Preface
This book is intended for anyone who is planning the construction, use or management of any kind of engineering materials property information system.
Readers are expected to have a basic familiarity with engineering materials at a level similar to that of a first-year student in general engineering. Only general knowledge of databases, programming and other software is assumed though the greater experience the reader has in these areas, the more useful this book will be. The book is intended to demonstrate to materials experts which types of software expertise they need to acquire and to convince software experts, in terms they can appreciate, how materials databases are perhaps somewhat different from those they are familiar with.
The book addresses the problem of designing databases, decision support aids, expert systems and communication systems that can be integrated with manual and software-supported tasks in design and manufacture, in CAD and CAM. The tasks covered are those of materials selection, materials modelling and materials process simulation: anything that involves access to materials identification or property information.
I would like to thank the following people for making comments and suggestions on early drafts of parts of this book (though they may not have realised it at the time): Roger Bamkin, John Rumble, Jane Vvedensky, Arthur Fairfull, Steve Roberts, Norman Swindells and Mike Ashby, and my innumerable colleagues at Cambridge University Engineering Department and the Engineering Design Research Center at Carnegie Mellon University, but especially to Hugh Shercliff, David Cebon, Chris Turner and Claire Barlow at the former and Paul Fussell, Eswaran Subrahmanian, Peter Piela and Rob Coyne at the latter. Adrian Demaid, Adrian Hopgood and John Zucker of the Open University and David Williams of Loughborough University of Technology have contributed much to my appreciation of knowledge-based tools and their place in distributed, manufacturing environments.
I wish to thank Mary Downs, Diane Rishel and Reid Greene who permitted me to participate in their project developing concept hierarchies of materials information at Alcoa Technical Center. Many thanks also to Bridget Buckley of Butterworth Heinemann for her painstaking and rapid editing.
Of course I would like to thank my wife Margot for putting up with the absence of her husband while this book has been written.
I would like to apologise for the inadequacy of the list of references at the end of the book. In a fast-developing subject such as this, most current work appears first as unpublished working papers, as presentations at workshops and standards committee meetings, by conversation, by electronic mail and on bulletin boards. Unreferenced assertions in the text often come from these sources which have not been listed since few readers will have access to them. Where appropriate, my published academic papers give more detailed references and technical depth.
Throughout the book I have used 'he', 'him' and 'his' to refer to an engineer. I hope very much that as time goes on this description will become less appropriate as a description of the typical engineer.
This is a short book, there is far more to say now and will be more in the future. My best wishes go to those who will write succeeding and, I hope, better volumes on this important subject.
This book was written and typeset using Microsoft Word 4.0 and MacDraw II on an Apple Macintosh SE/30.
After seven years, there have been no more books specifically on this subject, so this remains the only one. Since it has been out of print for years, an HTML edition seemed appropriate. Apart from this preface, the hypertext links and some email address and punctuation corrections, the book here is exactly as it was in 1991.
With the benefit of hindsight, what do I think are the most valuable bits of this book ? Chapter 3 on representing materials information is the core of the book and contained new insights into the different types of materials parameters which do not seem to have been appreciated or properly understood. The "capricious properties" idea hinges on a combination of (a) supervenient, layered theories (levels of description, emergent properties etc.) and (b) the inspiration of classifying a property by how you would design a software model to simulate it. I should have expressed myself better (and probably more often and to more people too). Today this should probably be rewritten using Cohen and Stewart's complicity and simplexity ideas (and I must read Holland's book "Emergence", ISBN 0-201-149435).
Commercial materials databases have developed little and many specifically materials data companies have gone out of business. While mechanical design environments have acquired mechanism and "designerly idiom" libraries, the materials information provision as discussed in this book is still rudimentary. Some design CAD tools, e.g. for carbon fibre laminae and tape-winding, metal casting and polymer injection moulding, now incorporate materials processing models as a precursor to predicting final properties. These processing materials parameters databases are, rightly, seen as more important. The materials selection charts of Ashby's group have grown in depth (a number are available for specific material types, e.g. aluminium alloys, steels etc.) and breadth (materials process selection is now handled) but are still found in universities more than in industry.
To my mind, the most significant theoretical work on representing materials properties in the past seven years has come from Demaid's group at the Open University. Here the inherently open-ended nature of materials information is tackled head-on and not side-stepped.
Since finishing this book, I continued to work in engineering design theory before joining a small Cambridge company. There I helped design the software and managed the software development of a new product data management (PDM) system to support engineering companies design and development needs. During my design research it had become clear that materials selection issues were nearly always going to be a minor and usually-avoidable part of design problems, and this industrial experience confirmed it. It also reinforced the importance of making effective interfaces between design processes and manufacturing operations (procurement, stock control etc., the "ERP" software) and also the ordinary desktop office software systems.
For a year now I have been working in geographic information systems, a much slower-moving industry for the most part, but with interesting information representation problems of its own.
The HTML files produced by Word97 still needed a lot of work which I did manually with the free utilities FrontPad (with Internet Explorer 4 beta), FrontPage Express (with Internet Explorer 4) and Programmers File Editor 7.001 which has excellent macro capabilities. The hyperlinking of the references required a program written in awk, though I actually used the commercial TAWK compiler from Thompson Automation Software.
I have left the page numbers from the original paper book edition in the index and the contents for two reasons: so that reference can be made to chapters of the book by page numbers for those requesting copies from a UK copyright library, and in case I get around to using the implicit information hidden in the page number references for some type of concept-thesaurus.
Errors in the cross-referencing of bibliographic references have been fixed in chapters 4, 6 and 8, and these chapters and the References chapter have been converted to HTML 4.0 (transitional). I used HTML-kit.