R4

Post-process to make the overhangs from the bulk Change the orientation of the product in the workspace to make easier the support removal, evaluating the new orientation in comparison to that eventually resulting from A7. In case of conflict, see A11 Split the model to avoid the need for supports Post-process to merge the split parts and eventually finish the surface Change the orientation of the product in the CMM volume Avoid requiring the measurement of inaccessible zones of the product Tolerance the product so that it is possible obtaining an estimation of the characteristics of inaccessible features indirectly Define an adequate procedure to deduce indirect measurement Change the orientation of the product in the measuring volume to get better accessibility

A11 R4 A12 R4 A13 R5 A14 R6 A15

R6

A16 A17

R6 R6

Manufact. Manufact.

10 1

A8 -

Design Manufact.

8 5

A12 A11

Verification 2 Design 3

-

Design

5

A16

Verification 3 Verification 2

A15 -

Now the content of the section of the knowledge matrix related to the Design domain floor, the Manufacturing domain floor and Verification domain floor appeared as shown in Tables 2.10, 2.11 and 2.12 respectively. Table 2.10 Part of the knowledge matrix corresponding to the Design domain floor DESIGN DOMAIN FLOOR Features F1

Characteristics Manufacturing M1 A1: Split the model to make dimensions compatible with the workspace (linked to A2) A3: Scale the model making it smaller if A6 has not been performed yet F2 -

F3

M2 A4: Over-dimension thin parts (linked to A5) A6: Scale the model making it bigger if A3 has not been performed yet A8: Over-dimension the part considering overall dimensions instead of overhangs (linked to A9) A11: Split the model to avoid the need for supports (linked to A12)

Verification V1 V2 -

-

-

-

A14: Avoid requiring the measurement of inaccessible zones of the product A15: Tolerance the product so that it is possible obtaining an estimation of the characteristics of inaccessible features indirectly (linked to A16)

56

2 The DGLs-CF – Introduction and Background

Table 2.11 Part of the knowledge matrix corresponding to the Manufacturing domain floor MANUFACTURING Characteristics DOMAIN FLOOR Manufacturing M1 Features F1 A2: Post-process to merge the split part and eventually finish the resulting surface (linked to A1) F2 -

F3

-

Verification M2 V1 A5: Post-process to make overdimensioned parts thinner (linked to A4) A7: Change the orientation of the product in the workspace to minimise the quantity of required supports A9: Post-process to make the overhangs from the bulk (linked to A8) A10: Change the orientation of the product in the workspace to make easier the support removal, evaluating the new orientation in comparison to that eventually resulting from A7. In case of conflict, see A11 A12: Postprocess to merge the split parts and eventually finish the surface (linked to A11)

V2 -

-

-

2.4 Third Release of the DGLs. Thinking Big

57

Table 2.12 Part of the knowledge matrix corresponding to the Verification domain floor VERIFICATION Characteristics DOMAIN FLOOR Manufacturing M1 M2 Features F1 -

F2 F3

-

-

Verification V1 A13: Change the orientation of the product in the CMM volume -

V2 -

A16: Define an adequate procedure to deduce indirect measurement (linked to A15) A17: Change the orientation of the product in the measuring volume to get better accessibility

The knowledge matrix started to have a structure so articulated as to suggest the use of some tool to manage the information content, in order to ensure consistency and validity. As what happens in 3D modelling by feature, where any modification to the model must be monitored and validated to keep the model valid (Shah and Mäntylä 1995), or in the DBMS — Data Base Management Systems — where there are different types of integrity checks, here had been hypothesised the presence of a tool helping to assure the validity of the knowledge database during its setup and update. All of this could be automatic, using a sort of consistency monitor based on the application of a set of validation rules to ensure correct data management. These validation rules, simple to be defined, inserted and integrated in the system, could help in increasing the DGLs flexibility. To clarify the role of the monitor, here are two examples of validation rules. • Validation Rule 1 For every action in the Design domain floor defined as coupled with another, the corresponding action in the Manufacturing domain floor or in the Verification domain floor must exist and it had to show the same backward link. • Validation Rule 2 The expressions to calculate compatibility values must cover all possible situations. In the end, the hypothesis about the introduction of this consistency monitor never found a real implementation; nevertheless the requirements about the database correctness remain and they will be kept under consideration in future work.

58

2 The DGLs-CF – Introduction and Background

Configuration Phase Quantification of the Technological Parameters. Given the brands and models of the available manufacturing and verification technologies, the parameter values of the technological characteristics were set as follows, and inserted in Table 2.7: • • • •

xMmax=200 mm, yMmax=200 mm, zMmax=300 mm xSmin=2 mm, ySmin=2 mm, zSmin=1 mm, αSmin=45°, αSmax=120° xVmax=600 mm, yVmax=600 mm, zVmax=550 mm φPmin=1 mm, lPmax=20 mm

Usage Phase Product Characterisation This step was similar to the previous one but here the parameters associated with the product features were set, and Table 2.7 was updated accordingly. The parameter values for this case study were as follows: • • • • •

Xmin=4 mm, Ymin=15 mm, Zmin=3 mm Xmax=34 mm, Ymax=316 mm, Zmax=30 mm α=90° MIN(xCmin, yCmin)=4 mm dCmax=9 mm

Calculation of the Compatibility Values and Knowledge Activation Given the values of all the parameters involved, expressions generated in step 4 determined the compatibility values for each entry of the knowledge matrix where they were present. Based on these values, the DGLs were able to activate the only pieces of information meaningful for the situation, i.e. actions tagged as must when the compatibility value was equal to zero and hints otherwise, except for the value equal to 1, meaning full compatibility — no need for any must or hints. Table 2.13 reports the compatibility values and the list of the activated actions, classified as must or hints, for this case study. Generation of the Reconfiguration Packages Must and hints were now aggregated to generate a set of reconfiguration packages where they were arranged in order to avoid discrepancies or repetitions and to allow the users to choose among them. If all the compatibilities resulted in non-zero values, packages contained only hints to improve product compatibility. Each package had two associated costs, referring to the application of the must and the hints respectively, and a value representing the compatibility value obtained after the adoption of the must (if any).

2.4 Third Release of the DGLs. Thinking Big

59

Table 2.13 Compatibility values and activated actions, classified as must or hints Compatibility value Description E1=0 (due to y max dimension)

E2=1

E3=0.674 (due to the presence of overhangs) E4=0.5 (due to x min dimension)

E5=1

E6=0.66 (due to x min dimension)

Activated Knowledge (must and hints)

Compatibility between model maximum dimensions and Workspace dimensions

Total cost 13

Must: Split the model (Design) Must: Post-process to merge the split part and eventually finish the resulting surface (Manufacturing) Must: Scale the model making it smaller 5 (Design) Compatibility between model minimum dimensions and need for supports Compatibility between Hint: Change the orientation of the product in 1 model overhangs and the workspace to minimise the quantity of required supports (Manufacturing) need for supports Hint: Change the orientation of the product to 1 Compatibilities between cavities and make easier the support removal, evaluating the new orientation in comparison to that need for supports eventually resulting from the application of the previous hint. In case of conflict, Split the model (Manufacturing) Compatibility between model maximum dimensions and measuring volume Compatibility between Hint: Change the orientation of the product in 2 the measurement volume to get better cavities and probes accessibility (Verification) characteristics

In this case study the DGLs generated three reconfiguration packages. Table 2.14 shows them, with must and hints distributed over the three domains — design, manufacturing, and verification. Table 2.14 The set of reconfiguration packages generated by the DGLs PACKAGE 1 Design domain floor Manufacturing domain floor

Must Split the model

Hints -

Post-process to merge the Change the orientation of the product in the workspace to minimise the quantity of required split part and eventually finish the resulting surface supports Change the orientation of the product to make easier the support removal, evaluating the new orientation in comparison to that eventually resulting from the application of the previous hint. In case of conflict, Split the model Verification domain Change the orientation of the product in the floor measurement volume to get better accessibility Total costs 13 4 Compatibility value after must application 1 (Average)

60

2 The DGLs-CF – Introduction and Background

Table 2.14 (continued) PACKAGE 2 Design domain floor

Must Scale the model making it smaller Over-dimension thin parts Manufacturing Post- process to make the domain floor over-dimensioned parts thinner Verification domain floor Total costs 20 Compatibility after must application (Average) PACKAGE 3 Design domain floor

Must Split the model Scale the model, making it bigger Manufacturing Glue the split parts and domain floor eventually finish the resulting surface Verification domain floor Total costs 18 Compatibility after must application (Average)

Hints Change the orientation of the product in the measurement volume to get better accessibility 2 1 Hints Change the orientation of the product in the measurement volume to get better accessibility 2 1

User Choice of a Reconfiguration Package and Implementation. Based on costs, resources availability, etc., the DGLs user could choose the package that best fitted the surrounding conditions and proceeded to its implementation. In this case, the chosen package was the number 1 and the model resulting from the application of its must and hints is shown in Fig. 2.14. Figure 2.14a is a screenshot of the FDM workspace setup; Fig. 2.14b shows the generation of the physical prototype of the split parts and, finally, Fig. 2.14c,d are pictures of the verification phase after the gluing action.

2.4.5 Discussion Positive Outcomes In this release of the DGLs the conceptual diagram was much better than before. Now it was clear how knowledge was generated, the cause-effect paradigm, the information classification in the different domains, etc. It was possible to extract, consult and print the must and hints related to each domain. For example, inspectors could easily extract the list of the verification actions that were meaningful from their point of view.

2.4 Third Release of the DGLs. Thinking Big

61

Fig. 2.14 The product adequately modelled and oriented inside: a the virtual manufacturing workspace; b the real manufacturing workspace; c, d the verification workspace

Here rules were not related to the moment when they were applied. So they could be derived in advance from technological characteristics and product features, during the DGLs setup and before the DGLs adoption. This structure let us separate and distinguish clearly these two different phases in using the DGLs. Moreover, the presence of the parameters in the Compatibility floor, and the concept of activated knowledge, made the DGLs a very flexible tool. Changing the available manufacturing or verification technologies required no modifications of the DGLs structure but just some modifications in the parameter values from the point of view of the knowledge content. The consistency monitor concept would have ensured that all of this occurred correctly. This release of the DGLs better exploited the ISO GPS adoption. For example, the ISO GPS concept of linking design and manufacturing/verification was applied by introducing the links between couple of actions placed on different floors.

62

2 The DGLs-CF – Introduction and Background

Finally, this release of the DGLs allowed the comparison of different product configurations as there was a quantification of the compatibility between the product features and the technological characteristics, together with some quantification associated with the reconfiguration packages. Relating to the categories of problems and open issues of the previous releases of the DGLs, here something positive has been done about the Conceptual diagram and knowledge organisation, the DGLs adoption process and the ISO GPS adoption. However, the experience described above gave some impressions of the presence of criticisms even of this revised version of the DGLs, and they appear in the following, as usual. Conceptual Diagram and Knowledge Organisation • The DGLs building was now clean and it represented quite well the knowledge organisation and the inference process. However some misunderstanding still remained regarding the Compatibility floor and the Rules floor. The compatibility is intrinsically a rule attribute so it should have been better to put these pieces of information on the same floor. Knowledge Description • Must and hints were concepts used here in the wrong way. It seemed that actions were classified before knowledge activation and this is a conceptual mistake. This classification made the knowledge description more confused and, in the end, it appeared useless because its meaning was obvious. This is why it disappears in the DGLs-CF. • Nothing really meaningful has been done here regarding the characterisation of products and processes. Again, the ISO GPS adoption was not exploited from this point of view. For example, tables containing the information about the adoption context for the DGLs were filled using non-homogeneous terminology and this made the inference process hard to perform or, even more, to implement. Costs • Here there was a second attempt to consider costs during the redesign and reconfiguration process. Their management appeared quite organic and coherent; however, the big challenge consisted of the determination of them. This topic has not been managed at all in this release of the DGLs. Cost values were set in a relative way only to allow the comparison among the different reconfiguration packages generated at the end of the DGLs adoption process.

Summary

63

Implementation/Automatisms • The data structures were not optimised for any implementation. Moreover, there was no attempt to highlight potential modules in order to introduce and/or increase the automatisms of the DGLs. ISO GPS Adoption • As said before, ISO GPS adoption was not exploited even in this release of the DGLs. One of the most important aspects of the ISO GPS standards is that they suggest a vocabulary for describing the pieces of information. Unfortunately, this aspect was not taken into consideration but it should have been. DGLs Adoption Process • Even if the DGLs roadmap was of great help in understanding the knowledge generation and inference, some problems relating to the DGLs adoption process still remained. For example, the generation of the reconfiguration package was an intrinsically iterative process and this was not clear enough and sufficiently highlighted in the DGLs roadmap. DGLs Architecture • Two issues started to be really clear from this release of the DGLs: the drawbacks coming from the bottom-up approach of its development and the need for a formalism to describe both the development and the adoption of the DGLs. For sure the DGLs roadmap helped with the comprehension but it was not enough for considering and highlighting the actors involved, the input/output of the phases and of the single activities, the tools used time and time again during the DGLs adoption, etc.

Summary This chapter has described the work preceding the DGLs-CF development. The starting point has been the desire for generating a DfX method, to be used as the basis for going ahead with some new concepts such as the concurrent redesign and reconfiguration of product and process, with the help of emerging standards such as the ISO GPS. Three releases of the DGLs allowed us to highlight some important issues and to indicate the way ahead for the development of the DGLsCF. This will be described in detail in the next chapters.

64

2 The DGLs-CF – Introduction and Background

References Agarwala M, Bourell D, Beaman J, Marcus H, Barlow J (1995) Direct Selective Laser Sintering of Metals. Rapid Prototyp J 1(1):26-36 Bandera C, Cristofolini I, Filippi S, Toneatto G (2004) The definition of the geometrical characteristics of products in a knowledge based system for industrial design (DGL-Design GuideLines) - exploring the possibility of introducing ISO GPS (Geometrical Specification of Products) concepts. In: Proc. XIV ADM - XXXIII AIAS, Bari (I), ISBN: 88-900637-3-4 Bandera C, Cristofolini I, Filippi S (2005) Customising a Knowledge-Based System for design optimisation in Fused Deposition Modelling RP-technique. In: CISM Courses and Lectures 486, Advanced Manufacturing Systems and Technology, Kuljanic E. (ed.), Springer, Wien New York: 486:607-616 Bandera C, Cristofolini I, Filippi S (2006) Using a Knowledge-Based System to Link Design, Manufacturing, and Verification in a Collaborative Environment. In: Proc. CIRP ICME '06, Ischia - (Napoli) - I:201-206 Bosch JA (1995) Coordinate Measuring Machines and Systems. Marcel Dekker Inc. Cristofolini I, Filippi S, Bandera C (2006) How Rapid Prototyping Process Parameters could affect the Product Design Phase: a KBS approach. In: Proc. 2006 ASME International Design Engineering Technical Conferences & Computers and Information In Engineering Conference DETC 2006, Philadelphia, Pennsylvania, USA - ISBN 0-7918-3784-X Filippi S, Cristofolini I (2007) The Design Guidelines (DGLs), a Knowledge Based System for industrial design developed accordingly to ISO-GPS (Geometrical Product Specifications) concepts. Res Eng Des 18(1):1-19 Filippi S, Bandera C, Toneatto G (2001) Generation and Testing of Guidelines for Effective Rapid Prototyping Activities. In: Proc. ADM International Conference on Design Tools and Methods in Industrial Engineering, Italy:A2.18-A2.27 Gatto A and Iuliano L (1998) Prototipazione rapida: la tecnologia per la competizione globale. Tecniche Nuove, Italy Haffey MKD and Duffy AHB (2000) Knowledge Discovery and Data Mining within a Design Environment. In: Proc. Fourth IFIP WG 5.2 Workshop on Knowledge Intensive CAD, Cugini U, Wozny M (eds.), 4:72-87 Jacobs PF (1995) Stereolithography & Other Rp&m Technologies: From Rapid Prototyping to Rapid Tooling. Society of Manufacturing Engineers Kumaran N and Chittaro L (eds.) (1998) Reasoning About Function, Artificial Intelligence in Engineering. ELSEVIER Kuzman K, Nardin B, Kovac M, Jurkosec B (2001) The integration of Rapid Prototyping and CAE in mould manufacturing. J Mater Process Technol111:279-285 La Trobe-Bateman J and Wild D (2003) Design for manufacturing: use of a spreadsheet model of manufacturability to optimize product design and development. Res Eng Des 14(2):107-117 Nelson C (2002) RapidSteel 2.0 Mold Inserts for Plastic Injection Molding. © by DTM Technology Nielsen J (1993) Usability Engineering. Academic Press, Cambridge, MA Otto K and Wood K (2000) Product Design Techniques in Reverse Engineering and New Product Development. Prentice Hall, New Jersey Shah JJ and Mantyla M (1995) Parametric and Feature-Based Cad/Cam: Concepts, Techniques, and Applications. John Wiley & Sons, New York (USA)

3 Detailed Description of the DGLs-CF

The current release of the DGLs, named DGLs-CF, requires a more articulated description than the previous ones because of the introduction of many new ideas. The first of these is the change of name. As said before, hereafter the project will be called Design GuideLines Collaborative Framework to emphasise the synergy among the several actors involved. The second is always related to the actors. Now there is a new one, the software developer. The project has become mature enough to think about some sort of systematic implementation and for this reason the architecture of the DGLs-CF presents seven modules with clear data structures and interfaces. All of this has been possible because this new release was created in a completely different way in comparison to the previous ones. Here the approach has been strongly top-down; the overall architecture came first, then all the modules, data structures and detail activities have been defined and developed accordingly. Another new idea is that the roadmap introduced by the third release of the DGLs, revised and integrated, was so complex that some formalism to describe it became necessary. Two candidates were considered: UML — Unified Modelling Language (Ambler 2005; Eriksson et al. 2004) and IDEF — Integration DEFinition. Even if the UML appeared the best one from the DGLs-CF implementation point of view, the simplicity of IDEF0 has been preferred to get immediate comprehension even for people not expert in formalisms. As before, this chapter starts with the description of the conceptual diagram. Here the term knowledge matrix has been abandoned because the knowledge base structure is too complex; the pieces of information in the DGLs-CF database are introduced and described when their role appears meaningful in the DGLs-CF roadmap. Then the chapter proceeds with a short introduction of the IDEF0 formalism and with some notes about the new DGLs-CF roadmap. After that, the detailed description of the activities and the modules takes place. Unlike the previous releases, the discussion will not be presented in this chapter; the next chapter will describe some case studies and the following one, consisting of the discussion about results and open issues, will close the book.

66

3 Detailed Description of the DGLs-CF

3.1 Conceptual Diagram The DGLs-CF building, the way the knowledge inside the DGLs-CF is organised, is very similar to that of the third release of the DGLs. The only difference is that the old Rule floor disappeared because rules and compatibility expressions are now together on the Compatibility floor. In this way the structure of the Conceptual diagram is clearer and there is more congruity with the data structures inside the modules. Figure 3.1 shows the new DGLs-CF building.

Fig. 3.1 Conceptual diagram of the DGLs-CF, the DGLs-CF building

3.2 IDEF0 Fundamentals This section presents an overview of the fundamentals of the IDEF0 formalism, used here to describe the DGLs-CF roadmap. IDEF is a family of modelling languages mainly used in the field of software engineering. In this family, IDEF0 is a graphical modelling method to produce a complete functional model of an organisation, system or process. The result of an IDEF0 representation is a hierarchical functional decomposition. There are five basic elements in an IDEF0 functional model: Inputs (I), Controls (C), Outputs (O), Mechanisms (M) and activities (or functions). Inputs indicate parameters that are modified by an activity, while outputs represent the results of it. Controls indicate factors that constrain an activity, while mechanisms are the tools used to perform it. As shown in Fig. 3.2, activities are represented by boxes, usually named using active verbs

3.3 Purpose, Viewpoint and the Node A-0 of the DGLs-CF IDEF0 Diagram

67

that explain the activity accomplishment. The ICOM constraints (Inputs, Controls, Outputs, Mechanisms) are represented by arrows: Input data enter from the left side of a box; Output data exit from the right side; Controls enter from the top and Mechanisms from the bottom.

Fig. 3.2 IDEF0 basic entities

Usually, the first thing to do when developing an IDEF0 model is to establish the Purpose and the Viewpoint of the model. These are two key-elements that influence all the following steps. Then the description of the process takes place, and a context diagram containing the top level activity — A0 context — is modelled together with its ICOMs. In the next step, the A0 context is broken down into multiple activity diagrams that define the process with the activity breakdown and the introduction of the appropriate ICOMs (FIPS PUB 183 1993; IEEE 1320.1 1998).

3.3 Purpose, Viewpoint and the Node A-0 of the DGLs-CF IDEF0 Diagram The DGLs-CF has been developed using a top-down approach. For this reason, the IDEF0 diagrams are not a synthesis of already existing modules and activities, but the map of a procedure that has been developed exactly as represented by the hierarchical structure formalised using IDEF. At the beginning, the DGLs-CF was considered as a black-box, in order to focus the attention on the analysis of the interfaces with its context of use. After that, the DGLs-CF roadmap has been developed step by step, always maintaining the same abstraction level inside each step. The result is a clean architecture where the modules are correctly placed and

68

3 Detailed Description of the DGLs-CF

interfaced with each other, and where the role of the external systems — human and not — are correctly identified.

3.3.1 Purpose and Viewpoint As previously stated, the starting point for the generation of an IDEF0 diagram is the definition of its Purpose and Viewpoint. In this case, the Purpose consists of describing the DGLs-CF roadmap as clearly as possible to all the actors, current and future, involved in the project. The Viewpoint is mainly the developer’s one, focusing on the characterisation of the modules used by the activities of the roadmap during the redesign/reconfiguration process.

3.3.2 The Node A-0 — TOP Level Figure 3.3 shows the IDEF0 diagram of the node A-0 — TOP level — of the DGLs-CF roadmap. This level contains only the overall activity A0 and it is important because it allows focusing on the main ICOM components of the whole process at a glance, even if they are used in different stages of the DGLs-CF adoption. These components are briefly described in the following. Input • Class of Manufacturing Technologies and Class of Verification Technologies They describe the type of technologies considered in the context of the DGLs-CF adoption; for example, the FDM or the CMM. This description is based on a specific data type, named parametric technological characteristic. • Class of Products It describes the kind of products that will be managed during the DGLs-CF adoption; for example, mechanical parts. This description is based on the data type named parametric product feature. • Specific Manufacturing Technology and Specific Verification Technology These items collect all the pieces of information related to the brands and models of the available technology; for example, the FDM equipment Stratasys Dimension (http://www.dimensionprinting.com), or the CMM equipment DEA Global Image 07-07-07 (http://www.dea.it). • Specific Product (Model) This item is the same as the previous one, but it is related to the product under evaluation each time. For example, the plastic shell of the left headlight of a specific car model. The term model is inside brackets because it is present only when considering design and manufacturing issues. The verification deals with physical parts only, so the model of the product is not considered.

3.3 Purpose, Viewpoint and the Node A-0 of the DGLs-CF IDEF0 Diagram

69

Fig. 3.3 Node A-0 - TOP level - of the IDEF0 diagram

Controls • ISO GPS Standards The parametric technological characteristics and the parametric product features are defined obeying the activities suggested by the ISO GPS standards. For this reason they appear as Controls in the IDEF0 diagrams of the DGLs-CF roadmap. Output • Redesign/Reconfiguration Packages This item represents the goal of the DGLs-CF adoption. At the end of the elaboration, the DGLs-CF generates the redesign/reconfiguration packages, lists of actions to redesign the product and to reconfigure the process, in order to optimise product manufacturing and verification, given the available technologies. Mechanisms • Designers, Manufacturers, Inspectors These are the actors that in different ways and at different stages contribute to the DGLs-CF adoption as explained in detail in the following. Seven Modules The other mechanisms used by the activities of the roadmap are the seven modules presented in the DGLs-CF. Their description is postponed to

70

3 Detailed Description of the DGLs-CF

the next paragraphs because their role is more easily comprehended if classified and associated with the phases where they are used.

Fig. 3.4 The node A0 of the IDEF0 diagram of the DGLs-CF roadmap

3.4 The Node A0. The Main Phases and the Modules

71

3.4 The Node A0. The Main Phases and the Modules Figure 3.4 shows the node A0 of the IDEF0 diagram. This level is important because it shows the three main phases of the DGLs-CF, derived from the third release of the DGLs. At the same time, this diagram presents all the pieces of information exchanged among the phases or, in other words, their interfaces. Two reasons explain why the DGLs-CF adoption comes in three phases: the first is that they are reasonably well executed sequentially, reflecting the evolution of the application context; the second refers to the need of highlighting the different actors and modules involved in each of them.

3.4.1 Overview of the Three Main Phases What follows is a brief description of the three main phases of the DGLs-CF roadmap. First Setup In the First Setup (A1), the DGLs-CF is customised, given the class of available manufacturing and verification technologies. Clearly, this phase is executed only once, at the adoption of the DGLs-CF, or when the technologies heavily change. Designers, manufacturers and inspectors work together in generating the knowledge base used afterwards during product redesign and process reconfiguration. Technological Configuration The Technological Configuration (A2) sets the parameter values of the manufacturing and verification characteristics, given the brands and models of the available technologies. This phase is executed every time the available technologies are updated. Only manufacturers and inspectors are involved here. Redesign/Reconfiguration Package Generation Finally, the Redesign/Reconfiguration Package Generation (A3), executed for each product, generates the lists of actions — the redesign/reconfiguration packages — to be applied to the product (model) and to the manufacturing/verification processes. Here all the actors are involved again; in fact, as will be shown in the following, the generation of the packages is based on an iterative procedure that requires user-assisted modifications to the product (model).

