BOOK REVIEWS 297

Computer-integrated Manufacturing Technology and in the book, again a point that must detract from its
Systems, U. Rembold, C. Blume and R. Dillmann, usefulness.
Marcel Dekker, 1985. Number of pages: 790. Price: All in all this book cannot be recommended to engineers
$79.95 (North America), $95.50 (elsewhere). involved with electronics testing. An up-to-date compre-
hensive engineering book on the subject is still very much
This is not a book about quality or reliability, but it is needed.
nevertheless a valuable reference source for Q&R people
in the modern world of integrated development and manu- FINNJENSEN
facturing. It describes computer applications in design,
machining, assembly and measurement. It also covers
system architectures, software and simulation techniques.
The authors have provided very detailed coverage of
current systems and methods, together with detailed 1. Advanced Strength and Applied Elasticity, A.C.
theoretical and analytical support. This comprehensive Ugural and S.K. Fenster, Second SI Edition, Elsevier
and detailed treatment of the subject stems from the fact Science Publishing Company Inc., 1987. Number of
that the book is based on the course material for the CAD/ pages: 471. Price: f35.00 (hardback only).
CAM course at the University of Karlsruhe, Germany.
The chapter entitled quality control is mainly a descrip- 2. Elasticity in Engineering Mechanics, A.P. Boresi and
tion of modern measurement systems. It does not describe K.P. Chong, Elsevier Publishing Company Inc., 1987.
process control methods such as SPC, but it does cover Number of pages: 645. Price: f38.00 (hardback only).
data systems and integration. The references at the end
of this chapter do not include any of the standard texts Although these are both good books within their own
on modem quality control and management. terms of reference-they are new editions of established
Although the authors have produced a very comprehen- books, with an additional author in the second case-I do
sive and up-to-date text in most respects, there is no not think they will appeal in many instances to readers of
description of just-in-time (kan ban) systems. Also, the this Journal. The reason is that a deterministic approach
specific problems of electronics manufacture are not has been adopted in both books. No reference is made to
covered. However, these omissions detract very little from statistically distributed quantities in the first book, though
its value as a comprehensive and interesting reference on variation is taken into account by conventional factor of
the field it covers. safety methods. The authors of the second book are aware
of statistical methods, but only mentioned them, with
P. D. T. O’CONNOR limited references confined to structures, in the first chap-
ter. The subsequent treatment if wholly deterministic. It
can be argued, of course, that it is necessary to understand
the standard theory of elasticity before proceeding to
Strategies for Electronics Test, Craig Pynn, McGraw- behaviour with distributed quantities: for example the
Hill, 1986. Number of pages: 174. Price: DM 49.90. calculation of the distributed stress in, say, a beam result-
ing from a distributed load follows the same principles as
This could have been a good book. Unfortunately the the calculation of the deterministic stress resulting from a
reading of the book was a disappointment. deterministic load. The same is true of nearly all mechan-
The audience for the book has not been defined in the ical items. However the treatment in both books gives the
introduction, and I fail to see to whom the book can be impression that definite solutions to problems are possible
useful. The style of writing is easy-going with very little rather than that values have a probability of occumng. I
mathematics, indicating, indirectly, that the book might do think it a pity that students in particular should be
be intended for technically orientated persons without misled as to the true nature of things in their formative
formal engineering education. However, I do not believe years. The chapter on the mechanical behaviour of
that the book can engage this audience, except maybe in materials in the first book is open to much criticism on
some of the well-written introductory parts. The practising this score. The more precisely defined scope of the second
engineer, on the other hand, will find the book too simplis- book renders it less vulnerable to criticism in this respect,
tic to be of any real use. The subject is complex, and an but it is difficult to see how variation can be ignored.
engineering use of mathematics can help bring order out These books are, of course, not alone by any means in
of chaos when applied judiciously. Unfortunately, the the approach they have adopted, and within the terms of
little mathematical modelling that is applied in the book reference the authors have created for themselves, and
is sometimes more misleading than helpful. within which the subject is most often treated, they are
The strongest point of the book is probably its general good books. The first offers a more general and compre-
discussion of what might go wrong in the assembly of hensive approach suitable for degree course students who
electronic systems, especially printed circuit boards. Part have already completed a first year course on materials.
One of the book, ‘Defining the test strategy’ gives a good It does contain an introduction to numerical methods,
overview of problem areas in electronic production, and including finite element techniques, which is essential for
introduces the concept of a fault spectrum to describe practical application of the subject. The second book is
the various classes of defects and their weighting. Test more advanced (in spite of the word appearing in the title
efficiency, the influence of production yield, production of the first and not the second), concentrating more on
rate and production mix are all factors that are dealt with the mathematical theory of elasticity necessary for a fuller
in this first part, which concludes with developing a test understanding of the subject. It would be applicable in
strategy and discussing problems of economics. the final year of appropriate degree courses, but I think
The second part of the book, ‘Implementing the test could also find use as a reference .book for practising
strategy’ is interesting, but again has too little substance engineers. Unfortunately it does not include any treatment
to be of real use. There is however a good discourse on of numerical methods, which are essential for such people.
the uses of simulation versus emulation, still a subject The authors state they are going to publish a second book
confusing to many engineers. The chapter on environmen- covering this topic. If this reaches the same standard as
tal stress screening has several fallacies. the present volume the two together could be a valuable
The book contains a good glossary of terms in elec- work of reference. It should be made clear, however, that
tronics test. There are no references and no bibliography it is a book on elasticity in the strict meaning of the word
298 REPORT REVIEW