72

3 Detailed Description of the DGLs-CF

3.4.2 Overview of the Seven Modules Previous releases of the DGLs showed a knowledge base distributed on a collection of tables, without an organic and homogeneous structure. Moreover, it was practically impossible to recognise the logical elaboration units associated with the tables. All of this generated misunderstandings and there were some great difficulty in thinking about any real implementation of the DGLs. To answer these problems, the DGLs-CF roadmap is based on a collection of seven modules that introduces a strict organisation of the pieces of information involved in the elaboration and highlights the elements that could be implemented. Figure 3.5 introduces these modules, classified by the phase where they are used, and collects the data structures inside each of them. Now these modules show a different development level; some of them have been already partially implemented, while for others only the procedures exist. The last one is an empty box because it is needed for the generation of the packages but for now all its activities are performed manually.

Fig. 3.5 The seven modules involved in the DGLs-CF roadmap

What follows is an overview of these modules and of their role and meaning. Details will be shown during the description of the roadmap activities where they are involved. MOD1. Characteristic and Feature Collecting Module This module allows the generation of the knowledge base used throughout the DGLs-CF roadmap. Given the classes of manufacturing and verification technologies and the class of products, three tables are filled with the parametric technological characteristics and the parametric product features. These items are

3.4 The Node A0. The Main Phases and the Modules

73

defined as parametric because they will be customised to specific technologies and products only in the next phases. MOD2. Rule and Action Generation Module This module is used to generate the relationships between technological characteristics and product features. These relationships are expressed by rules, and for each rule one or more actions for respecting it are developed. At the end of the activities the tables of this module contain the whole sets of rules and actions. MOD3. Feature Relationship Discovery Module The development of the previous releases of the DGLs highlighted the importance of considering possible relationships among product features during the redesign/reconfiguration process. The concept is simply that the modification of a product feature may impact on others, sometimes in an unexpected way. Starting from this, a procedure based on an inference algorithm has been developed, to discover these relationships. Moreover, this information is used effectively for the generation of the redesign/reconfiguration packages. At the end of the elaboration, the table of this module contains all the information needed by the redesign/reconfiguration package generation procedure. MOD4. Characteristic Data Input Module This module is used to manage the section of the DGLs-CF knowledge base related to the information on the available manufacturing and verification technologies. Each parametric characteristic is customised using the real data representing the brand and model of these technologies. The two tables of this module are filled accordingly. MOD5. Feature Data Input Module This module acts exactly as the previous one, but it focuses on the specific product (model) to be evaluated using the DGLs-CF. The reason why the two modules are set in separate phases is that their usage happens at different stages of the DGLsCF roadmap. This module is used every time a new product is evaluated; the previous one must be considered only when the available manufacturing and/or verification technologies should change. The feature parameters assume the values derived directly from the product (model) and are collected in a table. MOD6. Redesign/Reconfiguration Package Generation Module The generation of the redesign/reconfiguration packages is not straightforward. There are so many degrees of freedom, in terms of different combinations of the possible actions to perform, of required skills to compose and implement the

74

3 Detailed Description of the DGLs-CF

different packages, of relationships among product features, etc., that different procedures with different refinements have been developed up to now. The first one considered all the possible combinations of the activated actions (Filippi and Cristofolini 2007), while newer and more sophisticated ones also consider other variables, for example the relationships among product features. This module contains the latest version of the procedure, in which all the variables are considered, and iteration is introduced. The tables contain the compatibility values of the current state of the product (model), the set of actions that are meaningful for it — activated actions — and the result of the DGLs-CF elaboration — the redesign/reconfiguration packages. MOD7. Most Dynamic Action Simulation Module As said before, at present this module is an empty box because it only represents the activities needed to simulate the modifications to the product (model) during the iterative procedure that generates the redesign/reconfiguration packages. Designers, manufacturers and inspectors are required to simulate these activities by themselves. This has not to be seen as a drawback, because the main goal of this project is the development of a methodological approach to cover all the steps of the redesign/reconfiguration process, and not the implementation of a tool for doing it automatically. This is the reason why all the slots of the framework have been planned and foreseen, while not all of them have been already developed and/or implemented.

3.5 The Node A1. First Setup The purpose of this phase is the first setup of the DGLs-CF. Thinking about a real scenario, this phase is executed only once, as soon as the decision about the DGLs-CF adoption is taken. The purpose of this phase is to give the DGLs-CF all the information related to the technological context where it is used and about the type of the products to manage. An analysis of the available manufacturing and verification technologies and of the products allows the DGLs-CF to obtain the knowledge needed to generate the final redesign/reconfiguration packages. The information gathered is used, always during this phase, to derive some other pieces of information as rules, actions, relationships among product features, etc., as described in the following. Figure 3.6 shows the activities of the First Setup phase. The description of all of them follows in detail. At the end, the modules involved in these activities are described as well.

3.5 The Node A1. First Setup

Fig. 3.6 The activities of the First Setup phase

75

76

3 Detailed Description of the DGLs-CF

3.5.1 Activities A11. Select the Manufacturing Characteristics, and A12. Select the Verification Characteristics, and A13. Select the Product Features Given the classes of manufacturing and verification technologies and the class of products to be evaluated, these activities collect the characteristics and the features, with related parameters, used to describe formally the technologies and the product. At this stage the characteristics and the features are called parametric because the parameter values are not yet set. The module MOD1, named Characteristic and feature collecting module, is fundamental here because it allows the knowledge gathering by exploiting the description language offered by the ISO GPS. The results of these three activities are placed in the Compatibility floor of the DGLs-CF building. Manufacturers, inspectors and designers are the actors involved in these three activities respectively. A14. Generate the Rules This activity generates the rules by considering parametric technological characteristics and parametric product features. The rules describe the limitations of the technologies related to each product feature. Now the DGLs-CF does not allow exploiting the capabilities of the technologies because the data structures and the procedures cannot manage the positive rules described in the previous chapter. This possibility will be recalled in the last chapter, together with other hints for future work. The rules are coupled with the expressions needed to evaluate the compatibility of the current version of the product (model) with the available technologies. The rules are labelled to remember their origin — which manufacturing or verification characteristics are evaluated with respect to which product features — and placed in the Compatibility floor of the DGLs-CF building. The module MOD2, named Rule and action generation module, helps in collecting all these pieces of information by establishing the syntax for encoding the rules here and the actions in the next activity. Regarding the actors involved here, designers, manufacturers and inspectors work together, exploiting their knowledge and skill in synergy.

3.5 The Node A1. First Setup

77

A15. Generate the Actions The rules generated in the previous activity suggest actions to be executed to overcome the limitations of the technologies. The actions are classified by domain and placed in the Design, Manufacturing, and Verification domain floor of the DGLs-CF building. Some actions necessarily require other actions, and this information is also present in the DGLs-CF knowledge base, expressed as a link between them. As required and controlled by the module MOD2, all the actions are expressed using the same verb-accusative-goal pattern because this is required for the correct execution of the next activities. As before, all the actors involved in the DGLs-CF adoption are present here. A16. Discover the Relationships Among Product Features DGLs-CF users should be warned of the side-effects of the modification of a product feature required by the execution of an action. Moreover, the DGLs-CF should consider these interactions among product features in an automatic way, during the generation of the redesign/reconfiguration packages. This topic has received a lot of attention in previous works and now the DGLs-CF contains a clear and effective procedure to discover the relationships among product features, given the classes of technologies (Cristofolini et al. 2006, 2008; Cristofolini and Filippi 2008). The module MOD3, named Feature relationship discovery module, implements this procedure and fills the proper table of the DGLs-CF knowledge base with the results of its elaboration. Once again, the details are presented in the description of the module in the following. No actors are expected for this activity; in fact, the data elaboration in the module MOD3 is automatic.

3.5.2 Modules of Interest Here MOD1. Characteristic and Feature Collecting Module The aim of this module is the acquisition and formalisation of the manufacturing and verification characteristics, of the features describing the product to be manufactured and verified and of the parameters related to them. As stated before, a big issue here is the adoption of some language to make the descriptions uniform and usable. In fact, very often designers, manufacturers and inspectors refer to products and technologies using different languages coming from their everyday terminology and practices. The result is that features and characteristics are based on parameters presenting a non-homogeneous definition or with some pieces of information left implicit. All of this leads to real difficulties in relating these parameters each other.

78

3 Detailed Description of the DGLs-CF

Regarding this, the DGLs-CF exploits in depth the ISO GPS features of size. Because of their intrinsic association with dimensions, they can be used to describe both features and characteristics; in this way the description languages become the same and they can be easily related. Moreover, the relationships themselves can be managed exploiting the ISO GPS standards. A coherent and univocal formalisation of features, relationships, operations etc., is fully described in ISO 14660-1:1999, ISO 14660-2:1999, ISO/TS 17450-1:2000 and ISO 174502:2002, as already mentioned in Chap. 1. Technological characteristics and product features used as input for the DGLsCF can be considered as a sort of hi-level entities, containing several chunks of information and defined by hi-level parameters, roughly associated to them. On the other hand, the ISO GPS features of size could appear as lo-level features, and for this reason they seemed the best building blocks for translating the hi-level entities towards a common language. The third level used for the characteristic and feature description in the DGLs-CF is the mid one. This is the most important level and its importance lies in the fact that all the entities at this level are described by mid-level parameters, generated from the lo-level ISO GPS features of size, and for this reason expressed using the same language. This was exactly the challenge that drove the development of this module. Figure 3.7 schematises how the hi-level description of characteristics and features is translated to mid-level thanks to the lo-level ISO GPS features of size. The procedures defined and used during the translation allow one to recognise the lo-level features in the description of the hi-level ones and to define the relationships driving to the final mid-level description.

Fig. 3.7 The role of the ISO GPS in the knowledge formalization of the module MOD1

The topological relationships among these features, needed to ensure a univocal representation of them, are obtained thanks to the ISO GPS situation features, which allow defining the location and/or the orientation of the features of size.

3.5 The Node A1. First Setup

79

ISO GPS situation features are used in the DGLs-CF to define a datum system integral with the product. Finally, another important reason why the ISO GPS features of size are the best way to describe the geometrical issues of features and characteristics is that they allow the management of dimensional tolerances. As an example of geometrical features described by way of the ISO GPS concepts, consider a plate with a feature of the type “Through hole”, as in Fig. 3.8. This feature may be described by the ISO GPS features of size cylinder, characterising its diameter and axis, and two parallel planes orthogonal to the axis, to characterise the depth of the hole. The ISO GPS situation features defining the location and orientation of the through hole are represented by the three planes that determine the datum system shown in the figure as xyz, needed for establishing the topological relationships. If the hole is inclined as in the right side of Fig. 3.8, it may be described again using the feature of size cylinder, characterising the diameter and axis, using two parallel planes orthogonal to the axis, to characterise the hole depth, and two planes defining a wedge, one of them containing the axis of the hole and the other vertical, respecting the datum system integral with the model. To enhance comprehension, on the right side of Fig. 3.8 the planes defining the wedge are not shown, and only two axes belonging to them are represented.

Fig. 3.8 Example of product feature description using the ISO GPS concepts

As anticipated before, a univocal representation of product features and technological characteristics is obtained by establishing specific procedures both to derive the ISO GPS features of size from the hi-level descriptions and to establish their orientation/location with respect to the datum systems integral with the models. For the sake of brevity, only the procedures related to a subset of product features and technological characteristics are detailed hereafter. The product features considered here are “Bounding box”, “Minimum dimensions”, “Overhangs/sloped surfaces”, and “Cavities”. This set allows describing the macro-geometry of most of the mechanical parts and, although they have been developed specifically relating to particular technologies, they may be generalised, being the critical ones for the most of the situations. Table 3.1 shows the description of these product features. The upper part of each box contains the general information of the feature; in the lower part, left to right, there are the hi-

80

3 Detailed Description of the DGLs-CF

level formalisation, the ISO GPS features of size and the specific procedures used for the translation and, finally, the resulting mid-level formalisation. The figures of the mid-level formalisation show some examples of the features of size derived by the application of the procedures; these features are represented by double dotted lines and allow one to recognise the related parameters, identified by a number. Table 3.1 Examples of product feature formalisation based on the ISO GPS concepts FEATURE HI-LEVEL formalisation (parameters) X, Y, Z (maximum dimensions)

Name: Bounding box Description: Overall dimensions of the product ISO GPS concepts used for the translation Features of size Procedure (LO-LEVEL) Two parallel Considering the datum system planes integral with the model, three pairs of parallel planes belonging to the model, each time orthogonal to x—y, x—z, z—y directions, are identified to find those showing the maximum distances (maximum bounding box dimensions). In case of curved surfaces, same as above with tangent planes

Name: Minimum dimensions Description: Minimum dimensions in the product HI-LEVEL ISO GPS concepts used for the translation formalisation Features of size Procedure (parameters) (LO-LEVEL) Considering the datum system x, y (minimum Two parallel planes integral with the model, three pairs dimensions in of parallel planes belonging to the horizontal plane) model, each time orthogonal to x— y, x—z, z—y directions, are z (minimum identified to find those showing the thickness) minimum distances (minimum dimensions). In case of curved surfaces, same as above with tangent planes. In case of pins, the diameter of the cylinder is evaluated (the diameter of the minimum pin in the part). The height of the pin is evaluated considering the distance between two parallel planes orthogonal to the axis delimiting the pin (height of the minimum pin)

MID-LEVEL formalisation (parameters and figure) Bounding_box_X, Bounding_box_Y (1), Bounding_box_Z (maximum dimensions of the product)

FEATURE

MID-LEVEL formalisation (parameters and figure) Minimum_dimensions_x, Minimum_dimensions_y (minimum dimensions in xy plane) Minimum_dimensions_z (1) (minimum dimension in z direction) Minimum_dimensions_φmin (minimum diameter) Minimum_dimensions_hmin (minimum height)

3.5 The Node A1. First Setup

81

Table 3.1 (continued) FEATURE

Name: Overhangs/Sloped surfaces Description: Protruding parts, both inside the component (undercuts) and outside (overhangs) MID-LEVEL formalisation HI-LEVEL ISO GPS concepts used for the translation (parameters and figure) formalisation Features Procedure (parameters) of size (LOLEVEL) Wedge Considering the datum system integral α Overhangs_Sloped_surfaces_α with the model, the planes belonging to (overhangs/sloped (1) (overhangs/sloped surfaces the model are evaluated with respect to surfaces angle) angle) the vertical planes belonging to the model, thus defining a wedge, in order to find the angles among them (angles defining Overhangs and Sloped Surfaces). In case of curved surfaces, same as above with tangent planes

FEATURE

Name: Cavities Description: Through/blind holes and slots HI-LEVEL ISO GPS concepts used for the translation formalisation Features of Procedure (parameters) size (LOLEVEL) In case of through/blind holes, considering xCav, yCav Cylinder the datum system integral with the model, (minimum dimensions) Two parallel the diameter of the cylinder is evaluated (the diameter of the minimum cavity in the part). planes In the case of non-cylindrical holes, the dCav minimum related dimensions in function of Wedge (maximum the orientation are evaluated instead. depth) The depth of the hole is evaluated considering the distance between two parallel planes orthogonal to the axis delimiting the hole (depth of the deepest hole). The inclination of the hole is evaluated considering the angle formed by the wedge between the plane containing the axis of the hole and the vertical plane (inclination of the hole). In case of slots, considering the datum system integral with the model, the distance between the two parallel planes defining the slot is evaluated. The depth of the slot is evaluated considering the distance between the two parallel planes orthogonal to the median plane delimiting the slot.

MID-LEVEL formalisation (parameters and figure)

Cavities_x, Cavities_y, Cavities_z (minimum dimensions for non-cylindrical cavities) Cavities_φ (1) (minimum diameter for cylindrical cavities) Cavities_d (2) (maximum depth for cylindrical cavities) Cavities_β (angle of inclination of the axis of the cavity)

82

3 Detailed Description of the DGLs-CF

Table 3.1 (continued) xCav, yCav (minimum dimensions) dCav (maximum depth)

The inclination of the slot is evaluated considering the angle formed by the Two parallel wedge between the median plane of the slot and the vertical plane planes (inclination of the slot) Wedge Cylinder

Regarding the examples of technological characteristics considered here, they are: “Manufacturing workspace” and “Supports” for manufacturing, and “Indexed measuring head” for verification. Table 3.2 shows their description. Table 3.2 Examples of manufacturing and verification characteristic formalisation based on the ISO GPS concepts CHARACTERISTIC HI-LEVEL formalisation (parameters) xM, yM, zM (maximum dimensions of the manufacturing workspace)

Name: Manufacturing workspace Description: Volume of the manufacturing workspace MID-LEVEL formalisation ISO GPS concepts used for the translation (parameters and figure) Features of Procedure size (LOLEVEL) Two parallel Considering the datum system Manufacturing_Workspace_x (1), planes integral with the model describing Manufacturing_Workspace_y, the manufacturing workspace, three Manufacturing_Workspace_z pairs of parallel planes belonging to (maximum dimensions of the the model, each time orthogonal to manufacturing workspace) x—y, x—z, z—y directions, are identified to find those showing the maximum distances (maximum bounding box dimensions). In case of curved surfaces, same as above with tangent planes

3.5 The Node A1. First Setup

83

Table 3.2 (continued) CHARACTERISTIC

Name: Supports Description: Critical characteristics related to the need for supports when building overhangs/sloped surfaces or cavities MID-LEVEL HI-LEVEL ISO GPS concepts used for the translation formalisation formalisation Features Procedure (parameters and figure) (parameters) of size (LOLEVEL) Considering the datum system integral Two α Supports_x, Supports_y, (overhangs/sloped parallel with the model describing the smallest part Supports_z (1) (minimum which is built needing supports, in case of dimensions related to the planes surfaces angle) cylindrical parts, the diameter of the need for supports) cylinder is evaluated. Wedge The depth or height of the cylinder is Supports_φ (minimum Cylinder evaluated considering the distance diameter related to the between two parallel planes orthogonal to need for supports) the axis delimiting the cylinder. The inclination of the cylinder is evaluated Supports_d (maximum considering the angle formed by the depth or height related to wedge between the plane containing the the need for supports) axis of the hole and the vertical plane. Supports_α (minimum In case of non-cylindrical parts, the angle related to support minimum related dimensions in function removal) of the orientation are evaluated instead. The depth or height of the part is evaluated considering the distance between the two parallel planes orthogonal to the median plane delimiting the part. The inclination of the part is evaluated considering the angle formed by the wedge between the median plane of the slot and the vertical plane CHARACTERISTIC

Name: Indexed measuring head Description: Possible rotation and inclination of the measuring head HI-LEVEL ISO GPS concepts used for the translation MID-LEVEL formalisation formalisation Features of Procedure (parameters and figure) (parameters) size (LOLEVEL) βV (angle of inclination Wedge Considering the datum system Indexed_measuring_head_β of the head) integral with the model (1) (minimum angle of describing the measuring head, inclination of the head) the inclination of the axis belonging to the probe is evaluated with respect to the vertical planes, thus defining a wedge, in order to find the minimum angle among them

84

3 Detailed Description of the DGLs-CF

Comparing the formalisations before and after the introduction of the ISO GPS concepts, thanks to the procedures exploiting the features of size now the description of the features and characteristics is much clearer and unique and the coherence with the ISO GPS is properly ensured. In particular, in the case of “Minimum dimensions”, ”Cavities”, and ”Supports”, the coarse definition of the parameters has been replaced by a more rigorous one, also implying the coverage of a wider range of situations. All the data generated in this phase are collected in the tables MOD1_T1, MOD1_T2 and MOD1_T3 of the DGLs-CF data structure, shown here by Table 3.3, Table 3.4 and Table 3.5 respectively, containing some examples related to the FDM manufacturing process, the CMM verification process and to mechanical parts. Table 3.3 Examples of parametric manufacturing characteristics — table MOD1_T1 of the DGLs-CF data structure Parametric manufacturing characteristic Label Name Description M1 Manufacturing Volume of the workspace manufacturing workspace M2

Supports

Critical characteristics related to the need for supports when building overhangs/sloped surfaces or cavities

Mid-level parameters Manufacturing_Workspace_x, Manufacturing_Workspace_y, Manufacturing_Workspace_z (maximum dimensions of the manufacturing workspace) Supports_x, Supports_y, Supports_z (minimum dimensions related to the need for supports) Supports_φ (minimum diameter related to the need for supports) Supports_d (maximum depth or height related to the need for supports) Supports_α (minimum angle related to support removal) Supports_Ra_xy (minimum obtainable roughness in xy plane)

Table 3.4 Examples of parametric verification characteristics — table MOD1_T2 Parametric verification characteristic Label Name Description Mid-level parameters V1 Verification Volume of the Verification_workspace_x, workspace verification workspace Verification_workspace_y, Verification_workspace_z (maximum dimensions of the verification workspace) V2 Indexed Possible rotation and Indexed_measuring_head_β (minimum measuring head inclination of the angle of inclination of the head) measuring head

3.5 The Node A1. First Setup

85

Table 3.5 Examples of parametric product features — table MOD1_T3 Parametric product feature Label Name Description F1 Bounding box Overall dimensions of the product F2

Pins, ribs, minimum dimensions

Mid-level parameters Bounding_box_X, Bounding_box_Y, Bounding_box_Z (maximum dimensions of the product) Minimum dimensions Minimum_dimensions_x, Minimum_dimensions_y (minimum in the product dimensions in xy plane) Minimum_dimensions_z (minimum dimension in z direction) Minimum_dimensions_φmin (minimum diameter) Minimum_dimensions_hmin (minimum height)

Items in the tables are specified using labels; Mx for manufacturing characteristics, Vx for verification characteristics, Fx for product features, being x consecutive Arabic numbers. The name and the description of each characteristic or feature are followed by the parameters that constitute the mid-level definition, obtained by the translation based on the ISO GPS concepts, as specified above. Some non-geometrical characteristics and features should be considered in this project too, for example those related to the mechanical behaviour or to the surface finishing of the product. Even in these cases, characteristics and features are defined congruently to allow the compatibility evaluation. Basically, it will be ensured that the parameters describing the technological characteristics will be congruent with the product features they affect. For example, if the need for supports affects the surface finishing, among the parameters describing the manufacturing characteristic “Supports” there will be one allowing a comparison with the surface roughness. This will be the parameter named Supports_Ra_xy of Table 3.3, representing the minimum obtainable roughness in the xy plane. MOD2. Rule and Action Generation Module In this module, parametric technological characteristics and parametric product features are related to each other, in order to evaluate the compatibility between the product (model) and the manufacturing and verification processes. This relationship is expressed by rules; each manufacturing and each verification characteristic is related to each product feature and a rule to determine their compatibility may be generated. For example, relating the manufacturing characteristic “Manufacturing workspace” and the product feature “Bounding box” allows the actors involved — in this case, designers and manufacturing experts — generating a rule establishing that the dimensions defining the bounding box of the product model must be smaller than the maximum dimensions of the manufacturing workspace. The relationships between technological characteristics and product features do not necessarily determine

86

3 Detailed Description of the DGLs-CF

rules; for example, “Manufacturing workspace” does not affect “Minimum dimensions”, so that no rule is established in this case. Table 3.6 Examples of rules — table MOD2_T1 Rule Label Description R_M1F1 Dimensions defining the bounding box of the product must be smaller than maximum dimensions of the manufacturing workspace

Procedure Locate the features of size defining the bounding box in the product model — gives the associated parameters Bounding_box_X, Bounding_box_Y, Bounding_box_Z. Identify the parameters defining the maximum dimensions of the building room — gives Manufacturing_workspace_x, Manufacturing_workspace_y, Manufacturing_workspace_z R_V1F1 Dimensions Locate the features of size defining the defining the bounding box in bounding box the product model — gives of the product the associated parameters Bounding_box_X, must be smaller than Bounding_box_Y, Bounding_box_Z. maximum dimensions of Identify the parameters defining the maximum the dimensions of the building verification room — gives workspace Verification_workspace_x, Verification_workspace_y, Verification_workspace_z

Compatibility expressions E_M1F1=1 IF Bounding_box_Z<Manufacturing_workspace_z AND Bounding_box_X<Manufacturing_workspace_x AND Bounding_box_Y<Manufacturing_workspace_y ELSE E_M1F1=0

E_V1F1=1 IF Bounding_box_Z
As shown in Table 3.6 — table MOD2_T1 — each rule is formalised using a label, a description, a procedure, and a compatibility expression. The label is expressed in the form R_MxFy or R_VxFy, being R for rule, Mx or Vx for the manufacturing or verification characteristic, and Fy for the product feature which the rule derives from. The description is simply an explicative sentence that helps in identifying the topics of the rule. The procedure establishes how the technological characteristics and the product features can be identified and related to the associated parameters, thus determining the parameters to be introduced in the compatibility expressions. Finally, the compatibility expressions allow quantifying the compatibility between the technological characteristics and the product features, in order to highlight the rules that are violated. Compatibility expressions are labelled by E_MxFy or E_VxFy, being E for expression and the other items with the same meaning as before. Expressions are defined in order to cover all the situations between the characteristic and the feature parameters. For

3.5 The Node A1. First Setup

87

example, relating the dimensions defining the “Bounding box” (F1) of the product model and the maximum dimensions of the “Manufacturing workspace” (M1), the compatibility expression is: E_M1F1=1 IF (Bounding_box_Z<Manufacturing_workspace_z AND Bounding_box_X<Manufacturing_workspace_x AND Bounding_box_Y<Manufacturing_workspace_y) ELSE E_M1F1=0 For now, these expressions may assume only two values: 1 when there is full compatibility and 0 if not. The definition of the rules drives to the definition of the actions to be performed when the rules are not respected. In other words, actions are defined to describe what to do when the compatibility value is equal to 0. As shown in Table 3.7 - table MOD2_T2 -, each action is formalised by defining the domain where it takes place, a label, the eventual link to other actions, a description, and a procedure to follow in performing the action. Table 3.7 Examples of actions — table MOD2_T2 Action Domain Design

Label

Link

Description Verb Accusative the model AD_M1F1 AM_M1F1 Use existing plane surfaces in splitting

Procedure Goal to make dimensions compatible with the manufacturing workspace

Locate the features of size related to those parameters giving E_M1F1=0 Locate the situation features (planes and/or axes and/or median planes) belonging to the product model and orthogonal to the features of size previously identified Split the model using planes containing these situation features

88

3 Detailed Description of the DGLs-CF

Table 3.7 (continued) Manufacturing AM_M1F1 AD_M1F1 Merge

the product

to join the split Locate the features of parts size related to those parameters giving E_M1F1=0 Locate the situation features (planes and/or axes and/or median planes) belonging to the product model and orthogonal to the features of size previously identified Merge the model using planes containing these situation features

The domain recalls one of the three floors of the DGLs-CF building where the actions must be executed. In the case of the Design domain, actions are labelled by AD_MxFy or AD_VxFy, being A for action, D for design domain, and the other items as before. Analogously, for the Manufacturing domain each action label has AM as prefix and for the Verification domain the prefix is AV. The link highlights whether an action necessarily implies another one; in that case the linked action is specified too. The description of each action is expressed in the form verbaccusative-goal, thus specifying what must be performed — the verb — on which object — the accusative — and why — the goal. The procedure explains how the action must be performed, that means how to identify the features defined in the description and how to use them when executing the action. MOD3. Feature Relationship Discovery Module Goal of this module is to help in finding the possible relationships among the product features; in fact, this is considered a really important issue and the results are used in the generation of the redesign/reconfiguration packages. In other words, the DGLs-CF cannot ignore that a modification in the value of some product feature parameter could influence other product features not directly connected to it or not directly considered in that specific moment. The procedure is presented in the following; how all of this impacts on the next DGLs-CF activities will be described later. The key-concept of the procedure is that actions with the same verb-accusative pattern can be grouped together because, independently from their goals, they could influence the same product features. Product features driving to the same group can then be considered related each other. Figure 3.9 offers a graphical overview of the procedure for the relationship discovery.

3.5 The Node A1. First Setup

89

Fig. 3.9 Graphical representation of the procedure for product feature relationship discovery

The first part of the procedure recalls the activities performed in modules MOD1 and MOD2: step a) Identification of technological characteristics and product features, step b) Generation of rules, and step c) Generation of actions. These steps generate only the knowledge needed by the procedure and, as a matter of fact, they are not part of the inference process. The second part of the procedure consists in performing the activities to analyse the content of the knowledge base and to discover the relationships among the product features. Two steps constitute this part: step d) Generation of groups of actions, and step e) Discovery of the relationships among product features. Step a) — Identification of Manufacturing Characteristics and Product Features In Fig. 3.9, M1..Mn represent the manufacturing characteristics, V1..Vn the verification ones, while F1..Fn are the product features, as collected by the module MOD1. Please note that the inference procedure does not require the specific values of the parameters associated with characteristics and features to run. This allows discovering the relationships among features early, independently from the brand and model of the manufacturing and/or verification technologies used, and from the specific product to be processed.

90

3 Detailed Description of the DGLs-CF

Step b) — Generation of Rules In Fig. 3.9, R1..Rn represent the rules as collected using the module MOD2. Step c) — Generation of Actions In Fig. 3.9, A1..An represent the actions as collected using again the module MOD2 and respecting the format verbaccusative-goal. The different content of the brackets, verb1-acc.1-goal, verb2acc.1-goal, etc., highlights that different actions could present the same verb or the same accusative or both, while their goals are not meaningful here. Step d) — Generation of Groups of Actions The information collected before is used here to discover the possible relationships among product features. Actions with the same verb-accusative pattern are grouped together even if their goals are different. This makes sense because, by having the same pattern, they require the same activities to be performed — that means, the modification of the same parameters — even if they derive from different rules. For example, in Fig. 3.9 actions A1 and A2 share the same pattern verb1-acc.1 and for this reason they are grouped together. Step e) — Discovery of the Relationships among Product Features Given the statements of the previous steps, the activities required by the execution of the actions of the same group could influence all the features where the grouped actions come from. The way to discover and link these features together is very simple; looking once again at Fig. 3.9, grouped actions (bottom) allow grouping the rules where these actions come from. For example, G1 represents the group containing rule R1 (parent of A2) and rule R3 (parent of A1). After that, these grouped rules indicate all the features related each other, because there is a direct link between features and rules given that, as said before, rules are generated by considering all the characteristic/feature pairs. The group of related features is labelled Rel1, and with this result the procedure ends. Regarding the DGLs-CF data structure, table MOD3_T1, shown here in Table 3.8, collects the input and output for this module. First, each group of actions is formalised defining the domain where it takes place. This is made possible by the fact that each action defines an activity to be executed in a specific domain and actions sharing the same verb-accusative pattern require the same activity; thus, each group of actions can be associated to a specific domain. Going ahead in the description of Table 3.8, for each group of actions a label is specified, and the actions present in the group are shown. For each group, the product features related each other are then presented, reporting their mid-level parameters. A dynamic coefficient is associated to each group, whose value is equal to the number of related product features. This is a very important item, because it immediately shows how many features could be affected by an action belonging to a group. This coefficient will be used in the following, to define the sequence of the actions in the redesign/reconfiguration packages.

3.6 The Node A2. Technological Configuration

91

Table 3.8 Examples of groups of actions and of the derived relationships among product features — table MOD3_T1 Domain

Group Actions label Verbaccusative pattern Design GD1 Use existing plane surfaces in splitting — the model Manufacturing GM1 Merge — the product

Verification

GV1

Product feature mid-level Product parameters features related each other AD_M1F1 Cavities Cavities_x, Cavities_y, AD_M2F4_a Cavities_z, Cavities_φ, Cavities_d, Cavities_β Bounding Bounding_box_X, box Bounding_box_Y, Bounding_box_Z Labels

Dynamic coefficient

2

2 Bounding_box_X, Bounding_box_Y, Bounding_box_Z Cavities Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β Orient — AV_V2F3_a Overhangs/ Overhangs_sloped_surfaces_α 2 AV_V2F4_a sloped the product surfaces Cavities Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β AM_M1F1 Bounding AM_M2F4_b box

3.6 The Node A2. Technological Configuration The purpose of this phase is to customise the parametric characteristics given the brand and model of the available manufacturing and verification technologies. Figure 3.10 shows the IDEF0 diagram of this phase.

3.6.1 Activities A21. Characterise the Manufacturing Technology Once the DGLs-CF building contains the parametric manufacturing characteristics, this activity sets the correct values of the parameters given the available manufacturing technology. The module MOD4, named Characteristic data input module, helps in collecting these values and in monitoring their correctness. Here only the manufacturers are involved.

92

3 Detailed Description of the DGLs-CF

A22. Characterise the Verification Technology This activity is the same as the previous one, but here the parameter values of the verification characteristics are set, given the verification equipment. Again, the module MOD4 is in charge of it. Here only the inspectors are involved.

Fig. 3.10 The activities of the Technological Configuration phase

3.6.2 Modules of Interest Here MOD4. Characteristic Data Input Module This module helps in collecting the data describing the manufacturing and verification characteristics related to the available technologies. They are expressed in terms of values associated with the manufacturing and verification parameters defined in the First Setup phase. Table 3.9 — table MOD4_T1 — and Table 3.10 — table MOD4_T2 — contain some specific manufacturing and verification characteristics related to two examples: the RP FDM Stratasys Dimension (http://www.dimensionprinting.com) and the CMM DEA Global Image 07-07-07 (http://www.dea.it).

3.7 The Node A3. Redesign/Reconfiguration Package Generation

93

Table 3.9 Example of specific manufacturing characteristics related to the RP FDM equipment Stratasys Dimension — table MOD4_T1 Manufacturing Mid-level parameters characteristic M1 Manufacturing_workspace_x Manufacturing_workspace_y Manufacturing_workspace_z M2 Supports_x Supports_y Supports_z Supports_φ Supports_d Supports_α Supports_Ra_xy

Parameter values of the specific equipment 203 mm 203 mm 300 mm 4 mm 4 mm 4 mm 4 mm 10 mm 45° 1.2 μm

Table 3.10 Example of specific verification characteristics related to the CMM equipment DEA Global Image 07-07-07 — table MOD4_T2 Verification Mid-level parameters characteristic V1 Verification_workspace_x Verification_workspace_y Verification_workspace_z V2 Indexed_measuring_head_β

Parameter values of the specific technology 700 mm 700 mm 660 mm 90°

3.7 The Node A3. Redesign/Reconfiguration Package Generation Figure 3.11 shows the IDEF0 diagram of the Redesign/Reconfiguration Package Generation phase. This phase is executed for each product to be evaluated and, as explained in detail in the paragraph dedicated to the module MOD6, named Redesign/reconfiguration package generation module, the core is the execution of the iterative procedure used to generate the redesign/reconfiguration packages. Briefly, the product (model) is first described in the DGLs-CF; after that, if it is not fully compatible with the available technologies, an action to be executed on it or on the manufacturing or verification process is selected and simulated; all of this leads to a new product (model) that is considered as the input for the next iteration of the procedure. For the termination criteria and other details, please see again the paragraph describing the module MOD6.

94

3 Detailed Description of the DGLs-CF

Fig. 3.11 The IDEF0 diagram of the Redesign/Reconfiguration Package Generation phase

3.7 The Node A3. Redesign/Reconfiguration Package Generation

95

3.7.1 Activities A31. Characterise the Product This activity sets the parameter values of the features describing the specific product for any iteration of the procedure. The designers can afford this task personally as they are helped by the MOD5, named Feature data input module that, once again, monitors the data labelling and correctness. A32. Activate the Knowledge and Select the Most Dynamic Action For any iteration of the procedure, the actions corresponding to the violated rules — where the compatibility is equal to 0 — are activated. This generates a collection of potential actions to be executed on the product (model) to achieve the best compliance between the product and the manufacturing and verification processes. These actions are then sorted by precise criteria as shown in detail in the description of the module MOD6, and in the end the action to be executed is selected and inserted in a redesign/reconfiguration package. This activity is performed completely in an automatic way without human intervention, so no actors are involved as mechanisms. A33. Simulate the Most Dynamic Action Here the action selected in the previous activity is simulated. The result consists of a new product (model) that will be used as input for the next iteration of the redesign/reconfiguration package generation procedure. The module MOD7, named Most dynamic action simulation module, would be expected here because in some way there should be some sort of automatism in the simulation process. For now this simulation is completely left to the actors, each of them in charge of their specific domains. Future research will manage this topic, as discussed in the final chapter of this book.

3.7.2 Modules of Interest Here MOD5. Feature Data Input Module This module collects the data for the features related to the specific product to be manufactured and verified. It is very similar to the module MOD4; the reason why they are separated is that they are used in two distinct phases of the DGLs-CF adoption. Product features are expressed in terms of specific values associated to the parameters defined in the First Setup phase. As introduced before, the procedure used to generate the redesign/reconfiguration packages is iterative; for

96

3 Detailed Description of the DGLs-CF

this reason the data structure of this module, shown in Table 3.11 — table MOD5_T1 — allows more than one instance of data collection, one for any iteration. The values related to the iterations are reported time by time in the proper column, should they concern the model or the product. The order of the Model and Product columns in the table highlights that the actions implying some modifications on the model must be performed first, making it senseless to consider modifications to the physical representation of the product before them. Table 3.11 Examples of specific product features collected during some iterations of the redesign/reconfiguration package generation procedure — table MOD5_T1 Product feature Label Mid-level parameters

F1 F2

Parameter values (iterations) Product model IV I II III Split part A Bounding_box_X 304 mm 159 159 113 Bounding_box_Y 159 mm 106 106 106 Bounding_box_Z 106 mm 304 304 150 Minimum_dimensions_x 16 mm 3 5 5 Minimum_dimensions_y 3 mm 3 5 5 Minimum_dimensions_z 3 mm 16 16 16

Product V VI Split part B 159 106 154 10 10 16

159 106 304 5 5 16

159 106 304 5 5 16

MOD6. Redesign/Reconfiguration Package Generation Module As said before, current release of the redesign/reconfiguration package generation procedure, derived from previous work, is based on iteration. First, it analyses the product (model), finds the first action to apply and simulates it in a user-assisted way. The result is a new product (model), considered as input for the next iteration of the procedure. The procedure is completed when all the meaningful actions to make the product (model) compatible with the manufacturing/verification processes are applied. An ordered collection of simulated actions constitutes a redesign/reconfiguration package, or, in other words, the expected result of the DGLs-CF adoption. The procedure runs as follows. For any iteration, the compatibility expressions associated with the rules are evaluated considering the values assumed by the feature parameters of the product (model) time by time, in order to represent quantitatively the compatibility between the specific product (model) and the available technologies. The compatibility values are reported in the column of Table 3.12 — table MOD6_T1 — related to the current iteration of the procedure. This data structure is flexible enough to cover all the cases that can happen. For example, if an action requires splitting the product model, next iteration must manage two separated parts, so two columns for the same iteration must be allowed.

3.7 The Node A3. Redesign/Reconfiguration Package Generation

97

Table 3.12 Examples of compatibility values for each iteration of the redesign/reconfiguration package generation procedure — table MOD6_T1 Compatibility Compatibility values (iterations) expressions Product model Product IV I II III V VI Split part A Split part B E_M1F1 E_M2F2 E_M2F3

0 0 1

0 1 1

0 1 1

1 1 1

1 1 1

1 1 1

1 1 1

Every time the compatibility is equal to 0, proper actions are activated in their domains. The activated actions and their dynamic coefficients, coming from the table MOD3_T1, are considered, and a ranking of actions is established for any iteration. As a general criterion, except for particular situations, time by time the action with the highest dynamic coefficient is assumed to be simulated first. The reason for this is simple; the execution of this action, given that it could influence the widest set of product features, could determine the widest range of modifications, which should be preferably brought and evaluated at the first stages of the redesign/reconfiguration process. The action rankings corresponding to the iterations of the procedure are shown in the columns of the table MOD6_T2, shown here in Table 3.13. This table contains only the actions that have been activated during all the iteration of the procedure. Table 3.13 Examples of action rankings, based on the dynamic coefficients, for any iteration of the redesign/reconfiguration packages definition procedure — table MOD6_T2 Domain

Actions Label

Ranking of the activated actions (iterations) Product Dynamic Product model coefficient I II III IV V VI Split part A Split part B Design AD_M1F1 2 3° 2° 1° - AD_M2F2 2 2° 1° - Manufacturing AM_M1F1 2 3° 2° 1° 1° 1° - AM_M3F5b 4 1° - Verification AV_V2F3_a 2 2° AV_V2F3_b 2 1° -

Regarding particular situations where the simple criterion “highest dynamic coefficient first” could not apply, some examples are reported in the following. In case of more than one action to be simulated first, because they show the same dynamic coefficient, alternative packages could be evaluated; they are generated by applying the procedure many times, one for each alternative action. If, after the simulation of an action, the incompatibility related to the violation of the corresponding rule is not solved, it must be evaluated if an alternative action

98

3 Detailed Description of the DGLs-CF

exists; in that case, the procedure executes again the current iteration excluding the action which did not solve the incompatibility. If a rule showing compatibility equal to 0 has no actions associated, the procedure generates a void redesign/reconfiguration package and exits. In fact, this means that it is impossible to modify the product (model) in order to manufacture or verify it with the available technologies. For example, this could happen when the “Bounding box” of the product is bigger than the “Verification workspace”; in this case no actions can solve the problem except for changing the verification technology. In the special situation where only one rule is violated and this has two or more actions associated, the least dynamic action is selected instead of the most one, just to try limiting the interference with the other product features. The simulation of the action in any iteration generates new input values for the parameters of the product features and the whole procedure is repeated until all the compatibility values are equal to 1. A redesign/reconfiguration package is thus defined, establishing the ordered list of actions to be executed in their domains. Warnings concerning all the features which may be affected by each action are also reported as a further help for the DGLs-CF users and this can be considered as an enhancement of the DGLs-CF usability. All the pieces of information related to any redesign/reconfiguration package are collected in the table MOD6_T3, shown here with an example in Table 3.14. Table 3.14 Example of a redesign/reconfiguration package — table MOD6_T3 Execution order Domain 1°

2°

Action

Manufacturing Orient the model to make the roughness resulting from slicing compatible with the requirements Design Over-dimension thin parts to make them compatible with the need for supports

Warnings (related product features) Overhangs/sloped surfaces Cavities Surface finishing Mechanical properties Minimum dimensions Mechanical properties

It should not be surprising that an action related to the manufacturing domain must be performed before another one related to the design domain; both actions in fact imply modifications of the product model. Obeying the first action, a particular orientation of the model in the manufacturing workspace must be considered in the design phase. As pointed out at the beginning of the previous chapter, the DGLs-CF is a design for Multi-X, linking different stages of the product development process; however, all the redesign activities are performed exclusively on the product (model), even if they happen during manufacturing or verification.

Summary

99

MOD7. Most Dynamic Action Simulation Module The last activity of the IDEF0 diagram presents this module as a mechanism but for the moment this is an empty box. Given the heterogeneity of the actions that the DGLs-CF could require to apply to the product (model), it is quite impossible to develop a single module to perform them automatically. This simulation would involve interfacing the DGLs-CF with CAD/CAM — Computer Aided Design/Computer Aided Manufacturing — systems, FEM — Finite Element Method — codes, expert systems for manufacturing and verification, local or distributed structured and unstructured knowledge bases, etc., and all of this is out of the scope of this project. Anyway, the opportunity to study something that could come in help at this point will be discussed in the last chapter. For now, the goal is to give the DGLs-CF users clear instructions about what to do to the product (model), how to perform it, and where to insert the results of the simulation in the DGLs-CF data structure.

3.8 Discussion At this point of the chapter, the description of the previous releases of the DGLs would suggest the presence of the discussion about solved problems, open issues, etc., all of this related to the DGLs-CF. In this case, these considerations will be the focus of the last chapter of the book, because it is important to describe the adoption of the DGLs-CF in the field before it. The next chapter is completely dedicated to the description of some case studies where the DGLs-CF has been adopted, while, afterwards, Chap. 5 contains the discussion.

Summary Previous releases of the DGLs constituted a good starting point for the development of the DGLs-CF. This chapter has described the top-down development of this design for Multi-X method, using a well known formalism as IDEF0 and focusing on a modular architecture that simplifies comprehension, some further development, and a possible implementation. Getting the goals of the DGLs-CF, the redesign of products and the reconfiguration of manufacturing and verification processes, requires some important procedures to formalise the knowledge in a correct and effective way, to discover relationships among product features and to generate redesign/reconfiguration packages, etc. All of them have been described in detail. Now is the time to apply the result of this project in the field and this will be the topic of the next chapter.

100

3 Detailed Description of the DGLs-CF

References Ambler SW (2005) The Elements of UML™ 2.0 Style. Cambridge University Press. Cristofolini I and Filippi S (2008) Exploiting the features of ISO GPS standards to enhance a knowledge based method for product redesign and process reconfiguration. In: Proc. IDETC/CIE 2008 - ASME International Design Engineering Technical Conferences & Computers and Information In Engineering Conference DETC 2008, New York City, NY, I798CD ISBN 0-7918-3831-5 Cristofolini I, Filippi S, Bandera C (2006) How Rapid Prototyping Process Parameters could affect the Product Design Phase: a KBS approach. In: Proc. IDETC/CIE 2006 - ASME International Design Engineering Technical Conferences & Computers and Information In Engineering Conference DETC 2006, Philadelphia, PA, ISBN 0-7918-3784-X Cristofolini I, Filippi S, Bandera C (2008) The role of product feature relations in a knowledgebased methodology to manage design modifications for product measurability. Int J Prod Res47(9):2373,2389 Eriksson HE, Penker M, Lyons B, Fado D (2004) UML™ 2 Toolkit. Wiley Publishing, Inc., Indianapolis, Indiana. Filippi S and Cristofolini I (2007) The Design Guidelines (DGLs), a Knowledge-Based System for industrial design developed accordingly to ISO-GPS (Geometrical Product Specifications) concepts. Res Eng Des 18(1):1-19 FIPS PUB 183 (1993) Draft Federal Information Processing Standards, Standard for Integration Definition for Function Modeling (IDEF0).National Institute of Standards and Technology, Gaithersburg, MD, USA http://www.dea.it Accessed 24 April 2009 http://www.dimensionprinting.com 24 April 2009 IEEE 1320.1 (1998) IEEE Standard for Function Modeling Language-Syntax and Semantics for IDEF0 (Replaces FIPS PUB 183) ISO 14660-1:1999 Geometrical Product Specification (GPS) - Geometrical features - Part 1: General terms and definitions ISO 14660-2:1999 Geometrical Product Specification (GPS) - Geometrical features - Part 2: Extracted median line of a cylinder and a cone, extracted median surface, local size of an extracted feature ISO/TS 17450-1:2000 Geometrical Product Specifications (GPS) - General Concepts - Part 1: Model for geometrical specification and verification ISO/TS 17450-2:2002 Geometrical Product Specifications (GPS) - General concepts - Part 2: Basic tenets, specifications, operators and uncertainties

4 Adopting the DGLs-CF in the Field

The previous chapter has described the DGLs-CF in detail and this one deals with its application in the field. Three case studies are presented in order to clarify the DGLs-CF issues and to highlight the resulting outcomes. Different products have been considered such as different RP manufacturing technologies, while verification is always performed with CMM technology. Attention was focused on the RP technologies since they are considered new ones, presenting characteristics not always widely known in depth. CMM is used because ISO GPS focuses most attention on it (ISO 10360-1:2000; ISO 10360-2:2001; ISO 10360-3:2000; ISO 10360-4:2000; ISO 10360-5:2000; ISO 10360-6:2001; ISO/TS 15530-3:2004; ISO/TS 15530-4:2008; ISO/TS 23165:2006). It must be pointed out that when dealing with prototypes of methods, systems, etc., there are two dimensions to consider: the horizontality, representing the completeness of the prototype regarding the set of required functionalities, and the verticality, related to the level of definition/implementation of each function. The goal of these case studies has been essentially to check the horizontal, transversal completeness of the DGLs-CF, rather than its verticality, or depth. In other words, these case studies had mainly to prove that the whole redesign/reconfiguration process sounds good. This is the reason why, sometimes, the rules, the compatibility expressions, the actions, etc., may sometimes seem incomplete or coarse. In the future, when the active role of the actors are reduced by introducing some automatisms, they will be revised and refined. This chapter opens with a short description of the technologies used in the case studies. Then the three case studies are presented in detail. These descriptions are based on the seven modules of the DGLs-CF rather than on the activities of the IDEF0 diagrams because in this way the comprehension of the activities gets easier.

102

4 Adopting the DGLs-CF in the Field

4.1 Rapid Prototyping — RP — Technologies Rapid Prototyping — RP — technologies allow the construction of products directly from their models. Complex objects can be generated avoiding the use of standard technologies, such as NC — Numerical Control — milling machines, etc., so costs and times are significantly reduced. RP activities are preceded by the design phase, where the model needed for the prototyping process is generated (Horvàth et al. 1999; Qian and Dutta 2001; Tang et al. 2005). The prototyping phase then begins, which can be divided into three main steps: 1. Pre-processing CAD data are elaborated for the RP equipment, and the process parameter setting takes place. The STL file, the file format considered as a standard de facto in the RP domain, is converted into the correct format, then the support addition, the slicing, and the job generation take place. 2. Processing Here the prototype is built by the RP equipment. 3. Post-processing Here support removal, surface finishing, form and dimension adjustment, etc. could take place. If necessary, the object is finished using traditional methods to satisfy completely the design requirements. These characteristics of the RP technologies are the basic ones, which will help in the comprehension of the DGLs-CF adoption. For further information see the work of Gatto and Iuliano (1998), Jacobs (1995), Pham and Dimov (2001) and Rosochowski and Matuszak (2000). Several RP technologies have been developed; the most widespread of them are the Fused Deposition Modelling — FDM, the Stereolithography — SLA, and the Selective Laser Sintering — SLS. All of them are layer manufacturing technologies, but their characteristics may considerably differ (Byun and Lee 2005; Cooper 2001; Mahesh et al. 2004; Masood and Soo 2002). They will be introduced inside the case studies, highlighting their specific characteristics.

4.2 Coordinate Measuring Machines — CMMs Coordinate Measuring Machines — CMMs — represent one of the most powerful metrological instruments, widely used in many manufacturing contexts, due to their flexibility and accuracy, coupled with reduced measurement times and costs. Hereafter, only the main characteristics of the CMMs are summarised, in order to help again the comprehension of the DGLs-CF adoption. The main goal of coordinate metrology consists in the measurement of the actual shape of a workpiece, in its comparison with the nominal shape and in the evaluation of the conformity to the requirements, in terms of dimensional and geometrical tolerances. The actual shape of the workpiece is obtained by probing its surface at discrete measuring points. In the scanning mode, points may be

4.2 Coordinate Measuring Machines — CMMs

103

pitched close and ordered, so a profile can be extracted. In any case, each point is represented by its measured coordinates. Nevertheless, points or profiles do not allow a direct comparison with the nominal shape, so that an appropriate best-fit algorithm must be applied to obtain the geometric elements constituting the actual shape, in order to compare it to the nominal one. CMMs are the physical representation of a three-dimensional rectilinear Cartesian coordinate system. The configuration of a CMM plays an important role in meeting measurement requirements such as accuracy, flexibility, time, and costs. The most common configuration for high precision measurements is the bridge one, shown in Fig. 4.1. The accuracy of this kind of CMM can be about few microns.

Fig. 4.1 A CMM equipment presenting the bridge configuration (courtesy of Hexagon Metrology Italia)

Point coordinates can be measured by means of a contact probe — contact measurement, but most CMMs can be equipped with optical probe systems too, obtaining a non-contact measurement. In the latter case the accuracy is limited, normally to some dozens of microns (Bosch 1995). Figure 4.2 shows examples of two measuring heads, corresponding to contact and non-contact measurements. The procedures used for the inspections may affect the final result and they have been extensively studied, for example by Hwang et al. (2004), Kweon and Medeiros (1998), Lin and Chow (2001), Merat and Radack (2002), Roy et al. (1994), and Ziemian and Medeiros (1998). In particular, the algorithms used to fit the measured points may have a great influence in determining the resulting workpiece surfaces which are compared with the nominal ones; in other words, they could influence a lot in determining the conformity of the part (Cristofolini and Podda 2001). This argument is extensively treated in the ISO GPS regarding different aspects, from features definition, to measurement, to data filtration, etc. (ISO 10360-1:2000; ISO 10360-2:2001; ISO 10360-3:2000; ISO 10360-4:2000;

104

4 Adopting the DGLs-CF in the Field

ISO 10360-5:2000; ISO 10360-6:2001; ISO/TS 14253-1:1998; ISO/TS 142532:1999; ISO/TS 14253-3:2002; ISO 14660-1:1999; ISO 14660-2:1999; ISO/TS 15530-3:2004; ISO/TS 15530-4:2008; ISO/TS 16610-1:2006; ISO/TS 1661020:2006; ISO/TS 16610-22:2006; ISO/TS 16610-29:2006; ISO/TS 1661040:2006; ISO/TS 16610-41:2006; ISO/TS 16610-49:2006; ISO/TS 23165:2006).