and does not include any material on related Qualig Assurance in Process Plant Manufacture, J . H.
topics-fracture mechanics for example. Rogerson, Elsevier Applied Science Publishers, 1986.
In conclusion I can recommend both these books to Number of pages: 159. Price: f27.00
readers who are interested in the limited approach, i.e.
the conventionally accepted approach, to the subject mat-
ter they cover, but would warn that they have no special For anyone involved in management and engineering
significance in connection with either quality or reliability. related to process plant manufacture, this is a very timely
and useful introduction to quality assurance related to the
A.D.S. CARTER industry. The author uses the term ‘process plant’ to
embrace all plant installations in industries such as power
generation, chemical, food manufacture, etc., so the book
should have wide appeal, particularly as it is designed to
Testing in Software Development, M. A. Ould and C. provide an overview and critique of the major national
Unwin (eds) Cambridge University Press, 1986. Num- and international standards relevant to the field.
ber of pages: 124. Price: f9.95/$16.95. The book covers economic and safety arguments for
quality, and the message presented is relevant and up-to-
This excellent little book has been prepared on behalf of date, including mention of recent product liability legis-
the British Computer Society, and is published in its series lation. The major relevant standards in the U.K., Europe,
of ‘Monographs in Informatics’. The editors have put Canada, and international standards are described, but
together a text which is very well constructed and pre-
the only U.S. standards mentioned are the ASME Codes.
sented. They have also combined the material so that the The book briefly covers quality costs, rightly pointing
separate contributions are merged to provide a seamless, out that these can be minimized only by the pursuit of
consistent and concise introduction to the subject of excellence, and not by any less stringent approach.
software testing. Despite the clear, common-sense treatment of the sub-
The structure adopted is to present successively the ject, the book does omit some relevant topics. There is
views and responsibilities of the manager, the user, the
no mention of reliability engineering methods during
designer and the programmer. For each of these the book design and development, or reference to reliability stan-
provides a wealth of practical guidelines and checklists on
dards. The particular quality assurance requirements of
testing of the requirement, the system specification, the
the food and pharmaceutical processing industries are not
software product itself and the test plan. Each section is described. Also, the references given at the end of each
written in a way that respects the knowledge of the reader, chapter are insufficient, and in several cases incomplete.
and yet the whole book is easily understandable by man- A reasonably comprehensive list of relevant standards,
agers, students and workers whose work involves the with their coverage, application and sources would have
development of new software. The book is completely been a helpful addition. It is hard to imagine a process
general, being equally relevant to data processing or
plant being developed today without a major software
engineering and scientific applications. component, yet the book makes no mention of software
I strongly recommend this book as a necessary primer quality standards and methods.
and handy reference for anyone involved in system Bearing in mind the limitations mentioned, the book is
software development. Applying the principles and
a useful, if somewhat incomplete, introduction to quality
methods described is as close as one can get to guarantee- assurance in process industries.
ing a trouble-free development programme.

P. D. T. O’CONNOR P. D. T. O’CONNOR

Report Review
Reliability Indicator Methods - A Feasibility Study, generate energy concentrations, such as hot spots. In
F. Jensen and J. Moltoft, for the Danish Council of other cases energy input might be altered, for example in
Technology, 1987. Number of pages: 115. the generation of harmonic noise. The authors provide a
very interesting survey of the development and use of
‘Reliability indicators’ are methods for determining reliability indicators over a wide field of engineering. In
which components (electronic, electrical, mechanical) are addition to the examples mentioned, methods described
more likely than others to fail within the expected operat- include discriminant function analysis applied to oil seals
ing life. For example, noise measurements on resistors or and electronic components, marginal voltage testing of
acoustic tests on railway wheels can indicate the existence digital integrated circuits, and parameter drift methods.
of flaws. Such methods are not destructive, and they do The report is a very useful and interesting survey of a
not subtract from the expected operating life. Therefore branch, or rather of several branches, of practical
they contrast with the usually better-known technique of reliability engineering. It points out the need for more
‘burn-in’, particularly of electronic components, in which research into these applications, which can be more effec-
the weak components are caused to fail by the imposition tive and much less damaging than ‘brute force’ techniques
of long duration high stress levels. such as burn-in.
The report provides a very clear description of the The report is available from the Engineering Academy
underlying philosophy. If in a population of manufactured of Denmark, Building 451, DK-2800 Lyngby, Denmark.
items there is a subpopulation of ‘freaks’, which contain Attn: Hanne Nicolaisen. The price is DKK200 plus post-
flaws which can reduce life, and if the existence of these age and packaging.
flaws can somehow be detected, then this knowledge can
be used to improve reliability. In many cases the flaws P. D. T. O’CONNOR