Fig. 4.2 Measuring heads of CMMs, equipped with contact probe (left) and optical probe (right) (courtesy of Hexagon Metrology Italia)

4.3 First Case Study. A Mechanical Bracket Built Using FDM In the first case study, the FDM manufacturing technology and the CMM verification technology were considered for the generation of a mechanical bracket. The starting model is shown in Fig. 4.3.

Fig. 4.3 The starting model of the mechanical bracket used as product for this case study

A short overview of the main characteristics of the FDM manufacturing technology is reported hereafter, so it will be easier to understand the meaning of the parameters defined and used in the case study.

4.3 First Case Study. A Mechanical Bracket Built Using FDM

105

4.3.1 FDM Fundamentals FDM is an RP technology that builds prototypes in multiple thin layers, using spools of thermoplastic wire as material. As depicted in Fig. 4.4, the material is heated to just above the melting point in a delivery head. The molten thermoplastic is then extruded through a nozzle as a thin ribbon and deposited in computer-controlled locations appropriate for the object geometry, thus building a section of the prototype. Typically, the delivery head moves in the horizontal plane while the foam base, where the prototype is built, moves vertically, so that each section is built over the previous one. Deposition temperature is such that the deposing material bonds almost instantaneously with that deposited before.

Fig. 4.4 FDM process outline

Depending on the geometrical complexity of the prototype, some supporting material may be necessary to build the model. The amount and the shape of the supports, which will be removed from the final prototype, are automatically calculated, based on the orientation of the prototype. The first section of the model is always built on a layer of supports, slightly larger than the model, to allow an easy removal of the prototype from the foam base. Precision and surface finishing of the prototypes are affected by the so-called slicing or layering, which depends on the kind of equipment used. The layer thickness can vary typically from 0.17 mm to 0.33 mm. A wide range of thermoplastic materials can be used to build models, including ABS — Acrylonitrile Butadiene Styrene, polyolefin and polyamide. The final prototypes do not require post-processing, except for support removal and grinding, for a better surface finishing. Another advantage of the FDM is that

106

4 Adopting the DGLs-CF in the Field

it can be used not only in a laboratory, but in an office as well; no high powered lasers are used and the material, supplied in spool format, does not require special handling or present environmental concerns (Chua et al. 2003; Cooper 2001; Gatto and Iuliano 1998; Pham and Dimov 2001).

4.3.2 First Setup MOD1 — Characteristic and Feature Collecting Module Considering the classes of the FDM and CMM technologies described earlier, the main parametric manufacturing and verification characteristics were collected by the manufacturers and inspectors respectively. According to the translation procedures described in the previous chapter, the manufacturing and verification characteristics were described using mid-level parameters. The result is shown in Table 4.1, corresponding to the table MOD1_T1 of the DGLs-CF data structure, and in Table 4.2 — table MOD1_T2. Table 4.1 Parametric manufacturing characteristics of the class FDM — table MOD1_T1 of the DGLs-CF data structure Parametric manufacturing characteristic Label Name Description Mid-level parameters M1 Manufacturing Volume of the Manufacturing_Workspace_x, workspace manufacturing Manufacturing_Workspace_y, workspace Manufacturing_Workspace_z (maximum dimensions of the manufacturing workspace) M2 Supports Critical Supports_x, Supports_y, Supports_z (minimum characteristics dimensions related to the need for supports) related to the Supports_φ (minimum diameter related to the need need for supports for supports) when building Supports_d (maximum depth or height related to overhangs/sloped the need for supports) surfaces or Supports_α (minimum angle related to support cavities removal) Supports_Ra_xy (minimum obtainable roughness in xy plane) M3 Slicing Material deposed Slicing_zmin (minimum thickness of the slice) slice by slice Slicing_σx, Slicing_σy, Slicing_σz (minimum strength in the three dimensions) Slicing_Ra_z (minimum obtainable roughness in z direction) M4 Material Kind of material Material_φwire (minimum diameter of the wire) used Material_σ (breaking strength of the material)

4.3 First Case Study. A Mechanical Bracket Built Using FDM

107

Table 4.2 Parametric verification characteristics of the class CMM — table MOD1_T2 Parametric verification characteristic Label Name Description Mid-level parameters V1 Verification Volume of the Verification_workspace_x, workspace verification workspace Verification_workspace_y, Verification_workspace_z (maximum dimensions of the verification workspace) V2 Indexed Possible rotation and Indexed_measuring_head_β (minimum angle measuring inclination of the of inclination of the head) head measuring head V3 Probe Kind of probe used for Probe_φ (minimum diameter of the probe) verification Probe_l (maximum length of the probe) V4 Clamping Limitations due to the Clamping_tools_x, Clamping_tools_y, tools clamping tools Clamping_tools_z (minimum dimensions defining the contacting area with the clamping tools) Table 4.3 Parametric product features of the class Mechanical parts — table MOD1_T3 Parametric product feature Label Name Description F1 Bounding box Overall dimensions of the product F2 Rib, pins, Minimum minimum dimensions in dimensions the product

F3

F4

F5

F6

Mid-level parameters Bounding_box_X, Bounding_box_Y, Bounding_box_Z (maximum dimensions of the product) Minimum_dimensions_x, Minimum_dimensions_y (minimum dimensions in xy plane) Minimum_dimensions_z (minimum dimension in z direction) Minimum_dimensions_φ (minimum diameter) Minimum_dimensions_h (minimum height) Webs, Overhangs and Overhangs_Sloped_surfaces_α Overhangs/Sloped protrusions (overhangs/sloped surfaces angle) /Free form surfaces Cavities Through and Cavities_x, Cavities_y, Cavities_z (minimum blind holes, dimensions for not cylindrical cavities) undercuts and Cavities_φ (minimum diameter for cylindrical other cavities cavities) Cavities_d (maximum depth for cylindrical cavities) Cavities_β (angle of inclination of the axis of the cavities) Surface finishing Surface Surface_finishing_Ra_xy (maximum allowable texture roughness in xy plane) Surface_finishing_Ra_z (maximum allowable roughness in z direction) Mechanical Main Mechanical_properties_σx, properties mechanical Mechanical_properties_σy, properties Mechanical_properties_σz (minimum strength in the three directions)

108

4 Adopting the DGLs-CF in the Field

Considering these technological characteristics, the main product features of the class of products mechanical parts have been identified. Again, the mid-level parameters have been derived using the procedures, and the result is shown in Table 4.3 — table MOD1_T3. These features have been identified by the designers together with the manufacturers and inspectors as the basic features describing mechanical parts. They will be used in all the study cases, relating them to the specific technologies time by time. It is worth remembering that, according to the convention explained in the Appendix, in all the case studies angle values have been considered in the range [0°—180°], starting from the vertical line corresponding to the z direction, pointing downwards, and proceeding counter-clockwise. MOD2 — Rule and Action Generation Module Table 4.4 — table MOD2_T1 — shows the rules and the compatibility expressions derived by crossing the parametric technological characteristics and the parametric product features. Designers, manufacturers and inspectors performed this work together. Table 4.4 Rules and related compatibility expressions — table MOD2_T1 Rule Label Description Procedure R_M1F1 Dimensions Locate the features of size defining defining the the bounding box in the model — gives the associated parameters bounding Bounding_box_X, box of the product must Bounding_box_Y, Bounding_box_Z. be smaller Identify the parameters defining than the maximum dimensions of the maximum dimensions building room — gives Manufacturing_workspace_x, of the Manufacturing_workspace_y, manufact. workspace Manufacturing_workspace_z Locate the features of size defining R_M2F2 Minimum dimensions the minimum dimensions in the model — gives the associated of the product must parameters Minimum_dimensions_x, be bigger Minimum_dimensions_y, than Minimum_dimensions_z, minimum dimensions Minimum_dimensions_φ, related to the Minimum_dimensions_h. presence of Identify the parameters defining the critical dimensions related to supports the need for supports — gives Supports_x, Supports_y, Supports_z, Supports_φ, Supports_d

Compatibility expressions E_M1F1=1 IF Bounding_box_Z< Manufacturing_workspace_z AND Bounding_box_X< Manufacturing_workspace_x AND Bounding_box_Y< Manufacturing_workspace_y ELSE E_M1F1=0

E_M2F2=1 IF Minimum_dimensions_z>Supports_z AND Minimum_dimensions_x>Supports_x AND Minimum_dimensions_y>Supports_y AND Minimum_dimensions_φ>Supports_φ AND Minimum_dimensions_h<Supports_d ELSE E_M2F2=0

4.3 First Case Study. A Mechanical Bracket Built Using FDM

109

Table 4.4 (continued) Locate the features of size defining the overhangs/sloped surfaces in the model — gives the associated parameter Overhang_sloped_surfaces_α. Identify the parameters defining the critical inclination related to the presence of supports — gives Supports_α Locate the features of size R_M2F4 The dimensions, defining the minimum and/or deepest cavities in the model depth and orientation of — gives the associated parameters Cavities_φ, the cavities Cavities_d, Cavities_x, must be Cavities_y, Cavities_z, and compatible with the need Cavities_β. for supports Identify the parameters defining the critical dimensions and inclination related to the need for supports — gives Supports_φ, Supports_x, Supports_y, Supports_z, Supports_d, and Supports_α R_M2F3 The presence of overhangs/ sloped surfaces must be evaluated considering the need for supports

R_M2F5 Roughness must be evaluated considering the presence of supports

-

E_M2F3=1 IF Overhang_sloped_surfaces_α>Supports_α ELSE E_M2F3=0

BLIND HOLES E_M2F4=1 IF (Cavities_φ≥1.5•Supports_φ AND Cavities_d<Supports_d AND Cavities_x≥1.5•Supports_x AND Cavities_y≥1.5•Supports_y AND Cavities_z≥1.5•Supports_ z) OR (Cavities_φ<1.5•Supports_φ AND Cavities_x<1.5•Supports_x AND Cavities_y<1.5•Supports_y AND Cavities_z<1.5•Supports_ z AND Cavities_ β>(Supports_ α + 90°) ELSE E_M2F4=0 THROUGH HOLES E_M2F4=1 IF Cavities_φ≥Supports_φ AND Cavities_x≥Supports_x AND Cavities_y≥Supports_y AND Cavities_z≥Supports_ z ELSE E_M2F4=0 E_M2F5=1 IF Surface_Finishing_Ra_xy>Supports_Ra_xy ELSE E_M2F5=0

110

4 Adopting the DGLs-CF in the Field

Table 4.4 (continued) R_M3F2 Minimum zthickness must be greater than minimum dimensions related to the slicing

R_M3F5 Maximum roughness of the product must be compared with the minimum roughness related to the slicing R_M3F6 Mechanical properties must be considered in function of the slicing R_M4F2 Minimum xydimensions must be compatible with the diameter of the wire

Locate the features of size defining the minimum dimensions orthogonal to x— y plane in the model — gives the associated parameter Minimum_dimensions_z, Minimum_dimension_hmin. Identify the parameters defining the critical dimensions related to the slicing — gives Slicing_zmin -

-

Locate the features of size defining the minimum dimensions parallel to x—y plane in the model — gives the associated parameter Minimum_dimensions_x, Minimum_dimensions_y, Minimum_dimensions_φ. Identify the parameters defining the critical dimensions related to the diameter of the wire — gives Material_φwire R_M4F4 Minimum xy- Locate the features of size in dimensions of the x—y plane defining the minimum cavities in the the cavities model — gives the associated must be parameters Cavities_φ, or compatible Cavities_x, Cavities_y. with the Identify the parameters diameter of defining the critical the wire dimensions related to the diameter of the wire — gives Material_φwire

E_M3F2=1 IF Minimum_dimensions_z>10•Slicing_zmin AND Minimum_dimension_hmin>10•Slicing_zmin ELSE E_M3F2=0

E_M3F5=1 IF Surface_Finishing_Ra_z>Slicing_Ra_z ELSE E_M3F5=0

E_M3F6=1 IF Mechanical_properties_σz<Slicing_σz ELSE E_M3F6=0 E_M4F2=1 IF Minimum_dimensions_x>3•Material_φwire AND Minimum_dimensions_y>3•Material_φwire AND Minimum_dimensions_φ>3•Material_φwire ELSE E_M4F2=0

E_M4F4=0 IF Cavities_φ>2•Material_φwire AND Cavities_x>2•Material_φwire AND Cavities_y>2•Material_φwire ELSE E_M4F4=0

4.3 First Case Study. A Mechanical Bracket Built Using FDM

111

Table 4.4 (continued) R_M4F6 Mechanical properties must be compared with those related to the material used in the FDM technology R_V1F1 Dimensions defining the bounding box of the product must be smaller than maximum dimensions of the verification workspace

R_V2F3 The overhangs/ sloped surfaces must be accessible by the measuring head, considering its inclination R_V2F4 The cavities must be accessible by the measuring head, considering its inclination

-

E_M4F6=1 IF Mechanical_properties_σx<Material_σ AND Mechanical_properties_σy<Material_σ AND Mechanical_properties_σz<Material_σ ELSE E_M4F6=0

Locate the features of size defining the bounding box in the model — gives the associated parameters Bounding_box_X, Bounding_box_Y, Bounding_box_Z. Identify the parameters defining the maximum dimensions of the building room — gives Verification_workspace_x, Verification_workspace_y, Verification_workspace_z Locate the features of size defining the overhangs/sloped surfaces in the model — gives the associated parameter Overhang_sloped_surfaces_α. Identify the parameters defining the critical inclination related to the measuring head — gives Indexed_measuring_head_β Locate the features of size defining the inclination of the cavities in the model — gives the associated parameter Cavities_β. Identify the parameters defining the critical inclination related to the measuring head — gives Indexed_measuring_head_β

E_V1F1=1 IF Bounding_box_Z
E_V2F3=1 IF Overhang_sloped_surfaces_α>Indexed_meas uring_head_β ELSE E_V2F3=0

E_V2F4=1 IF Cavities_β>Indexed_measuring_head_β ELSE E_V2F4=0

112

4 Adopting the DGLs-CF in the Field

Table 4.4 (continued) E_V3F4=1 IF Cavities_φ>5•Probe_φ AND Cavities_dClamping_tools_x AND the model — gives the bounding Bounding_box_Y>Clamping_tools_y AND associated parameters box of the Bounding_box_Z>Clamping_tools_z product must Bounding_box_X, ELSE Bounding_box_Y, be E_V4F1=0 compatible Bounding_box_Z. Identify the parameters with the defining the minimum need for dimensions of the clamping clamping tools — gives tools Clamping_tools_x, Clamping_tools_y, Clamping_tools_z

R_V3F4 The dimensions of the cavities must be compatible with the diameter and length of the probe

Locate the features of size defining the minimum and/or deepest cavities in the model — gives the associated parameters Cavities_φ, Cavities_d or Cavities_x Cavities_y, or Cavities_z. Identify the parameters defining the critical dimensions and inclination related to the probe — gives Probe_φ, Probe_l

Table MOD2_T2 collected the actions derived from the rules of table MOD2_T1, generated again by designers, manufactures and inspectors, working together. This table is split into three parts corresponding to the domains where the actions applied — design, manufacturing, or verification — and also to the three floors of the DGLs-CF building. Tables 4.5, 4.6 and 4.7 represent these three parts of table MOD2_T2 respectively. Table 4.5 Actions in the design domain — first part of table MOD2_T2 Action Domain Label Design

Link

Description Verb Accusative the model AD_M1F1 AM_M1F1 Use existing plane surfaces in splitting

Procedure Goal to make dimensions compatible with the manufact. workspace

Locate the features of size related to those parameters giving E_M1F1=0. Locate the situation features (planes and/or axes and/or median planes) belonging to the model and orthogonal to the features of size previously identified. Split the model using planes containing these situation features

4.3 First Case Study. A Mechanical Bracket Built Using FDM

113

Table 4.5 (continued) Design AD_M2F2

-

Overdimension

AD_M2F3

-

Overdimension

AD_M2F4_ AM_M2F4 Use existing a _b plane surfaces in splitting

AD_M2F4_ b

Overdimension

Locate the features of size related to those parameters giving E_M2F2=0. Locate two situation features planes orthogonal to the features of size previously identified. Over-dimension thin parts using these situation features Locate the features of size to the overhangs/ minimise related to those parameters giving E_M2F3=0. the sloped problems Locate the situation feature surfaces related to plane corresponding to the critical inclination, in support respect of the need for removal supports. Over-dimension the overhangs/sloped surfaces using this situation feature Locate the features of size the model to avoid related to those parameters the need giving E_M2F4_a=0. for Locate the situation supports features (planes and/or axes and/or median planes) belonging to the model and orthogonal to the features of size previously identified. Split the model using planes containing these situation features Locate the features of size the cavities to eliminate related to those parameters giving E_M2F4_b=0. their Locate two situation critical dimensions features planes orthogonal to the features of size previously identified or locate a situation feature cylinder coaxial with the features of size previously identified. Over-dimension the cavities using these situation features thin parts

to make them compatible with the need for supports

114

4 Adopting the DGLs-CF in the Field

Table 4.5 (continued) Design AD_M3F2

-

Overthin parts dimension

to make them compatible with the slicing

AD_M3F6

-

Overthin parts dimension

to make them conform with the requested mechanical properties compatibly with the slicing

AD_M4F2

-

Overthin parts dimension

to make it compatible with the diameter of the wire

AD_M4F4

-

Overthe cavities to make them dimension compatible with the diameter of the wire

Locate the features of size related to those parameters giving E_M3F2=0. Locate two situation features planes orthogonal to the features of size previously identified. Over-dimension thin parts using these situation features Locate the features of size related to those parameters giving E_M3F6=0. Locate two situation features planes orthogonal to the features of size previously identified. Over-dimension thin parts using these situation features Locate the features of size related to those parameters giving E_M4F2=0. Locate two situation features planes orthogonal to the features of size previously identified. Over-dimension thin parts using these situation features Locate the features of size related to those parameters giving E_M4F4 =0. Locate a situation feature cylinder coaxial with the features of size previously identified. Over-dimension the cavities using these situation features

4.3 First Case Study. A Mechanical Bracket Built Using FDM

115

Table 4.5 (continued) Design AD_M4F6 -

Overdimension

AD_V3F4 -

Overdimension

Locate the features of size related to those parameters giving E_M4F6=0. Locate two situation features planes orthogonal to the features of size previously identified. Over-dimension thin parts using these situation features Locate the features of size the to obtain related to those parameters cavities compatibility giving E_V3F4=0. between the dimensions of the Locate two situation features planes orthogonal to the cavities and features of size previously probes identified or locate a situation feature cylinder coaxial with the features of size previously identified. Over-dimension the cavities using these situation features

thin parts

to improve the mechanical strength

Table 4.6 Actions in the manufacturing domain — second part of table MOD2_T2 Action Domain

Label

Link

Description Verb Accusative Goal Manufacturing AM_M1F1 AD_M1F1 Merge the to join product the split parts

AM_M2F3 -

Orient the model to minimize the quantity of required supports

Procedure Locate the features of size related to those parameters giving E_M1F1=0. Locate the situation features (planes and/or axes and/or median planes) belonging to the model and orthogonal to the features of size previously identified. Merge the model using planes containing these situation features Locate the features of size related to those parameters giving E_M2F3=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the need for supports. Orient the model using this situation feature

116

4 Adopting the DGLs-CF in the Field

Table 4.6 (continued) Locate the features of size related to those parameters giving E_M2F4_a=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the need for supports. Orient the model using this situation feature AM_M2F4_b AD_M2F4_a Merge the to join the Locate the features of size related product split parts to those parameters giving E_M2F4_b=0. Locate the situation features (planes and/or axes and/or median planes) belonging to the model and orthogonal to the features of size previously identified. Merge the model using planes containing these situation features AM_M2F5_a Orient the to avoid the Locate the features of size related to those parameters giving model need for supports on E_M2F5_a=0. the surfaces Locate the situation feature plane corresponding to the critical that need inclination, in respect of the need good for supports. finishing Orient the model using this situation feature AM_M2F5_b Grind the to obtain surfaces better roughness AM_M3F5_a Orient the to make the Locate the features of size related to those parameters giving model roughness E_M3F5_a=0. resulting from slicing Locate the situation feature plane compatible corresponding to the critical inclination, in respect of the with the requirements slicing. Orient the model using this situation feature AM_M3F5_b Grind the to obtain surfaces better roughness AM_M3F6 Orient the to obtain the Locate the features of size related to those parameters giving model best mechanical E_M3F6=0. Locate the situation feature plane properties compatibly corresponding to the critical inclination, in respect of the with the slicing. slicing Orient the model using this situation feature

Manuf. AM_M2F4_a -

Orient the model

to make support removal easier

4.3 First Case Study. A Mechanical Bracket Built Using FDM

117

Table 4.7 Actions in the verification domain — third part of table MOD2_T2 Action Domain

Label

Link

Verification AV_V2F3_a -

AV_V2F3_b -

AV_V2F4_a -

to obtain Rotate the measuring minimum reand positioning incline head of the product Orient the to obtain product best accessibility to the cavities

Procedure Locate the features of size related to those parameters giving E_V2F3_a=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the rotation and inclination of the measuring head. Orient the model using this situation feature -

Locate the features of size related to those parameters giving E_V2F4_a=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the rotation and inclination of the measuring head. Orient the model using this situation feature -

to obtain Rotate the measuring best and accessibility incline head to the cavities and the minimum repositioning AD_V4F1 Orient the To obtain the product best configuration compatible with the clamping tools

AV_V2F4_b -

AV_V4F1

Description Verb Accusative Goal Orient the to achieve product easy access compatibly with the rotation and inclination of the measuring head

118

4 Adopting the DGLs-CF in the Field

MOD3 — Feature Relationship Discovery Module According to the procedure defined in the previous chapter, actions sharing the same verb-accusative pattern have been grouped to determine their dynamic coefficients, as shown in Table 4.8 — table MOD3_T1. Table 4.8 The groups of actions sharing the same verb-accusative pattern, and the derived relationships among product features with the dynamic coefficients — table MOD3_T1 Domain Group Actions label VerbLabels Accusative pattern Design GD_1 Over-dimension AD_M2F4_b the cavities AD_M4F4 AD_V3F4 GD_2 Over-dimension AD_M2F3 the overhangs/ sloped surfaces GD_3 Over-dimension AD_M2F2 the thin parts AD_M3F2 AD_M3F6 AD_M4F2 AD_M4F6

GD_4 Use existing plane surfaces in splitting the model

AD_M1F1 AD_M2F4_a

Manuf. GM_1 Orient the model

AM_M2F3 AM_M2F4_a AM_M2F5_a AM_M3F5_a AM_M3F6

GM_2 Merge the product

AM_M1F1 AM_M2F4_b

GM_3 Grind the surfaces

AM_M2F5_b AM_M3F5_b

Product feature mid-level Product parameters features related each other

Dyn. coeff.

1 Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β Overhangs/ Overhangs_sloped_surfaces_α 1 sloped surfaces 2 Minimum Minimum_dimension_x, dimensions Minimum_dimensions_y, Minimum_dimensions_z, Minimum_dimensions_φ, Minimum_dimensions_h Mechanical Mechanical_propertis_σx, properties Mechanical_propertie_σy, Mechanical_properties_σz 2 Cavities Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β Bounding Bounding_box_X, box Bounding_box_Y, Bounding_box_Z Overhangs/ Overhangs_sloped_surfaces_α 4 sloped surfaces Cavities Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β Surface Surface_finishing_Ra_xy, finishing Surface_finishing_Ra_z Mechanical Mechanical_properties_σx, properties Mechanical_properties_σy, Mechanical_properties_σz 2 Bounding Bounding_box_X, box Bounding_box_Y, Bounding_box_Z Cavities Cavities x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β Surface Surface_finishing_Ra_xy, 1 finishing Surface_finishing_Ra_z Cavities

4.3 First Case Study. A Mechanical Bracket Built Using FDM

119

Table 4.8 (continued) Verification GV_1 Orient the AV_V2F3_a Bounding product AV_V2F4_a Box AV_V4F1 Overhangs/ sloped surfaces Cavities

Bounding_box_X, Bounding_box_Y, 3 Bounding_box_Z Overhangs_sloped_surfaces_α

Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β

AV_V2F3_b Overhangs/ Overhangs_sloped_surfaces_α GV_2 Rotate AV_V2F4_b sloped and incline surfaces the Cavities Cavities_x, Cavities_y, Cavities_z, measuring Cavities_φ, Cavities_d, Cavities_β head

2

4.3.3 Technological Configuration MOD4 — Characteristic Data Input Module The specific technology used in this case study was the FDM Stratasys Dimension, while the verification equipment used in all the case studies has been the CMM DEA Global Image 07-07-07. The Stratasys Dimension equipment is quite small, compatible with a typical office environment. The material is stored in coils; the wire feeds a hot head, where it is fused and sent forth by a nozzle. The hot head has two nozzles, one for the prototype material and one for the supporting material. For this kind of technology, supports are solid and must be removed manually (http://www.dimensionprinting.com). In other cases, supporting material can be melted away at the end of the manufacturing process. DEA Global Image 07-07-07 is a portal-shaped CMM; the head can measure point by point or scanning, and it guarantees an accuracy of 1.7 µm + L/333 (according to ISO 10360-2:2001) when measuring point by point, and of 3.4/120 µm (according to ISO 10360-4:2000) when measuring by continuous scan (http://www.dea.it). The specific technological characteristics have been collected by manufacturers and inspectors in table MOD4_T1 and table MOD4_T2, shown in Tables 4.9 and 4.10 respectively.

120

4 Adopting the DGLs-CF in the Field

Table 4.9 Specific manufacturing characteristics of the Stratasys Dimension — table MOD4_T1 Manufacturing characteristic M1 M2

M3

M4 a

Mid-level parameters Manufacturing_workspace_x Manufacturing_workspace_y Manufacturing_workspace_z Supports_x Supports_y Supports_z Supports_φ Supports_d Supports_α Supports_Ra_xy Slicing_zmin Slicing_σx Slicing_σy Slicing_σz Slicing_Ra_z Material_φwire Material_σ

Parameter values of the specific technology 203 mm 203 mm 300 mm 4 mm 4 mm 4 mm 4 mm 10 mm 45° 1.2 μm 0.254 mm 16 MPaa 20 MPaa 14 MPaa 17 μm 1 mm 20 MPa

Experimental data

Table 4.10 Specific verification characteristics of the DEA Global Image 07-07-07 — table MOD4_T2 Verification Mid-level parameters characteristic V1 Verification_workspace_x Verification_workspace_y Verification_workspace_z V2 Indexed_measuring_head_β V3 Probe_φ Probe_l V4 Clamping_tools_x Clamping_tools_y Clamping_tools_z

Parameter values of the specific technology 700 mm 700 mm 660 mm 90° 1 mm 200 mm 5 mm 5 mm 5 mm

4.3.4 Redesign/Reconfiguration Package Generation Figure 4.5 shows the drawing of the starting configuration of the mechanical bracket considered in this case study. Its dimensional and geometrical specifications are defined on the drawing, while the required mechanical properties consisted in a minimum strength of 12 MPa in the three dimensions.

4.3 First Case Study. A Mechanical Bracket Built Using FDM

121

Fig. 4.5 The drawing of the starting configuration of the mechanical bracket

MOD5 — Feature Data Input Module This module has been used for any iteration of the redesign/reconfiguration package generation procedure. At the beginning, the bracket was supposed oriented as in the isometric view of Fig. 4.5, with respect to the datum system shown. According to this, the product features and their mid-level parameters have been derived and collected by the designers in the column I of Table 4.11 — table MOD5_T1. The other columns (II, III, etc.) refer to the next iterations of the procedure. Table 4.11 Specific product features for any iteration of the redesign/reconfiguration package generation procedure — table MOD5_T1 Product feature Label Mid-level parameters

F1 F2

Bounding_box_X Bounding_box_Y Bounding_box_Z Minimum_dimensions_x Minimum_dimensions_y Minimum_dimensions_z Minimum_dimensions_φ Minimum_dimensions_h

Parameter values (iterations) Product model I II III IV Split part A 304 mm 159 159 113 159 mm 106 106 106 106mm 304 304 150 16 mm 3 5 5 3 mm 3 5 5 3 mm 16 16 16 -

Product V VI Split part B 159 106 154 10 10 16 -

159 106 304 5 5 16 -

159 106 304 5 5 16 -

122

4 Adopting the DGLs-CF in the Field

Table 4.11 (continued) F3 F4

F5 F6

Overhangs_sloped_surfaces_α Cavities_x Cavities_y Cavities_z Cavities_φ Cavities_d Cavities_β Surface_finishing_Ra_xy Surface_finishing_Ra_z Mechanical_properties_σx Mechanical_properties_σy Mechanical_properties_σz

90° -

90 -

90 -

0 -

0 -

90 -

90 -

20 mm 16 mm 90° 40 µm 3.2 µm 12 MPa 12 MPa 12 MPa

20 16 90 3.2 40 12 12 12

20 16 90 3.2 40 12 12 12

20 16 90 3.2 40 12 12 12

20 16 90 3.2 40 12 12 12

20 16 90 40 3.2 12 12 12

20 16 90 40 3.2 12 12 12

MOD6 — Redesign/Reconfiguration Package Generation Module In the first iteration of the procedure, some rules were violated and this is highlighted by the zero values of the compatibility expressions collected in the column I of Table 4.12 — table MOD6_T1. Even this table has been used for all the iterations of the procedure. The values in the table highlight that the “Maximum dimensions” were not compatible with the “Manufacturing workspace”, and the “Minimum dimensions” were not as large as needed considering the presence of “Supports”; in addition, the characteristics of “Material” and “Surface finishing” were not compatible with the “Slicing”. Moreover, the “Overhangs” were not compatible with the “Indexed measuring head”. Thus, related actions have been activated and their dynamic coefficients compared. On the basis of these values, the ranking of the actions has been established, as reported in column I of Table 4.13 — table MOD6_T2. Empty cells in the columns represent non-activated actions for the iterations represented by the same columns. Table 4.12 Compatibility values for any iteration of the redesign/reconfiguration package generation procedure — table MOD6_T1 Compatibility expressions Compatibility values (iterations) Product model IV I II III Split part A Split part B E_M1F1 0 0 0 1 1 E_M2F2 0 0 1 1 1 E_M2F3 1 1 1 1 1 E_M2F4 1 1 1 1 1 E_M2F5 1 1 1 1 1 E_M3F2 1 1 1 1 1 E_M3F5 0 1 1 1 1 E_M3F6 1 1 1 1 1

Product V VI 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1

4.3 First Case Study. A Mechanical Bracket Built Using FDM

123

Table 4.12 (continued) E_M4F2 E_M4F4 E_M4F6 E_V1F1 E_V2F3 E_V2F4 E_V3F4 E_V4F1

0 1 1 1 0 1 1 1

0 1 1 1 0 1 1 1

1 1 1 1 0 1 1 1

1 1 1 1 0 1 1 1

1 1 1 1 0 1 1 1

1 1 1 1 0 1 1 1

1 1 1 1 1 1 1 1

Table 4.13 Rankings of the activated actions for all the iterations of the redesign/reconfiguration package generation procedure — table MOD6_T2 Domain

Design

Actions Label

AD_M1F1 AD_M2F2 AD_M2F3 AD_M2F4_a AD_M2F4_b AD_M3F2 AD_M3F6 AD_M4F2 AD_M4F4 AD_M4F6 AD_V3F4 Manufacturing AM_M1F1 AM_M2F3 AM_M2F4_a AM_M2F4_b AM_M2F5_a AM_M2F5_b AM_M3F5_a AM_M3F5_b AM_M3F6 Verification AV_V2F3_a AV_V2F3_b AV_V2F4_a AV_V2F4_b AV_V4F1

Ranking of the activated actions (iterations) Product Dynamic Product model coefficient I IV II III V VI Split part A Split part B 2 3° 2° 1° 2 2° 1° 1 2 1 2 2° 1° 2 2 2° 1 2 1 2 3° 2° 1° 1° 1° 4 4 2 4 1 4 1° 1 4° 4 3 2° 2 1° 3 2 3 -

In the first iteration, “Change in the orientation of the model” was the action with the highest dynamic coefficient; for this reason it was selected and simulated, to obtain the model ready to be analysed in the second iteration of the procedure. Product features were evaluated again and new actions activated. A new ranking was defined as in column II of Table 4.13, and the action “Over-dimension thin parts” was simulated. Next iterations suggested the actions of splitting the model — and merging the parts afterwards. It should be noted that, at this stage, only one

124

4 Adopting the DGLs-CF in the Field

rule is violated, so that the least dynamic action is selected, as explained in Chap. 3. In this way, rotating and inclining the measuring head in the verification workspace, full compatibility was reached. The set of actions activated time by time in the different iterations constituted the redesign/reconfiguration package associated to this product. In other words, this was the set of actions which allowed one to build the mechanical bracket using the FDM Stratasys Dimension and to verify it by the CMM DEA Global Image 07-07-07. This redesign/reconfiguration package is shown in Table 4.14 — table MOD6_T3. Table 4.14 The redesign/reconfiguration package — table MOD6_T3 Execution Domain Action order 1° Manufacturing Orient the model to make the roughness resulting from slicing compatible with the requirements

2° 3°

4°

Warnings (related product features) Overhangs/sloped surfaces Cavities Surface finishing Mechanical properties Design Over-dimension thin parts to make them Minimum dimensions compatible with the need for supports, the Mechanical slicing and the material properties Design/ Use existing plane surfaces in splitting the Cavities Manufacturing model (and merge the product afterwards) to Bounding box make dimensions compatible with the manufacturing workspace Verification Rotate and incline the measuring head to Overhangs/sloped obtain minimum repositioning of the surfaces product Cavities

Following the actions collected in the redesign/reconfiguration package, the model was oriented in the manufacturing workspace to obtain the requested roughness, thin parts were over-dimensioned and, finally, the model was split into two parts to achieve the compatibility with the manufacturing workspace. The top of Fig. 4.6 shows the two split parts in the manufacturing workspace. Then the two parts were merged and the whole product was properly measured by rotating and inclining the measuring head in the verification workspace, as shown in the bottom of Fig. 4.6, to obtain minimum re-positioning of the product during measurement.

4.4 Second Case Study. A Mould Insert Built Using SLA

125

Fig. 4.6 The redesigned bracket at the end of the manufacturing activities (top) and during the verification with the CMM technology (bottom)

4.4 Second Case Study. A Mould Insert Built Using SLA In the second case study, the SLA manufacturing technology and the CMM verification technology were considered for the generation of the mould insert shown as the starting model in Fig. 4.7. This component is used in a mould to build plastic handles. The description of this case study runs exactly as the previous one’s, but the text focuses mainly on those issues that have not been covered in the previous case study.

126

4 Adopting the DGLs-CF in the Field

Fig. 4.7 The starting model of the mould insert used as product for this case study

A short overview of the main characteristics of the SLA manufacturing technology is reported hereafter, so it will be easier to understand the meaning of the parameters defined and used in the case study.

4.4.1 SLA Fundamentals As depicted in Fig. 4.8, in the RP technology called SLA, a mobile platform that can be lifted and lowered is located the thickness of a layer below the surface of a liquid photosensitive polymer contained in a tank. Each section of the model is etched onto the polymer surface by a laser beam that solidifies the polymer almost instantaneously. Once the laser has covered the whole surface of a layer, the platform lowers to a depth of another layer thickness, allowing the liquid resin to flow over the previously cured layer. A re-coating blade passes over the surface to ensure that a flat layer of liquid polymer is present, before the etching of the next layer. Supports are required when islands — portions of a layer disconnected from any other portion of the same layer — overhangs, or cantilevered sections exist in the prototype being built (Cheng et al. 1995). At the end of the building phase, the model is carefully removed from the platform, the liquid resin still present is drained, and a post-curing phase is performed in a UV — UltraViolet — beam oven to solidify completely the prototype (Banerjee et al. 2002).

4.4 Second Case Study. A Mould Insert Built Using SLA

127

Fig. 4.8 SLA process outline

Different building styles can be used with SLA, as shown in Fig. 4.9. Standard style builds full resin prototypes while other styles, for example the quickcast, leave some resin in the liquid state during the slice etching. This is done for different purposes, such as stresses minimization, generation of models for investment casting, etc. (Chockalingam et al. 2006).

Fig. 4.9 SLA building styles: standard (left), and quickcast (right)

SLA prototypes have good dimensional accuracy and surface texture; however, the orientation of the model in the workspace, due to the staircase effect, and the presence of supports can influence the surface finishing (Harris et al. 2002; Williams et al. 1996).

128

4 Adopting the DGLs-CF in the Field

4.4.2 First Setup MOD1 — Characteristic and Feature Collecting Module Considering the class of the SLA technologies described above, the main parametric manufacturing characteristics were collected by the manufacturers in table MOD1_T1, shown here in Table 4.15. Table 4.15 Parametric manufacturing characteristics of the class SLA — table MOD1_T1 Parametric manufacturing characteristic Label Name Description M1 Manufacturing Volume of the workspace manufacturing workspace M2

M3

M4 M5

M6

Mid-level parameters Manufacturing_workspace_x, Manufacturing_workspace_y, Manufacturing_workspace_z (maximum dimensions of the manufacturing workspace) Supports Critical Supports_x, Supports_y, Supports_z (minimum characteristics dimensions related to the need for supports) related to the need Supports_φ (minimum diameter related to the for supports when need for supports) building Supports_d (maximum depth or height related to overhangs/sloped the need for supports) surfaces or cavities Supports_α (minimum angle related to support removal) Supports_Ra_xy (minimum obtainable roughness in xy plane) Slicing Material deposed Slicing_zmin (minimum thickness of the slice) slice by slice Slicing_σ x, Slicing_ σ y, Slicing_σz (minimum strength in the three dimensions) Slicing_Ra_z (minimum obtainable roughness in z direction) Laser Characteristics of Laser_beam_φ (diameter of the laser beam) the laser beam Material Kind of material Material_σ (breaking strength of the material) used Material_Ra_xy (minimum obtainable roughness in xy plane) QC_building style Quickcast building QC_building_style_σx, QC_building_style_σy, style QC_building_style_σz (minimum strength in the three dimensions) QC_building_style_Ra_xy (minimum obtainable roughness in xy plane) QC_building_style_Ra_z (minimum obtainable roughness in z direction)

As said before, two different building styles may be defined in SLA, standard and quickcast, which could influence product properties such as mechanical characteristics and roughness. If the standard style is preferred, the manufacturing characteristics considered will be M1 to M5, which means that mechanical characteristics and roughness are affected mainly by the slicing and the kind of

4.4 Second Case Study. A Mould Insert Built Using SLA

129

material used. On the other hand, if the quickcast style is chosen, the considered manufacturing characteristics will be M1, M2, M4 and M6, which means that mechanical characteristics and roughness are affected mainly by the building style. It should be noted that SLA normally implies some shrinkage. Given that the model geometry is automatically compensated by the on-board software of the SLA equipment, shrinkage was not considered here. Table MOD1_T2 was the same as in the first case study. It is reported here in Table 4.16 both for the importance of the information represented in it and for the completeness of the case study description. Table 4.16 Parametric verification characteristics of the class CMM — table MOD1_T2 Verification characteristic Label Name Description V1 Verification Volume of the workspace verification workspace V2 V3 V4

Indexed measuring head Probe Clamping tools

Possible rotation and inclination of the measuring head Kind of probe used for verification Limitations due to the clamping tools

Mid-level parameters Verification_workspace_x, Verification_workspace_y, Verification_workspace_z (maximum dimensions of the verification workspace) Indexed_measuring_head_β (minimum angle of inclination of the of the head) Probe_φ (minimum diameter of the probe) Probe_l (maximum length of the probe) Clamping_tools_x, Clamping_tools_y, Clamping_tools_z (minimum dimensions defining the contacting area with the clamping tools)

Table 4.17 — table MOD1_T3 — shows the main parametric product features of the class of product mechanical parts, already identified in the previous case study and reported here again for completeness. They were evaluated this time with respect to the SLA technology and again to the CMM. Table 4.17 Parametric product features of the class mechanical parts — table MOD1_T3 Parametric product feature Label Name Description Mid-level parameters F1 Bounding box Overall dimensions Bounding_box_X, Bounding_box_Y, of the product Bounding_box_Z (maximum dimensions of the product) F2 Rib, pins, Minimum Minimum_dimensions_x, minimum dimensions in the Minimum_dimensions_y (minimum dimensions product dimensions in xy plane) Minimum_dimensions_z (minimum dimension in z direction) Minimum_dimensions_φ (minimum diameter) Minimum_dimensions _h (minimum height) F3 Webs, Overhangs and Overhangs_Sloped_surfaces_α overhangs/sloped protrusions (overhangs/sloped surfaces angle) /free form surfaces

130

4 Adopting the DGLs-CF in the Field

Table 4.17 (continued) F4

F5

F6

Cavities

Through and blind holes, Cavities_x, Cavities_y, Cavities_z undercuts and other (minimum dimensions for not cavities cylindrical cavities) Cavities_φ (minimum diameter for cylindrical cavities) Cavities_d (maximum depth for cylindrical cavities) Cavities_β (angle of inclination of the axis of the cavities) Surface finishing Surface texture Surface_finishing_Ra_xy (maximum allowable roughness in the xy plane) Surface_finishing_Ra_z (maximum allowable roughness in the z direction) Mechanical Main mechanical Mechanical_properties_σx, properties properties Mechanical_properties_σy, Mechanical_properties_σz (minimum strength in the three directions)

MOD2 — Rule and Action Generation Module Table 4.18 — table MOD2_T1 — shows the rules and the compatibility expressions derived by crossing the parametric technological characteristics and the parametric product features. Designers, manufacturers and inspectors performed this work together. Table 4.18 Rules and related compatibility expressions — table MOD2_T1 Rule Label Description R_M1F1 Dimensions defining the bounding box of the product must be smaller than maximum dimensions of the manufacturing workspace

Procedure Locate the features of size defining the bounding box in the model — gives the associated parameters Bounding_box_X, Bounding_box_Y, Bounding_box_Z. Identify the parameters defining the maximum dimensions of the building room — gives Manufacturing_workspace_x, Manufacturing_workspace_y, Manufacturing_workspace_z

Compatibility expressions E_M1F1=1 IF Bounding_box_Z< Manufacturing_workspace_z AND Bounding_box_X< Manufacturing_workspace_x AND Bounding_box_Y< Manufacturing_workspace_y ELSE E_M1F1=0

4.4 Second Case Study. A Mould Insert Built Using SLA

131

Table 4.18 (continued) Locate the features of size defining the minimum dimensions in the model — gives the associated parameters Minimum_dimensions_x, Minimum_dimensions_y, Minimum_dimensions_z, Minimum_dimensions_φ, Minimum_dimensions_h. Identify the parameters defining the critical dimensions related to the need for supports — gives Supports_x, Supports_y, Supports_z, Supports _d, Supports_φ Locate the features of size R_M2F3 The presence of overhangs/sloped defining the overhangs/sloped surfaces must be surfaces in the model — gives the associated parameter evaluated Overhang_sloped_surfaces_α. considering the need for supports Identify the parameters defining the critical inclination related to the presence of supports — gives Supports_α Locate the features of size R_M2F4 The dimensions and orientation of defining the minimum and/or the cavities must deepest cavities in the model — gives the associated parameters be compatible with the need for Cavities_φ, Cavities_x, Cavities_y, Cavities_z, supports Cavities_ β. Identify the parameters defining the critical dimensions and inclination related to the need for supports — gives Supports_φ, Supports_d, Supports_x, Supports_y, Supports_z R_M2F2 Minimum dimensions of the product must be bigger than minimum dimensions related to the presence of supports

E_M2F2=1 IF Minimum_dimensions_z> Supports_z AND Minimum_dimensions_x> Supports_x AND Minimum_dimensions_y> Supports_y AND Minimum_dimensions_φ> Supports_φ AND Minimum_dimensions_h< Supports_d ELSE E_M2F2=0 E_M2F3=1 IF Overhang_sloped_surfaces_α> Supports_α ELSE E_M2F3=0

BLIND HOLES E_M2F4=1 IF (Cavities_φ≤4•Supports_φ AND Cavities_x≤4•Supports_x AND Cavities_y≤4•Supports_y AND Cavities_z≤4•Supports_ z) OR (Cavities_φ<4•Supports_φ AND Cavities_x<4•Supports_x AND Cavities_y<4•Supports_y AND Cavities_z<4•Supports_ z AND Cavities_ β >Supports_ α ELSE E_M2F4=0) THROUGH HOLES E_M2F4=1 IF Cavities_φ≥2•Supports_φ AND Cavities_x≥2•Supports_x AND Cavities_y≥2•Supports_y AND Cavities_z≥2•Supports_ z ELSE E_M2F4=0

132

4 Adopting the DGLs-CF in the Field

Table 4.18 (continued) R_M2F5 Roughness must be evaluated considering the presence of supports R_M3F2 Minimum zthickness must be greater than minimum dimensions related to the slicing

E_M2F5=1 IF Surface_Finishing_Ra_xy> Supports_Ra_xy ELSE E_M2F5=0 Locate the features of size defining E_M3F2=1 IF the minimum dimensions Minimum_dimensions_z> orthogonal to x—y plane in the 15•Slicing_zmin model — gives the associated parameter Minimum_dimensions_z, AND Minimum_dimension_hmin> 15•Slicing_zmin Minimum_dimension_hmin. Identify the parameters defining the ELSE E_M3F2=0 critical dimensions related to the slicing — gives Slicing_zmin E_M3F5=1 IF Surface_Finishing_Ra_z> Slicing_Ra_z ELSE E_M3F5=0

R_M3F5 Maximum roughness of the product must be compared with the Minimum roughness related to the slicing R_M3F6 Mechanical properties must be considered in function of the slicing

Locate the features of size defining the minimum dimensions parallel to x—y plane in the model — gives the associated parameter Minimum_dimensions_x, Minimum_dimensions_y, Minimum_dimensions_φ. Identify the parameters defining the critical dimensions related to the diameter of the laser beam — gives laser_beam_φ Locate the features of size in the x-y R_M4F4 Minimum xydimensions of the plane defining the minimum cavities cavities must be in the model — gives the associated parameters Cavities_φ, or greater than the dimensions related Cavities_x, Cavities_y. Identify the parameters defining the to the characteristics of critical dimensions related to the diameter of the laser beam — gives the laser beam laser_beam_φ R_M4F2 Minimum xydimensions must be greater than the dimensions related to the characteristics of the laser beam

E_M3F6=1 IF Mechanical_properties_σx< Slicing_σx AND Mechanical_properties_ σ y< Slicing_σy AND Mechanical_properties_σz< Slicing_σz ELSE E_M3F6=0 E_M4F2=1 IF Minimum_dimensions_x> 5•Laser_beam_φ AND Minimum_dimensions_y> 5•Laser_beam_φ AND Minimum_dimensions_φ> 5•Laser_beam_φ ELSE E_M4F2=0 E_M4F4=0 IF Cavities_φ>5•Laser_beam_φ AND Cavities_x>5•Laser_beam_φ AND Cavities_y>5•Laser_beam_φ ELSE E_M4F4=0

4.4 Second Case Study. A Mould Insert Built Using SLA

133

Table 4.18 (continued) R_M5F5 Maximum roughness of the product must be compared with the minimum roughness related to the material R_M5F6 Mechanical properties must be compared with those related to the material used in the SLA technology

R_M6F5 Surface finishing must be considered in function of the QC_building style (when applied)

R_M6F6 Mechanical properties must be considered in function of the QC_building style (when applied)

R_V1F1 Dimensions defining the bounding box of the product must be smaller than maximum dimensions of the verification workspace

E_M5F5=1 IF Surface_Finishing_Ra_xy> Material_Ra_xy ELSE E_M5F5=0

E_M5F6=1 IF Mechanical_properties_σx< Material_σ AND Mechanical_properties_σy< Material_σ AND Mechanical_properties_σz< Material_σ ELSE E_M5F6=0 E_M6F5=1 IF Surface_finishing_Ra_xy> QC_building_style_Ra_xy AND Surface_finishing_Ra_z> QC_building _style_Ra_z ELSE E_M6F5=0 E_M6F6=1 IF Mechanical_properties_ σx< QC_building_style_ σx AND Mechanical_properties_ σy< QC_building_style_ σy AND Mechanical_properties_ σz< QC_building_style_σ z ELSE E_M6F6=0 Locate the features of size defining E_V1F1=1 IF the bounding box in the model — Bounding_box_Z< gives the associated parameters Verification_workspace_z AND Bounding_box_X, Bounding_box_X< Bounding_box_Y, Verification_workspace_x AND Bounding_box_Z. Identify the parameters defining the Bounding_box_Y< Verification_workspace_y maximum dimensions of the ELSE building room — gives E_V1F1=0 Verification_workspace_x, Verification_workspace_y, Verification_workspace_z

134

4 Adopting the DGLs-CF in the Field

Table 4.18 (continued) R_V2F3 The overhangs/sloped surfaces must be accessible by the measuring head, considering its inclination

R_V2F4 The cavities must be accessible by the measuring head, considering its inclination

R_V3F4 The dimensions of the cavities must be compatible with the diameter and length of the probe

R_V4F1 Dimensions defining the bounding box of the product must be compatible with the need for clamping tools

Locate the features of size defining the overhangs/sloped surfaces in the model — gives the associated parameter Overhang_sloped_surfaces_α. Identify the parameters defining the critical inclination related to the measuring head — gives Indexed_measuring_head_β Locate the features of size defining the inclination of the cavities in the model — gives the associated parameter Cavities_β. Identify the parameters defining the critical inclination related to the measuring head — gives Indexed_measuring_head_β Locate the features of size defining the minimum and/or deepest cavities in the model — gives the associated parameters Cavities_φ, Cavities_d or Cavities_x Cavities_y, or Cavities_z. Identify the parameters defining the critical dimensions and inclination related to the probe — gives Probe_φ, Probe_l

E_V2F3=1 IF Overhang_sloped_surfaces_α> Indexed_measuring_head_β ELSE E_V2F3=0

E_V2F4=1 IF Cavities_β> Indexed_measuring_head_β ELSE E_V2F4=0

E_V3F4=1 IF Cavities_φ>5•Probe_φ AND Cavities_d gives the associated parameters Clamping_tools_x AND Bounding_box_X, Bounding_box_Y> Bounding_box_Y, Clamping_tools_y AND Bounding_box_Z. Identify the parameters defining the Bounding_box_Z> Clamping_tools_z minimum dimensions of the ELSE clamping tools — gives E_V4F1=0 Clamping_tools_x, Clamping_tools_y, Clamping_tools_z

Table MOD2_T2 collected the actions derived from the rules of table MOD2_T1, generated again by designers, manufactures and inspectors, working together. This table is split into three parts corresponding to the domains where the actions applied — design, manufacturing, or verification — and also to the three

4.4 Second Case Study. A Mould Insert Built Using SLA

135

floors of the DGLs-CF building. Tables 4.19, 4.20 and 4.21 represent these three parts of table MOD2_T2 respectively. Table 4.19 Actions in the design domain — first part of table MOD2_T2 Action Domain Label Design

Link

Description Verb Accusative the model AD_M1F1 AM_M1F1 Use existing plane surfaces in splitting

AD_M2F2 -

AD_M2F3 -

Procedure Goal to make dimensions compatible with the manufacturing workspace

Locate the features of size related to those parameters giving E_M1F1=0. Locate the situation features (planes and/or axes and/or median planes) belonging to the model and orthogonal to the features of size previously identified. Split the model using planes containing these situation features Locate the features of Overthin parts to make them size related to those dimension compatible with parameters giving the need for E_M2F2=0. supports Locate two situation features planes orthogonal to the features of size previously identified. Over-dimension thin parts using these situation features Locate the features of to minimise the Overthe dimension overhangs/ problems related size related to those to support removal parameters giving sloped E_M2F3=0. surfaces Locate the situation feature plane corresponding to the critical inclination, in respect of the need for supports. Over-dimension the overhangs/sloped surfaces using this situation feature

136

4 Adopting the DGLs-CF in the Field

Table 4.19 (continued) Design

Use existing plane surfaces in splitting

AD_M2F4_b -

Overthe dimension cavities

AD_M3F2

Overthin parts dimension

-

Locate the features of size related to those parameters giving E_M2F4_a=0. Locate the situation features (planes and/or axes and/or median planes) belonging to the model and orthogonal to the features of size previously identified. Split the model using planes containing these situation features to eliminate Locate the features of their critical size related to those parameters giving dimensions E_M2F4_b=0. Locate two situation features planes orthogonal to the features of size previously identified or locate a situation feature cylinder coaxial with the features of size previously identified. Over-dimension the cavities using these situation features to make them Locate the features of size related to those compatible parameters giving with the E_M2F3=0. slicing Locate two situation features planes orthogonal to the features of size previously identified. Over-dimension thin parts using these situation features

the model to avoid the need for supports

AD_M2F4_a AM_M2F4_b

4.4 Second Case Study. A Mould Insert Built Using SLA

137

Table 4.19 (continued) Design AD_M3F6 -

Overthin dimension parts

to make them conform with the requested mechanical properties compatibly with the slicing

AD_M4F2 -

Overthin dimension parts

to make them compatible with the characteristics of the laser beam

AD_M4F4 -

Overthe to make them dimension cavities compatible with the characteristics of the laser beam

AD_M5F6 -

Overthin dimension parts

AD_V3F4 -

Overthe to obtain dimension cavities compatibility between the dimensions of the cavities and probes

to improve the mechanical strength

Locate the features of size related to those parameters giving E_M2F6=0. Locate two situation features planes orthogonal to the features of size previously identified. Over-dimension thin parts using these situation features Locate the features of size related to those parameters giving E_M4F2=0. Locate two situation features planes orthogonal to the features of size previously identified. Over-dimension thin parts using these situation features Locate the features of size related to those parameters giving E_M4F4=0. Locate a situation feature cylinder coaxial with the features of size previously identified. Over-dimension the cavities using these situation features Locate the features of size related to those parameters giving E_M5F6=0. Locate two situation features planes orthogonal to the features of size previously identified. Over-dimension thin parts using these situation features Locate the features of size related to those parameters giving E_M3F4=0. Locate two situation features planes orthogonal to the features of size previously identified or locate a situation feature cylinder coaxial with the features of size previously identified. Over-dimension the model using these situation features

138

4 Adopting the DGLs-CF in the Field

Table 4.20 Actions in the manufacturing domain — second part of table MOD2_T2 Action Domain

Label

Manufacturing AM_M1F1

AM_M2F3

Link AD_M1F1

-

AM_M2F4_a -

Description Verb Accusative Goal Merge the to join product the split parts

Orient the model to minimise the quantity of required supports

Orient the model to avoid the need for supports

AM_M2F4_b AD_M2F4_a Merge the product

to join the split parts

Procedure Locate the features of size related to those parameters giving E_M1F1=0. Locate the situation features (planes and/or axes and/or median planes) belonging to the model and orthogonal to the features of size previously identified. Merge the model using planes containing these situation features Locate the features of size related to those parameters giving E_M2F3=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the need for supports. Orient the model using this situation feature Locate the features of size related to those parameters giving E_M2F4_a=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the need for supports. Orient the model using this situation feature Locate the features of size related to those parameters giving E_M2F4_b=0. Locate the situation features (planes and/or axes and/or median planes) belonging to the model and orthogonal to the features of size previously identified. Merge the model using planes containing these situation features

4.4 Second Case Study. A Mould Insert Built Using SLA

139

Table 4.20 (continued) Manuf. AM_M2F5_a -

AM_M2F5_b -

AM_M3F5_a -

Orient the model to avoid the need for supports on the surfaces that need good finishing

Grind the surfaces

to obtain better roughness Orient the model to make the roughness resulting from slicing compatible with the requirements

Locate the features of size related to those parameters giving E_M3F5_a=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the slicing. Orient the model using this situation feature -

AM_M3F5_b -

Grind the surfaces

AM_M3F6

Locate the features of size related to those parameters giving E_M3F6=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the slicing. Orient the model using this situation feature Orient the model to make the Locate the features of size related to those parameters roughness related to the giving E_M5F5_a=0. Locate the situation feature plane material compatible corresponding to the critical inclination, in respect of the with the requirements roughness resulting from the material. Orient the model using this situation feature Grind the to obtain surfaces better roughness

AM_M5F5_a -

AM_M5F5_b -

to obtain better roughness Orient the model to obtain the best mechanical properties compatibly with the slicing

Locate the features of size related to those parameters giving E_M2F5_a=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the need for supports. Orient the model using this situation feature -

140

4 Adopting the DGLs-CF in the Field

Table 4.20 (continued) Manuf, AM_M6F5_a -

AM_M6F5_b -

AM_M6F6

-

Orient the model to make the roughness resulting from slicing compatible with the requirements

Grind the surface

to obtain best roughness Orient the model to obtain the best mechanical properties in function of the building style

Locate the features of size related to those parameters giving E_M6F5_a=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the QC_building style. Orient the model using this situation feature -

Locate the features of size related to those parameters giving E_M6F6=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the QC_building style. Orient the model using this situation feature

Table 4.21 Actions in the verification domain — third part of table MOD2_T2 Action Domain

Label

Link Description Verb Accusative Goal Verification AV_V2F3_a Orient the to achieve product easy access compatibly with the rotation and inclination of the measuring head

Procedure

AV_V2F3_b -

Locate the features of size related to those parameters giving E_V2F3_a=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the rotation and inclination of the measuring head. Orient the model using this situation feature -

AV_V2F4_a -

to obtain Rotate the measuring minimum reand positioning incline head of the product Orient the to obtain best product accessibility to the cavities

Locate the features of size related to those parameters giving E_V2F4_a=0. Locate the situation feature plane corresponding to the critical inclination, in respect of the rotation and inclination of the measuring head. Orient the model using this situation feature

4.4 Second Case Study. A Mould Insert Built Using SLA

141

Table 4.21 (continued) Verif.

AV_V2F4_b -

AV_V4F1

-

to obtain best Rotate the measuring accessibility to and the cavities and incline head the minimum repositioning Orient the to obtain the product best configuration compatible with the clamping tools

-

Locate the features of size related to those parameters giving E_V4F1=0. Locate the situation feature plane corresponding to the critical configuration, in respect of the clamping tools. Orient the model using this situation feature

MOD3 — Feature Relationship Discovery Module According to the procedure defined in the previous chapter, actions sharing the same verb-accusative pattern have been grouped to determine their dynamic coefficient, as shown in Table 4.22 — table MOD3_T1. Table 4.22 The groups of actions sharing the same verb-accusative pattern, and the derived relationships among product features with the dynamic coefficients — table MOD3_T1 Product feature mid-level Product parameters features related each other Cavities Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β Overhangs/ Overhangs_sloped_surfaces_α sloped surfaces Minimum Minimum_dimension_x, dimensions Minimum_dimensions_y, Minimum_dimensions_z, Minimum_dimensions_φ, Minimum_dimensions_h Mechanical Mechanical_propertis_σx, properties Mechanical_propertie_σy, Mechanical_properties_σz AD_M1F1 Cavities Cavities_x, Cavities_y, AD_M2F4_a Cavities_z, Cavities_φ, Cavities_d, Cavities_β Bounding Bounding_box_X, box Bounding_box_Y, Bounding_box_Z

Domain Group Actions label VerbLabels Accusative pattern Design GD_1 Over-dimension AD_M2F4_b the cavities AD_M4F4 AD_V3F4 GD_2 Over-dimension AD_M2F3 the overhangs/ sloped surfaces GD_3 Over-dimension AD_M2F2 thin parts AD_M3F2 AD_M3F6 AD_M4F2 AD_M5F6

GD_4 Use existing plane surfaces in splitting the model

Dyn. coeff.

1

1

2

2

142

4 Adopting the DGLs-CF in the Field

Table 4.22 (continued) Manuf.

GM_1 Orient the AM_M2F3 model AM_M2F4_a AM_M2F5_a AM_M3F5_a AM_M3F6 AM_M5F5_a AM_M6F5_a AM_M6F6

Overhangs/ sloped surfaces Cavities

Overhangs_sloped_surfaces_α 4

Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β Surface finishing Surface_finishing_Ra_xy, Surface_finishing_Ra_z Mechanical Mechanical_properties_σx, properties Mechanical_properties_σy, Mechanical_properties_σz

GM_2 Merge the AM_M1F1 Bounding box product AM_M2F4_b

Bounding_box_X, Bounding_box_Y, Bounding_box_Z Cavities Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β GM_3 Grind the AM_M2F5_b Surface finishing Surface_finishing_Ra_xy, Surface_finishing_Ra_z surfaces AM_M3F5_b AM_M5F5_b AM_M6F5_b Verification GV_1 Orient the AV_V2F3_a Bounding box Bounding_box_X, product AV_V2F4_a Bounding_box_Y, AV_V4F1 Bounding_box_Z Overhangs/ sloped surfaces Cavities

2

1

3

Overhangs_sloped_surfaces_α Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β

GV_2 Rotate and AV_V2F3_b Overhangs/sloped Overhangs_sloped_surfaces_α 2 incline the AV_V2F4_b surfaces measuring Cavities Cavities_x, Cavities_y, head Cavities_z, Cavities_φ, Cavities_d, Cavities_β

4.4.3 Technological Configuration MOD4 — Characteristic Data Input Module The specific technology used in this case study was the SLA3500 by 3D Systems (http://www.3dsystems.com). This is a medium-size workstation, equipped with a solid state laser and an automatic resin dispensing system. The SLA3500 is compatible with a range of different materials. It can be used for building prototypes and also for building style patterns to replace traditional wax patterns for investment casting processes. The SLA3500 needs a dedicated location in a

4.4 Second Case Study. A Mould Insert Built Using SLA

143

humidity controlled area. Prototypes are built layer by layer; supports use the same resin material but a different etching type, to allow an easy removal. Table 4.23 — table MOD4_T1 — contains the parameter values related to the SLA3500 and collected by manufacturers, while, again, Table 4.24 — table MOD4_T2 — is the same as in the first case study but it is reported here for information completeness. Table 4.23 Specific manufacturing characteristics of the SLA3500 — table MOD4_T1 Manufacturing Mid-level parameters characteristic M1 M2

M3

M4 M5 M6

a

Parameter values of the specific technology Manufacturing_workspace_x 350 mm Manufacturing_workspace_y 350 mm Manufacturing_workspace_z 300 mm Supports_x 2 mm Supports_y 2 mm Supports_z 2 mm 2 mm Supports_φ Supports_d 10 mm 60° Supports_α Supports_Ra_xy 2.5 μm Slicing_zmin 0.15 mm 38 MPaa Slicing_σx 38 MPaa Slicing_σy 35 MPaa Slicing_σz Slicing_Ra_z 9.5 μm 0.4 mm Laser_beam_φ 47 MPa Material_σM Material_Ra_xy 0.4 μm QC_Building_style_σx 30 MPa QC_Building_style_σy 30 MPa QC_Building_style_σz 27 MPa QC_Building_style_Ra_xy 3.2 μm QC_Building_style_Ra_z 12.5 μm

Experimental data

Table 4.24 Specific verification characteristics of the DEA Global Image 07-07-07 — table MOD4_T2 Verification Mid-level parameters characteristic V1 Verification_workspace_x Verification_workspace_y Verification_workspace_z V2 Indexed_measuring_head_β V3 Probe_φ Probe_l V4 Clamping_tools_x Clamping_tools_y Clamping_tools_z

Parameter values of the specific technology 700 mm 700 mm 660 mm 90° 1 mm 200 mm 5 mm 5 mm 5 mm

144

4 Adopting the DGLs-CF in the Field

4.4.4 Redesign/Reconfiguration Package Generation Figure 4.10 shows the drawing of the starting configuration of the mould insert considered in this case study. Its dimensional and geometrical specifications are defined on the drawing, while the required mechanical properties consisted in a minimum strength of 30 MPa in the three directions.

Fig. 4.10 The drawing of the starting configuration of the mould insert (only the elements meaningful for the case study are present)

MOD5 — Feature Data Input Module In the first iteration of the redesign/reconfiguration package generation procedure, the mould insert was supposed oriented as in the isometric view of Fig. 4.10, with respect to the datum system shown. According to this, the product features and their mid-level parameters were derived and collected by the designers as shown in column I of Table 4.25 — table MOD5_T1. The other columns (II, III, etc.) refer to the next iterations of the procedure. It should be noted that only the meaningful parameters are reported; for example, “Minimum dimensions” have been defined only in x, y, z directions and not by cylindrical features, given that there are no pins in this part.

4.4 Second Case Study. A Mould Insert Built Using SLA

145

Table 4.25 Specific product features for any iteration of the redesign/reconfiguration package generation procedure — table MOD5_T1 Product feature Label Mid-level parameters F1 F2

F3 F4

F5 F6

Parameter values (iterations) Product model Product I II III IV V Bounding_box_X 97 mm 97 97 97 97 Bounding_box_Y 49 mm 31 49 49 49 Bounding_box_Z 31 mm 49 31 31 31 Minimum_dimensions_x 3 mm 3 3 3 3 Minimum_dimensions_y 1 mm 4 1 4 4 Minimum_dimensions_z 4 mm 1 4 4 4 Minimum_dimensions_φ Minimum_dimensions_h 90° 160° 160° 160° Overhangs_sloped_surfaces_α 167° Cavities_x Cavities_y Cavities_z 5 mm 5 5 5 5 Cavities_φ Cavities_d 30 mm 30 30 30 30 90° 90° 90° 90° 90° Cavities_β Surface_finishing_Ra_xy 3.2 µm 3.2 3.2 3.2 3.2 Surface_finishing_Ra_z 3.2 µm 3.2 3.2 3.2 3.2 30 MPa 30 30 30 30 Mechanical_properties_σx 30 MPa 30 30 30 30 Mechanical_properties_σy 30 MPa 30 30 30 30 Mechanical_properties_σz

MOD6 — Redesign/Reconfiguration Package Generation Module In the first iteration of the procedure, some rules were violated and this is highlighted by the zero values of the compatibility expressions collected in column I of Table 4.26 — table MOD6_T1. Table 4.26 Compatibility values for each iteration of the redesign/reconfiguration package generation procedure — table MOD6_T1 Compatibility expressions Compatibility values (iterations) Product model Product I II III IV V VI E_M1F1 1 1 1 1 1 E_M2F2 0 0 0 1 1 E_M2F3 1 1 1 1 1 E_M2F4 1 1 1 1 1 E_M2F5 1 0 1 1 1 E_M3F2 1 0 1 1 1 E_M3F5 0 1 0 0 1 E_M3F6 1 1 1 1 1 E_M4F2 0 1 0 1 1 E_M4F4 1 1 1 1 1 E_M5F5 1 1 1 1 1 E_M5F6

1

1

1

1

1

-

146

4 Adopting the DGLs-CF in the Field

Table 4.26 (continued) E_M6F5 E_M6F6 E_V1F1 E_V2F3 E_V2F4 E_V3F4 E_V4F1

1 1 0 1 1

1 0 0 1 1

1 1 0 1 1

1 1 0 1 1

1 1 0 1 1

1 1 -

The values in the table highlight that the “Minimum dimensions” were not as large as needed considering the slicing and the need for supports, and “Surface finishing” was also affected by the slicing. Moreover, the “Cavities” were not easily accessible for measurement. Considering the dynamic coefficient values, the action “Orient the model” was simulated at the end of the first iteration, and the product features were evaluated again, and so the compatibility expressions. The results have been reported in column II of Table 4.26. However, this action did not solve the incompatibility; given that in this case an alternative action could be simulated, “Grind the product”, the procedure started from scratch excluding the action “Orient the model” for the next iteration. Thus, a new ranking was defined, as shown in column III of Table 4.27 — table MOD6_T2 — and so on. It should be highlighted that in this case there were actions showing the same dynamic coefficient value. As said in the previous chapter, this situation requires that alternative packages must be evaluated. However, here these actions were effectively the same action, “Over-dimension thin parts”, except for coming from two different rules. The action “Over-dimension thin parts” has been simulated and product features have been re-evaluated. Next iterations suggested performing the action “Grind the product” and then “Rotate and incline the measuring head” in the verification workspace, until full compatibility was reached. Table 4.27 Rankings of the activated actions for all the iterations of the redesign/reconfiguration package generation procedure — table MOD6_T2 Domain

Design

Actions Label

Dynamic coefficient AD_M1F1 2 AD_M2F2 2 AD_M2F3 1 AD_M2F4_a 2 AD_M2F4_b 1 AD_M3F2 2 AD_M3F6 2 AD_M4F2 2 AD_M4F4 1 AD_M5F6 2 AD_M6F6 2 AD_V3F4 1

Ranking of the activated actions (iterations) Product model Product I II III IV V 2° 2° 1° 2° 1° 2° -

4.4 Second Case Study. A Mould Insert Built Using SLA

147

Table 4.27 (continued) Manufacturing AM_M1F1 AM_M2F3 AM_M2F4_a AM_M2F4_b AM_M2F5_a AM_M2F5_b AM_M3F5_a AM_M3F5_b AM_M3F6 AM_M5F5_a AM_M5F5_b AM_M6F5_a AM_M6F5_b Verification AV_V2F3_a AV_V2F3_b AV_V2F4_a AV_V2F4_b AV_V4_F1

2 4 4 2 5 1 4 1 4 4 1 4 1 3 2 3 2 3

1° 3° -

1° 3° -

2° -

1° -

2° 1° -

The set of actions activated time by time in the different iterations constituted the redesign/reconfiguration package associated with this product. In other words, this was the set of actions which allowed one to build the mould insert using the SLA3500 and to verify it by the CMM DEA Global Image 07-07-07. This redesign/reconfiguration package is shown in Table 4.28 — table MOD6_T3. Table 4.28 The redesign/reconfiguration package — table MOD6_T3 Execution order 1° 2° 3°

Domain Design

Action

Warnings (related product features) Minimum dimensions Mechanical properties

Over-dimension thin parts to make them compatible with the need for supports and the slicing Manufacturing Grind the product to obtain the Surface finishing required roughness Verification Rotate and incline the measuring Overhangs/ sloped surfaces head to obtain best accessibility Cavities to cavities and the minimum repositioning

Following the actions collected in the redesign/reconfiguration package, thin parts were over-dimensioned to achieve the compatibility with the slicing. Finally, the product was grinded due to the roughness requirements and the measuring head was properly oriented and inclined in the verification workspace to obtain the best accessibility. The result is shown in Fig. 4.11 left side, while Fig. 4.11 right side shows the plastic handle generated by the mould where a metallic replica of the insert was placed.

148

4 Adopting the DGLs-CF in the Field

Fig. 4.11 The redesigned mould insert (left) and the plastic handle generated by a metallic replica of it (right)

4.5 Third Case Study. An Angle-Shaped Connector Built Using SLS In the third case study, the SLS manufacturing technology and the CMM verification technology were considered for the generation of the angle-shaped connector shown as the starting model in Fig. 4.12.

Fig. 4.12 The starting model of the angle-shaped connector used as product in this case study

A short overview of the main characteristics of the SLS manufacturing technology is reported hereafter, so it will be easier to understand the meaning of the parameters defined and used in the case study.

4.5 Third Case Study. An Angle-Shaped Connector Built Using SLS

149

4.5.1 SLS Fundamentals As depicted in Fig. 4.13, the RP technology called SLS builds the prototype over a platform where a layer of plastic, metal or ceramic powder with particle size of about 50 m is spread and kept heated. A laser beam melts the powder particles selectively. As a layer is finished, the platform moves down by the thickness of one layer, approximately 0.10—0.15 mm, and a new layer of powder is spread on the previous one. When the laser generates the new layer, particles melt and bond to the previous layer. The process repeats until the prototype is complete. To avoid mechanical stresses, retirements, etc., the SLS usually uses the building style named skin&core. The first and last layers of the prototype and the boundaries of the intermediate layers are completely sintered, and this is the skin. The inner part of the intermediate layers is not fully sintered and some sort of 3D grid is generated. This is the core. Surrounding powder acts as supporting material for the objects but in any case additional supports are required for the generation of overhangs. At the end, the built volume cools down to room temperature; after that, it can be removed from the workspace. In this case, the removal of the prototype from the building platform and of the supports is not straightforward and requires special machining such as Electrical Discharge Machining — EDM. Sandblasting or other finishing manufacturing techniques are used to remove all un-sintered particles and to improve the final surface finishing of the prototypes (Cooper 2001; Gatto and Iuliano 1998).

Fig. 4.13 SLS process outline

The SLS process can consider different etching styles, deriving from different laser powers and moving speeds.

150

4 Adopting the DGLs-CF in the Field

SLS prototypes have average surface texture and dimensional accuracy; their quality is mainly influenced by the powder particle size (Agarwala et al. 1995; Pham and Dimov 2001).

4.5.2 First Setup MOD1 — Characteristic and Feature Collecting Module Considering the class of the SLS technologies described above, the main parametric manufacturing characteristics were collected by the manufacturers as shown in Table 4.29 — table MOD1_T1. The laser characteristics have been considered only regarding the diameter of the laser beam, as for SLA technology. The shrinkage effect of the material has not been considered because it is automatically controlled by the on-board software of SLS equipment. Table 4.29 Parametric manufacturing characteristics of the class SLS — table MOD1_T1 Parametric manufacturing characteristic Label Name Description M1 Manufacturing Volume of the workspace manufacturing workspace

Mid-level parameters Manufacturing_workspace_x, Manufacturing_workspace_y, Manufacturing_workspace_z (maximum dimensions of the manufacturing workspace) Supports_x, Supports_y, Supports_z (minimum dimensions related to the need for supports) Supports_φ (minimum diameter related to the need for supports) Supports_d (maximum depth or height related to the need for supports) Supports_α (minimum angle related to support removal) Slicing_zmin (minimum thickness of the slice) Slicing_ σ x, Slicing_σy, Slicing _σz (minimum strength in the three dimensions) Slicing_Ra_z (minimum obtainable roughness in z direction) Laser_beam_φ (diameter of the laser beam)

M2

Supports

Critical characteristics related to the need for supports when building overhangs/sloped surfaces or cavities

M3

Slicing

Material deposed slice by slice

M4

Laser

M5

Material

Characteristics of the laser beam Kind of material Material_σ (breaking strength of the material) used Material_powder_φ (particle medium diameter)

Table 4.30 — table MOD1_T2 — is the same as in the first case study but it is reported here for the importance of the information represented in it and for the completeness of the case study description.

4.5 Third Case Study. An Angle-Shaped Connector Built Using SLS

151

Table 4.30 Parametric verification characteristics of the class CMM — table MOD1_T2 Parametric verification characteristic Label Name Description Mid-level parameters V1 Verification Volume of the Verification_workspace_x, workspace verification Verification_workspace_y, Verification_workspace_z workspace (maximum dimensions of the verification workspace) V2 Indexed Possible rotation Indexed_measuring_head_β (minimum angle of measuring and inclination of inclination of the of the head) head the measuring head V3 Probe Kind of probe used Probe_φ (minimum diameter of the probe) for verification Probe_l (maximum length of the probe) V4 Clamping Limitations due to Clamping_tools_x, Clamping_tools_y, tools the clamping tools Clamping_tools_z (minimum dimensions defining the contacting area with the clamping tools)

Also in this case study, Table 4.31 — table MOD1_T3 — shows the main parametric product features of the class of product mechanical parts already used in the previous cases. Here they were related to the SLS technology and again to the CMM. Table 4.31 Parametric product features of the class “Mechanical parts” — table MOD1_T3 Parametric product feature Label Name Description F1 Bounding box Overall dimensions of the product F2 Rib, pins, Minimum minimum dimensions dimensions in the product

F3

F4

F5

F6

Mid-level parameters Bounding_box_X, Bounding_box_Y, Bounding_box_Z (maximum dimensions of the product)

Minimum_dimensions_x, Minimum_dimensions_y (minimum dimensions in xy direction) Minimum_dimensions_z (minimum dimension in z direction) Minimum_dimensions_φ (minimum diameter) Minimum_dimensions _h (minimum height) Webs, Overhangs Overhangs_Sloped_surfaces_α (overhangs/sloped Overhangs/sloped and surfaces angle) /free form protrusions surfaces Cavities Through Cavities_x, Cavities_y, Cavities_z (minimum and blind dimensions for not cylindrical cavities) holes, Cavities_φ (minimum diameter for cylindrical cavities) undercuts Cavities_d (maximum depth for cylindrical cavities) and other Cavities_β (angle of inclination of the axis of the cavities cavities) Surface finishing Surface Surface_finishing_Ra_xy (maximum allowable texture roughness in xy plane) Surface_finishing_Ra_z (maximum allowable roughness in z direction) Mechanical Main Mechanical_properties_σx, Mechanical_properties_σy, properties mechanical Mechanical_properties_σz (minimum strength in the properties three directions)

152

4 Adopting the DGLs-CF in the Field

MOD2 — Rule and Action Generation Module Table 4.32 — table MOD2_T1 — shows the rules and the compatibility expressions derived by crossing the parametric technological characteristics and the parametric product features. Designers, manufacturers and inspectors performed this work together. Table 4.32 Rules and related compatibility expressions — table MOD2_T1 Rule Label Description R_M1F1 Dimensions defining the bounding box of the product must be smaller than maximum dimensions of the manufacturing workspace

R_M2F2 Minimum dimensions of the product must be bigger than minimum dimensions related to the presence of supports

R_M2F3 The presence of overhangs/slo ped surfaces must be evaluated considering the need for supports

Procedure Locate the features of size defining the bounding box in the model — gives the associated parameters Bounding_box_X, Bounding_box_Y, Bounding_box_Z. Identify the parameters defining the maximum dimensions of the building room — gives Manufacturing_workspace_x, Manufacturing_workspace_y, Manufacturing_workspace_z Locate the features of size defining the minimum dimensions in the model — gives the associated parameters Minimum_dimensions_x, Minimum_dimensions_y, Minimum_dimensions_z, Minimum_dimensions_φmin, Minimum_dimensions_h. Identify the parameters defining the critical dimensions related to the need for supports — gives Supports_x, Supports_y, Supports_z, Supports _d, Supports_φ Locate the features of size defining the overhangs/sloped surfaces in the model — gives the associated parameter Overhang_sloped_surfaces_α. Identify the parameters defining the critical inclination related to the presence of supports — gives Supports_α

Compatibility expressions E_M1F1=1 IF Bounding_box_Z< Manufacturing_workspace_z AND Bounding_box_X< Manufacturing_workspace_x AND Bounding_box_Y< Manufacturing_workspace_y ELSE E_M1F1=0

E_M2F2=1 IF Minimum_dimensions_z>Supports_z AND Minimum_dimensions_x>Supports_x AND Minimum_dimensions_y>Supports_y AND Minimum_dimensions_φ>Supports_φ AND Minimum_dimensions_h<Supports_d ELSE E_M2F2=0

E_M2F3=1 IF Overhang_sloped_surfaces_α> Supports_α ELSE E_M2F3=0

4.5 Third Case Study. An Angle-Shaped Connector Built Using SLS

153

Table 4.32 (continued) R_M2F4 The dimensions, depth and orientation of the cavities must be compatible with the need for supports

R_M3F2 Minimum z-thickness must be greater than minimum dimensions related to the slicing

R_M3F5 Maximum roughness of the product must be compared with the Minimum roughness related to the slicing

Locate the features of size defining the minimum and/or deepest cavities in the model — gives the associated parameters Cavities_φ, Cavities_x, Cavities_y, Cavities_z, and Cavities_β. Identify the parameters defining the critical dimensions and inclination related to the need for supports — gives Supports_φ, Supports_x, Supports_y, Supports_z, Supports_d, and Supports_α

BLIND HOLES E_M2F4=1 IF (Cavities_φ≤3•Supports_φ AND Cavities_x≤3•Supports_x AND Cavities_y≤3•Supports_y AND Cavities_z≤3•Supports_ z) OR (Cavities_φ<3•Supports_φ AND Cavities_x<3•Supports_x AND Cavities_y<3•Supports_y AND Cavities_z<3•Supports_ z AND Cavities_ β >Supports_ α ELSE E_M2F4=0 THROUGH HOLES

E_M2F4=1 IF Cavities_φ≥1.5•Supports_φ AND Cavities_x≥1.5•Supports_x AND Cavities_y≥1.5•Supports_y AND Cavities_z≥1.5•Supports_ z ELSE E_M2F4=0 E_M3F2=1 Locate the features of size IF defining the minimum Minimum_dimensions_z> dimensions orthogonal to x—y plane in the model — 20•Slicing_zmin AND Minimum_dimension_hmin> gives the associated 20•Slicing_zmin parameter ELSE Minimum_dimensions_z, Minimum_dimension_hmin. E_M3F2=0 Identify the parameters defining the critical dimensions related to the slicing — gives Slicing_zmin E_M3F5=1 IF Surface_Finishing_Ra_z> Slicing_Ra_z ELSE E_M3F5=0

154

4 Adopting the DGLs-CF in the Field

Table 4.32 (continued) R_M3F6 Mechanical properties must be considered in function of the slicing R_M4F2 Minimum xydimensions must be greater than the dimensions related to the characteristics of the laser beam

Locate the features of size defining the minimum dimensions parallel to x—y plane in the model — gives the associated parameter Minimum_dimensions_x, Minimum_dimensions_y, Minimum_dimensions_φ. Identify the parameters defining the critical dimensions related to the diameter of the laser beam — gives laser_beam_φ Locate the features of size in the x—y plane defining the minimum cavities in the model — gives the associated parameters Cavities_φ, or Cavities_x, Cavities_y. Identify the parameters defining the critical dimensions related to the diameter of the laser beam — gives laser_beam_φ -

R_M4F4 Minimum xydimensions of the cavities must be greater than the dimensions related to the characteristics of the laser beam R_M5F5 Maximum roughness of the product must be compared with the minimum roughness related to the material R_M5F6 Mechanical properties must be compared with those related to the material used in the SLS technology

E_M3F6=1 IF Mechanical_properties_σx<Slicing_σx AND Mechanical_properties_σy<Slicing_σy AND Mechanical_properties_σz<Slicing_σz ELSE E_M3F6=0 E_M4F2=1 IF Minimum_dimensions_x>8•Laser_beam_φ AND Minimum_dimensions_y>8•Laser_beam_φ AND Minimum_dimensions_φ>8•Laser_beam_φ ELSE E_M4F2=0

E_M4F4=0 IF Cavities_φ>8•Laser_beam_φ AND Cavities_x>8•Laser_beam_φ AND Cavities_y>8•Laser_beam_φ ELSE E_M4F4=0

E_M5F5=1 IF Surface_finishing_Ra_xy≥ Material_powder_φ ELSE E_M5F5=0

E_M5F6=1 IF Mechanical_properties_σx<Material_σ AND Mechanical_properties_σy<Material_σ AND Mechanical_properties_σz<Material_σ ELSE E_M5F6=0

4.5 Third Case Study. An Angle-Shaped Connector Built Using SLS

155

Table 4.32 (continued) R_V1F1 Dimensions defining the bounding box of the product must be smaller than maximum dimensions of the verification workspace

R_V2F3 The overhangs/slo ped surfaces must be accessible by the measuring head, considering its inclination

R_V2F4 The cavities must be accessible by the measuring head, considering its inclination

R_V3F4 The dimensions of the cavities must be compatible with the diameter and length of the probe

Locate the features of size defining the bounding box in the model — gives the associated parameters Bounding_box_X, Bounding_box_Y, Bounding_box_Z. Identify the parameters defining the maximum dimensions of the building room — gives Verification_workspace_x, Verification_workspace_y, Verification_workspace_z Locate the features of size defining the overhangs/sloped surfaces in the model — gives the associated parameter Overhang_sloped_surfaces _α. Identify the parameters defining the critical inclination related to the measuring head — gives Indexed_measuring_head_ β Locate the features of size defining the inclination of the cavities in the model — gives the associated parameter Cavities_β. Identify the parameters defining the critical inclination related to the measuring head — gives Indexed_measuring_head_ β Locate the features of size defining the minimum and/or deepest cavities in the model — gives the associated parameters Cavities_φ, Cavities_d or Cavities_x Cavities_y, or Cavities_z. Identify the parameters defining the critical dimensions and inclination related to the probe — gives Probe_φ, Probe_l

E_V1F1=1 IF Bounding_box_Z
E_V2F3=1 IF Overhang_sloped_surfaces_α>Indexed_measu ring_head_β ELSE E_V2F3=0

E_V2F4=1 IF Cavities_β>Indexed_measuring_head_β ELSE E_V2F4=0

E_V3F4=1 IF Cavities_φ>5•Probe_φ AND Cavities_d
156

4 Adopting the DGLs-CF in the Field

Table 4.32 (continued) R_V4F1 Dimensions defining the bounding box of the product must be compatible with the need for clamping tools

Locate the features of size defining the bounding box in the model — gives the associated parameters Bounding_box_X, Bounding_box_Y, Bounding_box_Z. Identify the parameters defining the minimum dimensions of the clamping tools — gives Clamping_tools_x, Clamping_tools_y, Clamping_tools_z

E_V4F1=1 IF Bounding_box_X>Clamping_tools_x AND Bounding_box_Y>Clamping_tools_y AND Bounding_box_Z>Clamping_tools_z ELSE E_V4F1=0

Table MOD2_T2 collected the actions derived from the rules of table MOD2_T1, generated again by designers, manufactures and inspectors, working together. This table is split into three parts corresponding to the domains where the actions applied — design, manufacturing, or verification — and also to the three floors of the DGLs-CF building. Tables 4.33, 4.34 and 4.35 represent these three parts of table MOD2_T2 respectively. Table 4.33 Actions in the design domain — first part of table MOD2_T2 Action Domain Label

Link

Design AD_M1F1

AM_M1F1

Description Verb Accusative the model Use existing plane surfaces in splitting

Procedure Goal Locate the features of to make size related to those dimensions parameters giving compatible E_M1F1=0 with the manufacturing Locate the situation features (planes and/or workspace axes and/or median planes) belonging to the model and orthogonal to the features of size previously identified Split the model using planes containing these situation features

4.5 Third Case Study. An Angle-Shaped Connector Built Using SLS

157

Table 4.33 (continued) Locate the features of size related to those parameters giving E_M2F2=0 Locate two situation features planes orthogonal to the features of size previously identified Over-dimension thin parts using these situation features to minimise Locate the features of size AD_M2F3 Overthe dimension overhangs/ the problems related to those parameters giving E_M2F3=0 related to sloped Locate the situation feature support surfaces plane corresponding to the removal critical inclination, in respect of the need for supports Over-dimension the overhangs/sloped surfaces using this situation feature the model to avoid the Locate the features of size AD_M2F4_a AM_M2F4_b Use related to those parameters need for existing giving E_M2F4_a=0 supports plane Locate the situation surfaces features (planes and/or in axes and/or median planes) splitting belonging to the model and orthogonal to the features of size previously identified Split the model using planes containing these situation features AD_M2F4_b Overthe cavities to eliminate Locate the features of size dimension their critical related to those parameters dimensions giving E_M2F4=0 Locate two situation features planes orthogonal to the features of size previously identified or locate a situation feature cylinder coaxial with the features of size previously identified Over-dimension the cavities using these situation features

Design AD_M2F2

-

Overthin parts dimension

to make them compatible with the need for supports

158

4 Adopting the DGLs-CF in the Field

Table 4.33 (continued) Design AD_M3F2 -

AD_M3F6 -

AD_M4F2 -

AD_M4F4 -

AD_M5F6 -

AD_V3F4 -

Locate the features of size related to those parameters giving E_M2F3=0 Locate two situation features planes orthogonal to the features of size previously identified Over-dimension thin parts using these situation features Overthin to make them Locate the features of size related to dimension parts conform with those parameters giving E_M2F6=0 the requested Locate two situation features planes orthogonal to the features of size mechanical previously identified properties Over-dimension thin parts using these compatibly situation features with the slicing Overthin to make them Locate the features of size related to those parameters giving E_M4F2=0 dimension parts compatible Locate two situation features planes with the characteristics orthogonal to the features of size previously identified of the laser Over-dimension thin parts using these beam situation features Overthe to make them Locate the features of size related to those parameters giving E_M4F4=0 dimension cavities compatible Locate two situation features planes with the characteristics orthogonal to the features of size previously identified or locate a of the laser situation feature cylinder coaxial with beam the features of size previously identified Over-dimension the cavities using these situation features Overthin to improve the Locate the features of size related to those parameters giving E_M5F6=0 dimension parts mechanical Locate two situation features planes strength orthogonal to the features of size previously identified Over-dimension thin parts using these situation features Locate the features of size related to Overthe to obtain dimension cavities compatibility those parameters giving E_M3F8=0 between the Locate two situation features planes dimensions of orthogonal to the features of size previously identified or locate a the cavities situation feature cylinder coaxial with and probes the features of size previously identified Over-dimension the model using these situation features Overthin dimension parts

to make them compatible with the slicing

4.5 Third Case Study. An Angle-Shaped Connector Built Using SLS

159

Table 4.34 Actions in the manufacturing domain — second part of table MOD2_T2 Action Domain

Label

Link

Description Verb Accusative Goal

Procedure

to join Locate the features of size the split related to those parameters giving E_M1F1=0 parts Locate the situation features (planes and/or axes and/or median planes) belonging to the model and orthogonal to the features of size previously identified Merge the model using planes containing these situation features Locate the features of size AM_M2F3 Orient the model to minimise related to those parameters giving E_M2F3=0 the quantity Locate the situation feature plane corresponding to the of required critical inclination, in supports respect of the need for supports Orient the model using this situation feature AM_M2F4_a Orient the model to make Locate the features of size support related to those parameters removal giving E_M2F4_a=0 Locate the situation feature easier plane corresponding to the critical inclination, in respect of the need for supports Orient the model using this situation feature AM_M2F4_b AD_M2F4_a Merge the to join Locate the features of size product the split related to those parameters giving E_M2F4_b=0 parts Locate the situation features (planes and/or axes and/or median planes) belonging to the model and orthogonal to the features of size previously identified Merge the model using planes containing these situation features

Manufacturing AM_M1F1

AD_M1F1

Merge the product

160

4 Adopting the DGLs-CF in the Field

Table 4.34 (continued) Manuf.

AM_M3F5_a -

AM_M3F5_b -

AM_M3F6

-

AM_M5F5_a -

AM_M5F5_b -

Orient the model to make the Locate the features of size roughness related to those parameters giving E_M3F5_a=0 resulting from slicing Locate the situation feature compatible plane corresponding to the critical inclination, in respect with the requirements of the slicing Orient the model using this situation feature Grind the to obtain surfaces better roughness Orient the model to obtain the Locate the features of size related to those parameters best mechanical giving E_M3F6=0 properties Locate the situation feature compatibly plane corresponding to the critical inclination, in respect with the of the slicing slicing Orient the model using this situation feature Orient the model to make the roughness resulting from material powder compatible with the requirements Grind the to obtain surface best roughness

Table 4.35 Actions in the verification domain — third part of table MOD2_T2 Action Domain

Label

Link

Verification AV_V2F3_a -

Description Procedure Verb Accusative Goal Locate the features of size Orient the to achieve product easy access related to those parameters compatibly giving E_V2F3_a=0 Locate the situation feature with the rotation and plane corresponding to the inclination of critical inclination, in respect of the rotation and inclination the measuring of the measuring head Orient the model using this head situation feature

4.5 Third Case Study. An Angle-Shaped Connector Built Using SLS

161

Table 4.35 (continued) Verif. AV_V2F3_b -

AV_V2F4_a -

to obtain Rotate the measuring minimum reand positioning of the incline head product Orient the to obtain best product accessibility to the cavities

Locate the features of size related to those parameters giving E_V2F4_a=0 Locate the situation feature plane corresponding to the critical inclination, in respect of the rotation and inclination of the measuring head Orient the model using this situation feature -

to obtain best Rotate the measuring accessibility to and the cavities and incline head the minimum repositioning AD_V4F1 Orient the to obtain the best product configuration compatible with the clamping tools

AV_V2F4_b -

AV_V4F1

-

MOD3 — Feature Relationship Discovery Module According to the procedure defined in the previous chapter, actions sharing the same verb-accusative pattern have been grouped to determine their dynamic coefficient as shown in Table 4.36 — table MOD3_T1. Table 4.36 The groups of actions sharing the same verb-accusative pattern and the derived relationships among product features with the dynamic coefficients — table MOD3_T1 Domain Group Actions label VerbAccusative pattern Design GD_1 Overdimension the cavities GD_2 Overdimension the overhangs/ sloped surfaces

Labels

Product features Product feature mid-level parameters related each other

AD_M4F4 Cavities AD_V3F4 AD_M2F4_b AD_M2F3

Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β

Dyn. coeff.

1

Overhangs/sloped Overhangs_sloped_surfaces_α 1 surfaces

162

4 Adopting the DGLs-CF in the Field

Table 4.36 (continued) Design

AD_M2F2 GD_3 OverAD_M3F2 dimension the thin parts AD_M3F6 AD_M4F2 AD_M5F2 AD_M5F6

GD_4 Use existing AD_M1F1 AD_M2F4_a plane surfaces in splitting the model Manuf.

2 Minimum_dimension_x, Minimum_dimensions_y, Minimum_dimensions_z, Minimum_dimensions_φ, Minimum_dimensions_h Mechanical Mechanical_propertis_σx, properties Mechanical_propertie_σy, Mechanical_properties_σz 2 Bounding box Bounding_box_X, Bounding_box_Y, Bounding_box_Z Cavities Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β Overhangs/sloped Overhangs_sloped_surfaces_α 4 surfaces Cavities Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β Surface finishing Surface_finishing_Ra_xy, Surface_finishing_Ra_z Mechanical Mechanical_properties_σx, properties Mechanical_properties_σy, Mechanical_properties_σz Minimum dimensions

GM_1 Orient the model

AM_M2F3 AM_M2F4_a AM_M3F5_a AM_M3F6 AM_M5F5_a

GM_2 Merge the product

AM_M1F1 Bounding box AM_M2F4_b Cavities

GM_3 Grind the surfaces Verification GV_1 Orient the product

GV_2 Rotate and incline the measuring head

Bounding_box_X, Bounding_box_Y, Bounding_box_Z Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β

AM_M3F5_b Surface finishing Surface_finishing_Ra_xy, AM_M5F5_b Surface_finishing_Ra_z AV_V2F3_a Bounding box AV_V2F4_a AV_V4F1 Overhangs/sloped surfaces Cavities

2

1

3 Bounding_box_X, Bounding_box_Y, Bounding_box_Z Overhangs_sloped_surfaces_α

Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β AV_V2F3_b Overhangs/sloped Overhangs_sloped_surfaces_α 2 AV_V2F4_b surfaces Cavities Cavities_x, Cavities_y, Cavities_z, Cavities_φ, Cavities_d, Cavities_β

4.5 Third Case Study. An Angle-Shaped Connector Built Using SLS

163

4.5.3 Technological Configuration MOD4 — Characteristic Data Input Module The specific technology used in this case study was the SLS EOSINT M250X by EOS gmbh (http://www.eos.info). This is a direct metal laser sintering equipment, able to build tools and metal prototypes using different metal powders. This equipment needs a dedicated and humidity controlled location. The building chamber must be maintained in an inactive gas atmosphere having a low oxygen concentration. Table 4.37 — table MOD4_T1 — shows the parameter values related to this equipment and collected by the manufacturers while, again, Table 4.38 — table MOD4_T2 — is the same as in the first case study but it is reported here for information completeness. Table 4.37 Specific manufacturing characteristics of the EOSINT M250X — table MOD4_T1 Manuf. char.

Mid-level parameters

Parameter values of the specific technology

M1

Manufacturing_workspace_x Manufacturing_workspace_y Manufacturing_workspace_z Supports_x Supports_y Supports_z Supports_φ Supports_d Supports_α Slicing_zmin Slicing_σx Slicing_σy Slicing_σz Slicing_Ra_z Laser_beam_φ Material_σM Material_powder_φ

250 mm 250 mm 185 mm 3 mm 3 mm 3 mm 3 mm 10 mm 60° 0.10 mm 490 MPaa 490 MPaa 490 MPaa 15 μm 0.4 mm 500 MPa 50 μm

M2

M3

M4 M5 a

Experimental data

164

4 Adopting the DGLs-CF in the Field

Table 4.38 Specific verification characteristics of the DEA Global Image 07-07-07 — table MOD4_T2 Verif. char. V1

V2 V3 V4

Mid-level parameters Verification_workspace_x Verification_workspace_y Verification_workspace_z Indexed_measuring_head_β Probe_φ Probe_l Clamping_tools_x Clamping_tools_y Clamping_tools_z

Parameter values of the specific technology 700 mm 700 mm 660 mm 90° 1 mm 200 mm 5 mm 5 mm 5 mm

4.5.4 Redesign/Reconfiguration Package Generation Figure 4.14 shows the drawing of the starting configuration of the angle-shaped connector considered in this case study. Its dimensional and geometrical specifications are defined on the drawing, while the required mechanical properties consisted in a minimum strength of 400 MPa in the three directions.

Fig. 4.14 The drawing of the starting configuration of the angle-shaped connector (only the elements meaningful for the case study are present)

4.5 Third Case Study. An Angle-Shaped Connector Built Using SLS

165

MOD5 — Feature Data Input Module In the first iteration of the redesign/reconfiguration package generation procedure, the mould insert was supposed oriented as in the isometric view of Fig. 4.14, with respect to the datum system shown. According to this, the mid-level product features and their parameters were derived and collected by the designers as shown in column I of Table 4.39 — table MOD5_T1. The other columns (II, III, etc.) refer to the next iterations of the procedure. Table 4.39 Specific product features for any iteration of the redesign/reconfiguration package generation procedure — table MOD5_T1 Product feature Label Mid-level parameters F1

F2

F3 F4

F5 F6

Bounding_box_X Bounding_box_Y Bounding_box_Z Minimum_dimensions_x Minimum_dimensions_y Minimum_dimensions_z Minimum_dimensions_φ Minimum_dimensions_h Overhangs_sloped_surfaces_α Cavities_φ Cavities_d Cavities_β Surface_finishing_Ra_xy Surface_finishing_Ra_z Mechanical_properties_σx Mechanical_properties_σy Mechanical_properties_σz

Parameter values (iterations) Product model Product I II III IV 120 mm 120 120 120 66 66 66 66 mm 92 92 92 92 mm 8 mm 8 8 8 5 mm 4 3 4 4 mm 5 5 5 90° 4 mm 4 6 6 19 mm 19 19 19 90° 180° 180° 180° 400 MPa 400 400 400 400 MPa 400 400 400 400 MPa 400 400 400

MOD6 — Redesign/Reconfiguration Package Generation Module In the first iteration of the procedure some rules were violated and this is highlighted by the zero values of the compatibility expressions collected in column I of Table 4.40 — table MOD6_T1.

166

4 Adopting the DGLs-CF in the Field

Table 4.40 Compatibility values for any iteration of the redesign/reconfiguration package generation procedure — table MOD6_T1 Compatibility expressions

E_M1F1 E_M2F2 E_M2F3 E_M2F4 E_M3F2 E_M3F5 E_M3F6 E_M4F2 E_M4F4 E_M5F5 E_M5F6 E_V1F1 E_V2F3 E_V2F4 E_V3F4 E_V4F1

Compatibility values (iterations) Product model Product I II III IV 1 1 1 1 1 1 0 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 0 0 1 1 1 1 1 1

The values in the table highlight that the cavities showed some problems related to the need for supports and one of them was not compatible with the diameter of the probe and the inclination of the measuring head. Considering the dynamic coefficients of the activated actions, shown in Table 4.41 — table MOD6_T2, the most dynamic action suggested to change the orientation of the model in the manufacturing workspace to get an easier support removal.

4.5 Third Case Study. An Angle-Shaped Connector Built Using SLS

167

Table 4.41 Rankings of the activated actions for all the iterations of the redesign/reconfiguration package generation procedure — table MOD6_T2 Domain

Actions Label

Design

AD_M1F1 AD_M2F2 AD_M2F3 AD_M2F4_a AD_M2F4_b AD_M3F2 AD_M3F6 AD_M4F2 AD_M4F4 AD_M5F6 AD_V3F4 Manufacturing AM_M1F1 AM_M2F3 AM_M2F4_a AM_M2F4_b AM_M3F5_a AM_M3F5_b AM_M3F6 AM_M5F5a AM_M5F5_b Verification AV_V2F3_a AV_V2F3_b AV_V2F4_a AV_V2F4_b AV_V4F1

Ranking of the activated actions (iterations) Dynamic Product model Product coefficient I II III IV 2 2 1° 1 2 2° 1 3° 2 2 2 1° 1 2 1 3° 1° 2 4 4 1° 2 2° 4 1 4 4 1 3 2 3 2 3 -

After the simulation of the first action, “Orient the model”, the product features have been re-evaluated and thus the compatibility expressions. The results have been reported in column II of Table 4.41. This action solved the problems related to the need for supports and the inclination of the measuring head, although the incompatibility with the diameter of the probe remained. Thus, the new ranking suggested simulating the action “Over-dimension cavities”; after that, the product features were evaluated again. This time cavities were compatible with the dimensions of the probe, but the minimum dimensions derived by the cavities

168

4 Adopting the DGLs-CF in the Field

over-dimensioning were not compatible with the need for supports and with the laser beam characteristics, as it emerged by the new ranking in column III of Table 4.41. Finally, the action “Over-dimension minimum parts” allowed the obtaining of full compatibility of the redesigned part with the available technologies. The set of all the aforementioned actions constituted the redesign/reconfiguration package associated with this product. In other words, this was the set of actions which allowed one to build the angle-shaped connector using the SLS EOSINT M250X and to verify it by the CMM DEA Global Image 07-07-07. This redesign/reconfiguration package is shown in Table 4.42 — table MOD6_T3. Table 4.42 The redesign/reconfiguration package — table MOD6_T3 Execution Domain Action order 1° Manufacturing Orient the model to make support removal easier 2°

Design

3°

Design

Warnings (related product features) Overhangs/sloped surfaces Cavities Surface finishing Mechanical properties Cavities

Over-dimension the cavities to obtain compatibility between the dimensions of the cavities and probes Over-dimension thin parts to make Minimum dimensions them compatible with the Mechanical properties characteristics of the laser beam

Following the actions collected in the redesign/reconfiguration package, the model was oriented in the manufacturing workspace to obtain the required surface finishing, cavities were over-dimensioned to make their verification possible, and thin parts were over-dimensioned to make them compatible with the laser beam characteristics. The result of all of this is shown in Fig. 4.15.

Fig. 4.15 The redesigned angle-shaped connector (rendering of the digital model)

References

169

Summary The description of the DGLs-CF in the previous chapter gave the idea of a methodological approach to product redesign and process reconfiguration with a clean architecture, clear modules, well identified activities, and univocal roles for the actors involved. In this chapter the DGLs-CF has been applied in the field, using three different scenarios. The whole roadmap has been followed for each of them; the content of the tables of the modules has been set and described; the redesign/reconfiguration package has been applied and the result, the redesigned product, has been shown. These case studies gave many hints to evaluate the effectiveness of the DGLs-CF and to discuss the open issues and the problems still present. The next chapter describes them in detail.

References Agarwala M, Bourell D, Beaman J, Marcus H, Barlow J (1995), Direct Selective Laser Sintering of Metals. Rapid Prototyp J1(1):26-36 Banerjee PS, Sinha A, Banerjee MK (2002) A study on effect of variation of SLA process parameters over strength of Built model. In: Proc. of the 2nd National Symposium on Rapid Prototyping & Rapid Tooling Technologies :79-84 Bosch JA (1995) Coordinate Measuring Machines and Systems. Marcel Dekker Inc Byun HS, Lee KH (2005) A decision support system for the selection of a rapid prototyping process using the modified TOPSIS method. Int J Adv Manuf Technol 26:1338-1347 Cheng W, Fuh JYH, Nee AYC, Wong YS, Miyazawa T (1995) Multi objective optimization of part building orientation in stereolithography. Rapid Prototyp J 1(4):12-23 Chockalingam K, Jawahar N, Ramanathan KN, Banerjee PS (2006) Optimization of stereolithography process parameters for part strength using design of experiments. Int J Adv Manuf Technol 29(1-2):79-88 Chua CK, Leong KF, Lim CS (2003) Rapid Prototyping: Principles and Applications, 2nd edition. World Scientific Publishing Company Cooper K (2001) Rapid prototyping Technology: Selection and Application. eds. Taylor and Francis Group Ltd Cristofolini I and Podda G (2001) Diameter and Center Line of a 1st Reference Cylindrical Feature of Size with a Straightness Tolerance. In: Proc. XII ADM International Conference, Rimini (I), D1:28-37 Gatto A and Iuliano L (1998) Prototipazione rapida: la tecnologia per la competizione globale. eds. Tecniche Nuove, Italy Harris R, Hopkinso N, Newlyn H, Hague R, Dickens P (2002) Layer thickness and draft angle selection for stereolithography injection mould tooling. Int J Prod Res 40(3):719-729 Horvàth I, Broek JJ, Rus´ak Z, Kuczogi G, Vergeest JSM (1999) Morphological segmentation of objects for thick-layered manufacturing. In: Proc. of the 1999 ASME conference on design for manufacturing, Las Vegas :18-24 Hwang CY, Tsai C-Y, Chang CA (2004) Efficient inspection planning for coordinate measuring machines. Int J Adv Manuf Technol23(9-10):732-742 ISO 10360-1:2000 Geometrical Product Specifications (GPS) - Acceptance and reverification tests for coordinate measuring machines (CMM) - Part 1: Vocabulary ISO 10360-2:2001 Geometrical Product Specifications (GPS) - Acceptance and reverification tests for coordinate measuring machines (CMM) - Part 2: CMMs used for measuring size

170

4 Adopting the DGLs-CF in the Field

ISO 10360-3:2000 Geometrical Product Specifications (GPS) - Acceptance and reverification tests for coordinate measuring machines (CMM) - Part 3: CMMs with the axis of a rotary table as the fourth axis ISO 10360-4:2000 Geometrical Product Specifications (GPS) - Acceptance and reverification tests for coordinate measuring machines (CMM) - Part 4: CMMs used in scanning measuring mode ISO 10360-5:2000 Geometrical Product Specifications (GPS) - Acceptance and reverification tests for coordinate measuring machines (CMM) - Part 5: CMMs using multiple-stylus probing systems ISO 10360-6:2001 Geometrical Product Specifications (GPS) - Acceptance and reverification tests for coordinate measuring machines (CMM) - Part 6: Estimation of errors in computing Gaussian associated features ISO 14253-1:1998 Geometrical Product Specifications (GPS) - Inspection by measurement of workpieces and measuring equipment - Part 1: Decision rules for proving conformance or non-conformance with specifications ISO 14660-1:1999 Geometrical Product Specification (GPS) - Geometrical features - Part 1: General terms and definitions, 1999. ISO 14660-2:1999 Geometrical Product Specification (GPS) - Geometrical features - Part 2: Extracted median line of a cylinder and a cone, extracted median surface, local size of an extracted feature, 1999. ISO/TS 14253-2:1999 Geometrical Product Specifications (GPS) - Inspection by measurement of workpieces and measuring equipment - Part 2: Guide to the estimation of uncertainty in GPS measurement, in calibration of measuring equipment and in product verification ISO/TS 14253-3:2002 Geometrical Product Specifications (GPS) - Inspection by measurement of workpieces and measuring equipment - Part 3: Guidelines for achieving agreements on measurement uncertainty statements ISO/TS 15530-3:2004 Geometrical Product Specifications (GPS) - Coordinate measuring machines (CMM): Technique for determining the uncertainty of measurement - Part 3: Use of calibrated workpieces or standards, 2004 ISO/TS 15530-4:2008 Geometrical Product Specifications (GPS) - Coordinate measuring machines (CMM): Technique for determining the uncertainty of measurement - Part 4: Evaluating task-specific measurement uncertainty using simulation ISO/TS 16610-1:2006 Geometrical product specifications (GPS) - Filtration - Part 1: Overview and basic concepts ISO/TS 16610-20:2006 Geometrical product specifications (GPS) - Filtration - Part 20: Linear profile filters: Basic concepts ISO/TS 16610-22:2006 Geometrical product specifications (GPS) - Filtration - Part 22: Linear profile filters: Spline filters ISO/TS 16610-29:2006 Geometrical product specifications (GPS) - Filtration - Part 29: Linear profile filters: Spline wavelets ISO/TS 16610-40:2006 Geometrical product specifications (GPS) - Filtration - Part 40: Morphological profile filters: Basic concepts ISO/TS 16610-41:2006 Geometrical product specifications (GPS) - Filtration - Part 41: Morphological profile filters: Disk and horizontal line-segment filters ISO/TS 16610-49:2006 Geometrical product specifications (GPS) - Filtration - Part 49: Morphological profile filters: Scale space techniques ISO/TS 23165:2006 Geometrical product specifications (GPS) - Guidelines for the evaluation of coordinate measuring machine (CMM) test uncertainty Jacobs P F (1995) Stereolithography & Other Rp&m Technologies: From Rapid Prototyping to Rapid Tooling. eds. Society of Manufacturing Engineers Kweon S, Medeiros DJ (1998) Part orientations for CMM inspection using dimensioned visibility maps. CAD 30:741-749 Lin Z-C and Chow J-J (2001) Near optimal measuring sequence planning and collision-free path planning with a dynamic programming method. Int J Adv Manuf Technol. 18:29-43

References

171

Mahesh M, Wong YS, Fuh JYH, Loh HT (2004) Benchmarking for comparative evaluation of RP systems and processes. Rapid Prototyp J10/2:123-135 Masood SH and Soo A (2002) A rule based expert system for rapid prototyping system selection. Robot Comput Integr Manuf 18:267-274 Merat FL and Radack GM (1992) Automatic inspection planning within a feature-based cad system. Robot Comput Integr Manuf 9:61-69 Pham DT and Dimov S.S (2001) Rapid manufacturing. Springer, Berlin Heidelberg New York. Qian X and Dutta D (2001) Feature based fabrication in layered manufacturing. Trans ASME J Mech Des 123(3):337-345 Rosochowski A and Matuszak A (2000) Rapid Tooling: the State of The Art. J Mater Process Technol106:191-198 Roy U, Xu Y, Wang L (1994) Development of an intelligent inspection planning system in an object oriented programming environment. Comput Integr Manuf Syst 7:240-246 Tang Y, Loh HT, Fuh JYH, Wong YS, Lee SH (2005) An algorithm for disintegrating large and complex rapid prototyping objects in a CAD environment. Int J Adv Manuf Technol 25:895901 Williams RE, Komaragiri SN, Melton VL, Bishu RR (1996) Investigation of the effect of various build methods on the performance of rapid prototyping (stereolithography). J Mater Proc Technol 61:173-178 http://www.3dsystems.com Accessed 24 April 2009 http://www.dea.it Accessed 24 April 2009 http://www.dimensionprinting.com Accessed 24 April 2009 http://www.eos.info Accessed 24 April 2009 Ziemian CW and Medeiros DJ (1998) Automating probe selection and part setup planning for inspection on a coordinate measuring machine. Int J Comput Integr Manuf. 11:448-460

5 Discussion and Hints for Future Work

This chapter has the same role as the discussion paragraphs inside the description of the three releases of the DGLs in Chap. 2. It is a chapter instead of a paragraph simply because the DGLs-CF has been more widely described in Chap. 3 (fundamentals) and in Chap. 4 (application in the field). This chapter is organised as the aforementioned discussion paragraphs, but it reports the positive outcomes of the DGLs-CF for each category of problem/issues instead of all together; many indications for future research and work follow as usual.

5.1 Conceptual Diagram and Knowledge Organisation The conceptual diagram of the DGLs-CF is simple, clean, and intuitive, and it is the best representation of the knowledge organisation and of the inference process. Moreover, the information loci of the conceptual diagram find a perfect correspondence in the DGLs-CF data structure. As far as possible improvements are concerned, the following issues could be considered: • Aiming at enlarging the application of the DGLs-CF to the whole product lifecycle, some more floors could be considered in the DGLs-CF building; for example, the assembly and the retirement/recycling domains could be a couple of possible candidates. Thanks to all of this, the modules, the procedures, the data structures, etc. of the DGLs-CF would be easily updated and the DGLs-CF could become a wider design for multi-X framework. • Some further work has already been done about knowledge generation. For example, there is an attempt to weigh the importance of each characteristic/feature pair by tagging it as strong or weak in order to classify the dependence of the particular feature from the particular characteristic or vice versa. This could be used to decide which pairs to define the rules for, and the definition order. As far as the rule generation is concerned, the strong pairs are

174

5 Discussion and Hints for Future Work

processed first. All of this should be refined and extended to the other pieces of information in the DGLs-CF data structures. Moreover, this tagging could have some other interesting applications; for example, from a DfR point of view, these pairs could also be tagged highlighting how critical they are regarding the product/process reliability. • Now the DGLs-CF does not take into consideration the consistency of the information collected in the data structures. It could be a good idea to reconsider the consistency monitor already described in the DGLs and refine, implement, and adapt it to the new data structures. Checks can be performed at different levels, from the very basic ones, simply considering the matrix structures, the labels of the pieces of information, etc., to more articulated checks, based, for example, on semantic considerations. • The procedure for discovering the relationships among product features could be refined. The one used in the DGLs-CF is the result of considerations on the releases of the DGLs, where there was a bottom-up approach to the development. A new definition based on the DGLs-CF data structures could improve the procedure by identifying better what must be considered and what is less important. • The package generation procedure could be refined in order to be able to manage more exceptions than now, and in a more effective way.

5.2 Knowledge Description and ISO GPS Adoption The knowledge used by the DGLs-CF is now described in the tables of the modules. The rigorous labelling helps in understanding both the correct positions of the different kinds of information and the father/son relationships between characteristics, features, rules, actions, etc. The ISO GPS concepts helped a lot in enhancing the homogeneity of the information describing the processes (characteristics) and the products (features), thanks to the translation from specific-domain terms and concepts (hi-level formalisation) into same-language items (mid-level formalisation). This resulted in an easier inference process. Regarding possible improvements, there may be the following issues to consider, classified in two sets, one related to features and characteristics and the other to rules, compatibility expressions, and actions.

5.2.1 Product Features and Technological Characteristics • Now, parametric product features and parametric technological characteristics are defined anew each time. The idea would be to generate and manage a database where the users can select the information, ready at disposal, related to several classes of products and technologies. The language uniformity

5.2 Knowledge Description and ISO GPS Adoption

•

•

•

•

175

introduced thanks to the ISO GPS concepts opens the possibility of generating these databases of features and characteristics to be used transversally in different contexts and domains. For sure, these databases would enhance the DGLs-CF adoption and usability even for non-expert users, by avoiding the need to redefine many pieces of information every time. Some topics related to usability are described in a specific paragraph hereafter. The datum systems used to locate the technological characteristics and the product features in an ISO GPS context could be more precisely specified, to make easier the implementation of the related procedures in a semi-automatic way. The procedures to define the product features should be refined in order to pay more attention to the verification domain, for example by introducing dimensional and geometrical tolerances. In this way, inspection procedures to evaluate the conformity of the product could be introduced. These procedures, which have to be coherent with those already defined for the product redesign, could enhance the compatibility with the ISO GPS standards, particularly with the duality principle. ISO GPS standards proved to be really helpful in defining product features of geometrical nature. Nevertheless, the consideration of additional standards, for example those related to the verification of the mechanical characteristics such as impact strength, yield strength, fatigue behaviour, and so on, would be helpful for other types of product features. Moreover, standards are the keyfeature to gain simplicity in interfacing the DGLs-CF with other methodological approaches or systems. For all these reasons some research should be carried out on trying to widen the set of standards considered in the DGLs-CF. In the hypothesis of introducing an Assembly domain floor and a Retirement/recycling domain floor, items deriving from DfA and DfE contexts should be introduced in the features formalisation. Moreover, technological characteristics could be tagged using new indicators; for example, the MET score could highlight their environmental friendliness.

5.2.2 Rules, Compatibility Expressions and Actions • Rules and actions are now defined only to overcome the limits of the technologies. Future work should also consider the definition of rules and actions in order to exploit the particular opportunities offered by the technologies — in a word, the positive rules already cited throughout the book. An example of a positive rule could refer to the possibility of generating free form cooling channels in mould inserts, a possibility offered by the DMLS technology. • The expressions used to evaluate the compatibility between the product (model) and the technologies give only 1 or 0, yes or no, values. As seen in the

176

5 Discussion and Hints for Future Work

releases of the DGLs — a clear example is present in the Appendix — some attempts to switch this binary result to a continue interval — for example in a normalised range [0..1] — have been done and the results appeared very interesting. A continued interval would provide a more complete range of information; compatibility equal to 0 would indicate that some actions surely are needed; the value equal to 1 would mean full compatibility between the product (model) and the available technologies; values included between 0 and 1 would hint at possible activities to be performed to improve compatibility. • The re-introduction of the hints, already seen in the DGLs, has been postponed up to now because they require a consolidated structure, since modifications are necessary in many modules of the DGLs-CF. Now everything appears to be ready to perform some action on this topic.

5.3 Costs Costs have been considered definitely too complex to be managed up to this stage of the DGLs-CF development. However, current architecture and data structures are clear enough and well organised to allow some synergy with experts in cost management. The actions should also be quantitatively considered from this point of view, since this information would be useful in a lot of ways. For example, the selection of the most dynamic action during the Redesign/Reconfiguration Package Generation phase could exploit costs, when two or more actions have the same dynamic coefficient value.

5.4 Implementation/Automatisms The DGLs-CF has proved to be very useful during product redesign and process reconfiguration. However, most operations are user-assisted since there are no automatisms or software implementations. Now the DGLs-CF modules with their data structures, procedures, etc., are ready to be considered for some real implementation. The development of some prototypes in the releases of the DGLs encourages this switch to a computer-based release of the DGLs-CF. This could lead to shorter execution time and increased concurrencies, as will be shortly cited in the following paragraph. Nevertheless, a delicate situation is represented by the seventh module of the DGLs-CF, the Most dynamic action simulation module. As said before, at the moment this is an empty box and every activity required by the procedure for the generation of the redesign/reconfiguration packages is required to be performed by the designers or by the manufacturing/verification experts. The simulation of actions on the product (model) during the design phase or during the manufacturing/verification processes is an intrinsically difficult task because a lot

5.5 DGLs-CF Adoption Process

177

of competencies should be considered. Moreover, introducing automatisms in this case is really difficult because of all the problems coming from interfacing different software packages considered and developed as stand-alone systems. For example, the action “Split the model” would require linking the DGLs-CF with some parametric modelling software package containing enough knowledge to select the best plane surface for splitting, etc. The problems in using the API — Application Programming Interface — of this kind of packages are not straightforward. However, this is only an example; the development of this module could start with something easier, maybe always user-assisted at the beginning, but able to switch to semiautomatic, once developed and well understood.

5.5 DGLs-CF Adoption Process The use of the IDEF0 formalism in describing the DGLs-CF roadmap made the redesign/reconfiguration process more readable, the role of the actors clearer, and highlighted the data flow, the exploitation of the modules, and so on. Now the iterative procedure for the generation of the redesign/reconfiguration packages appears very clear. These are two main issues of interest about possible future work related to the DGLs-CF adoption process: activity timing and DGLs-CF usability. Both of these are described in the following.

5.5.1 Activity Timing and Concurrencies The top of Fig. 5.1 shows a qualitative Gantt diagram of the DGLs-CF adoption. It shows the relative lengths of the activities and the possible concurrencies; each activity bar is paired with the human resources involved. This diagram highlights that now the DGLs-CF shows really few concurrencies (the diagram considers only the first iteration of the redesign/reconfiguration package generation procedure). In some way, this absence of concurrencies is due to the limitations coming from the history of the DGLs, where the roadmap adoption was divided into three phases to map the evolution of the application context and to highlight the role of the different actors involved. But the new elements introduced during the study and development of the DGLs-CF show that the three-phase roadmap can be considered more conceptually than from the adoption point of view, and that the synergy among the actors runs throughout the whole roadmap. All of this said, the limitation of the time sequence among the three phases could be avoided, so that the number of concurrencies would increase and, of course, the time needed to perform the whole process would reduce, as shown in the bottom of Fig. 5.1

178

5 Discussion and Hints for Future Work

(again, only the first iteration of the redesign/reconfiguration package generation procedure is considered here).

Fig. 5.1 Gantt diagram of the current activities in the DGLs-CF (top) and in a hypothetical scenario with higher concurrency (bottom) (the time scale is the same)

5.5 DGLs-CF Adoption Process

179

Moreover, this new diagram reports another difference, a shorter length of the first activities of the DGLs-CF roadmap. This reflects the hypothesis of considering the existence of databases of features, characteristics, rules, actions, etc.; in this case, during the first run of the DGLs-CF adoption in a new application context, all the information related to the product and the technologies would already be at one’s disposal. Then, the DGLs-CF adoption could start by selecting the proper sets of characteristics and features from the database, given the classes of technologies and of the product, and soon, after that, could focus directly on the problems to solve.

5.5.2 DGLs-CF Usability The current stage of the DGLs-CF development suggests starting to think about usability issues or, in other words, clarifying once again the role of the actors involved and understanding how the DGLs-CF could best support their work (Nielsen 1993). Figure 5.2 shows the simple Use Case Diagram used for this purpose, coming from the UML theory (Eriksson and Penkner 2000; Ambler 2005). It contains only the Primary Use Cases; in other words, only the main tasks that the users can accomplish using the DGLs-CF. This time the UML has been preferred to the IDEF0, given the purpose of the diagrams; nevertheless, there is some literature that deals with the complementary use of these two formalisms, such as the work of Cheol-Han et al. (2003) and Noran (2000). For all the actors, this diagram depicts their involvement, or “what” each actor does/requires with/to the DGLs-CF. A second part of the diagram, not yet developed, will describe “how” the DGLs-CF could meet these requirements. This diagram is used as the starting point for all the considerations about usability issues, from the evaluation and testing of the DGLs-CF to the usability design for future releases. Of course, once that development is performed, it will be necessary to update the diagram. For example, now it is important to evaluate the usability of the DGLs-CF from the designer’s point of view, regarding the definition of the DGLs-CF modules and the generation of the working knowledge — the knowledge used by the DGLs-CF procedures to infer the redesign/reconfiguration package content. Once the knowledge databases are developed and the DGLs-CF structure is definitely clear and frozen, this designer’s contribution will disappear and the Use Case Diagram can be updated to focus on other topics. Even if Fig. 5.2 appears very simple, it will become really helpful in the next phase, when the DGLs-CF will actually be tested from the usability point of view. Some evaluation metrics will be developed and this diagram will act as an index for the formalisation of the testing activities and for their results. Table 5.1 depicts the relationship matrix, always representing the information related to the usability issues but using a different layout. Once the metrics are developed, this matrix will allow one to evaluate and compare the use of the DGLs-CF by the different

180

5 Discussion and Hints for Future Work

actors and to generate a to-do list of improving actions. The asterisks in each entry of the matrix weigh the importance of the specific use case for the specific user. For example, product characterisation is a really important use case for the designer (three asterisks) but is of no interest to the developer (no asterisks).

Fig. 5.2 Use Case Diagram for the DGLs-CF Table 5.1 Relationship matrix used for the usability evaluation of the DGLs-CF Users Metrics Designer Manufacturer Inspector Developer Characterise the product *** * * Descr. Generate working knowledge ** *** *** Descr. Define DGLs-CF modules * * * * Descr. Implement DGLs-CF *** Descr. modules Use Share knowledge about cases *** Descr. manufacturing processes Share knowledge about *** Descr. verification processes Redesign the product and *** ** ** Descr. reconfigure the processes

References

181

Summary The DGLs-CF is the result of some years of research. During the development of the releases of the DGLs some questions, problems, issues to consider, etc. arose, as happens during the life of every project. All the important pieces of information have remained under consideration; some problems have been solved, some questions answered. The present chapter has illustrated what remains as open issues and has given some suggestions about new topics that could be investigated, given the robustness of the DGLs-CF from the conceptual, architectural and procedural points of view. Now, the easiest activity is to populate the knowledge base to make the DGLs-CF ready for a wider set of application contexts in which to test it; all the other hints for further research suggested here present different implementation times and difficulties and, once again, highlight the interdisciplinary vision of the DGLs-CF.

References Ambler S.W (2005) The Elements of UML(TM) 2.0 Style. Cambridge University Press, ISBN13: 978-0521616782 Cheol-Han K, Weston RH, Hodgson A, Lee K-H (2003) The complementary use of IDEF and UML modelling approaches. Comput Ind50(1):35-56 Eriksson H-E and Penker M (2000) Business Modeling With UML: Business Patterns at Work. Wiley, ISBN-13: 978-0471295518 Nielsen J (1993) Usability Engineering. Academic Press, Cambridge, MA Noran O (2000) Business Modelling: UML vs. IDEF, [Report / Slides]. School of CIT, Griffith University. Electronic version available: http://www.cit.gu.edu.au/~noran Accessed 24 April 2009

Appendix Generation of Some Meaningful Compatibility Expressions

In the third release of the DGLs, the compatibility evaluation between product features and technological characteristics required some skill because sometimes the generation of the expressions used to calculate the compatibility values was not easy. This is why this appendix presents two examples of expression generation, extracted from the case study — Table 2.6 — and related to the evaluation of compatibility between “Overhangs” and “Supports” — expression E3 — and between “Cavities” and “Probes” — expression E6. Compatibility Between “Overhangs” and “Supports” — E3 RP technologies are layer-based construction methods, heavily affected by the presence of overhangs. Sometimes overhangs do not create problems, but quite often they require supports, depending on the slope of the walls. Sometimes the supports are impossible to generate and/or to remove. For this reason, the compatibility between the product morphology and the technologies must be evaluated. Figure A.1 shows the situation and the limits of the zones used to generate the compatibility expressions.

Fig. A.1 Overhang problem for layer-based methods

184

Generation of Some Meaningful Compatibility Expressions

The management of the two extreme zones was very simple. Overhang generation when α <αSmin should cause an impossible support removal, so this kind of overhangs must be avoided. On the other hand, when α > αSmax, overhangs are generated without supports and they do not present any limitation. The following expressions generated the corresponding compatibility values: E3 = 1 IF α ≥ αSmax E3 = 0 IF α < αSmin As regards the intermediate zone, some considerations must be mentioned. In this zone, supports are needed and their removal is easier as the angle of the overhang is larger. For some reasons, not reported here, the need for supports and, consequently, the compatibility value should follow a curve with a shape as in Fig. A.2.

Fig. A.2 Compatibility curve related to the problem of the overhangs

Hence, to get a similar shape the frequency response equation of a Butterworth filter was used. The result was as follows: E3 = 1 −

1 ⎛ ⎛ α − αSmin ⎞ ⎞ ⎜⎜ ⎟⎟ αSmax − αSmin ⎠ ⎟ ⎝ ⎜ 1+ ⎜ ⎟ 0.5 ⎜ ⎟ ⎝ ⎠

4

This expression completed the set of formulas that could generate the compatibility values covering every situation, together with a couple of ELSE statements.

Generation of Some Meaningful Compatibility Expressions

185

Compatibility Between “Cavities” and “Probes” — E6 Sometimes the cavities of the product are too narrow or too deep for the probe of the measuring machine. In this case the evaluation of compatibility was split into two; first of all, a simple expression controlled the depth; if this test was passed, the other expressions managed the relationship between the area of the cavity and the probe dimensions. Figure A.3 shows this situation.

Fig. A.3 Compatibility management between probes and cavities

The expression corresponding to the first test was as follows, where the length of the probe stylus was compared to the cavities depth; the size of the measuring head is not considered here: E6 = 0 IF dCmax > lPmax The other expressions were conditioned from this one; for this reason they were preceded by the ELSE statement. The following two were straightforward; in fact, if the minimum dimension of the cavities was smaller than the double of the probe dimension, the measurement was impossible, while if the minimum dimension of the cavities was bigger than five times the probe dimension, the measurement was possible and easy: ELSE E6 = 1 IF MIN(xCmin, yCmin) ≥ 5•φPmin ELSE E6 = 0 IF MIN(xCmin, yCmin) ≤ 2•φPmin

186

Generation of Some Meaningful Compatibility Expressions

For intermediate values, the measurement was possible, easier as the cavity was larger. The last expression represented a linear interpolation between the two values 2•φPmin and 5•φPmin, given that in this case a straight line has been estimated to be appropriate. The derived expression was as follows:

ELSE E6 =

MIN(xCmin, yCmin) − 2 ⋅ ΦPmin 3 ⋅ ΦPmin