Finite element methods are a computational technique widely used in structural engineering which have been applied more recently to problems of fluid dynamics. In March this year, The Cosener's House Abingdon - the hostel of the Rutherford and Atlas Laboratories - was the pleasant setting for a symposium on the subject organised by the Atlas Laboratory.
About thirty engineers and scientists attended the three day meeting. Papers were presented on various applications of the method, computer program packages, numerical techniques and computing requirements of finite element calculations. An "Any Questions" session concluded the occasion with Tony Egginton of London Office chairing a panel of SRC Committee members who answered participants questions about SRC policy in the field of engineering computing.
A social programme and symposium dinner was held at which Professor Sam Edwards, FRS, Chairman of the Council, gave a down-to-earth address.
Summing up, Jean Crow, Symposium organiser, discussed increased collaboration between universities and the Atlas Computer Laboratory in engineering computing. Dr Jack Howlett, Director of the Atlas Computer Laboratory concluded with the vote of thanks. All agreed on the success of the occasion.
The Science Research Council (SRC) has deferred a decision on the possible shutdown of the Atlas Computing laboratory at Chilton pending further investigation.
At last month's council meeting the SRC decided that the matter required a more detailed look at the effects of such a move.
The original idea was to merge the Atlas Computing Laboratory with the Daresbury Laboratory in Cheshire. This followed moves to reorganise the structure of high energy physics in Britain. Atlas users were circularised in January asking for their opinion of the move.
Photographic collection of intensity data may be inherently inferior to electronic measurements on a diffractometer but photography does have the advantage that very many reflections are recorded simultaneously. Besides, photography is cheap. The tedium and the relatively poor quality of typical photographic intensity data is mostly due to the difficulties of visual estimation of the spots on the films. Could this not be automated?
Crystallographers using photographic methods will recall that the Science Research Council circulated a questionnaire about three years ago to assess the demand for a central service providing automatic densitometer measurement of diffraction intensities on films, especially Weissenberg films. The response was encouraging. A steering panel for the project was set up; an Optronics automatic scanning microdensitometer was purchased and installed at the Atlas Laboratory at Chilton and, with the appointment of Dr Mike Elder to the project, work started on the development of the necessary computer programs. Programming the densitometry of Weissenberg films proved a difficult task, particularly as no satisfactory programs of any kind had been set up for this kind of photograph. However, success has been achieved. The equipment and its associated software is now able to index reflections and measure their integrated intensities on Weissenberg films, achieving precision definitely better than visual estimates and, of course, much greater speed.
UK structural crystallographers interested in making use of this new service should write to Mike Elder, Atlas Laboratory, Chilton, Didcot, Berks.
Two major computer projects at the Science Research Council are being postponed following Government cutbacks in finance to STC.
The projects are the enhancement of computer systems used for nuclear physics research and the provision for a major computer controlled microfilm recorder and plotter system required for the Atlas Laboratory.
JEAN CROW, a numerical analyst at the Science Research Council's
Atlas Laboratory, near Harwell, has been adopted
prospective Parliamentary Labour candidate for mid-Bedfordshire.
Jean, who used to work at Culham and lives in Wantage, Oxon, made her first Parliamentary bid in the February elections at Guildford, Surrey, where the Conservative member retained the seat.
Mid-Bedfordshire also returned a Conservative member in the same election.
IT'S make your mind up time is a catch-phrase which occurs in one of TV's less intellectually demanding programmes and although that programme may not commend itself to civil servants or even politicians, both would do well to heed the TV catch-phrase.
To be specific, it's make your mind up time for both the National Computing Centre and the Atlas Computer Laboratory at Chilton. The NCC, it may be recalled, is currently trying to find a new director to replace Dr Alec Robinson, and the uncharitable might say that it is also still searching for its true role in the British computing scene.
The Atlas Computer Laboratory is in a rather different situation. It has an established role and in Dr Jack Howlett it has a director who is widely held in high regard, but with the growth of university computing power and the relative decline of the laboratory's computing power in the context of the market for its services, its raison d'etre is not as self-evident as once it was.
It might, with some justification, be thought that to bracket these two organisations together was somewhat arbitrary in view of their widely differing roles. However, there have been murmurings for the greater part of this year, of a possible plan to combine these organisations, together with a number of others in what might be regarded as a UK version of the INRIA research institute in France.
Other possible participants in this grand design would be the Computer Science Division of the National Physical Laboratory and the CAD centre at Cambridge.
While the idea of creating such a centre of excellence in computing, has much to commend it, not least because it could make some of the important work which is being carried out by constituent bodies more visible to the world at large, it must be said that NCC participation in the scheme would raise some doubts.
Whereas, the NPL's Computer Science Division and the CAD Centre are already both primarily involved in research, and the Atlas Laboratory has, by virtue of its strong links with universities, an orientation which could make it the vehicle for research into various aspects of computing, the NCC has no such leaning.
In fact, the NCC has considerable achievements to its credit in the field of education and training, the compilation of surveys and reports, and publishing in general. The area in which its development has been most valuable is thus that of providing advice and services to the computer user.
While it would be dogmatic to suggest that the NCC could play no part in a research-oriented organisation, it seems likely that the particular skills it embodies could be put to the best use in a specifically user-oriented body. That said, it must be admitted that INRIA does include spreading expertise through conferences and training and research into applications, among the missions for which it is responsible.
Looking at the type of work which a British INRIA might undertake, two possibilities immediately spring to mind - networks and computer-aided design. The Computer Sciences Division at NPL has a world reputation for its work on packet switching and the design of computer networks and the possibility of putting its ideas to work on a wider scale could be of real benefit to computing and the nation as a whole.
Earlier this year a report published by the Science Research Council spelled out the importance of networks and recommended that the SRC should support research aimed at securing improvements in the efficiency of the design, implementation. management and use of networks.
The SRC is responsible for the Atlas Laboratory at Chilton and should a research organisation be created, there would, no doubt, be those among the laboratory's 500 or so existing users, most of whom are based at universities, who would be prepared to involve themselves in network research.
For its part, the CAD Centre at Cambridge has built up a considerable body of expertise in computer-aided design in a number of application areas and here, too, the impetus which a new national research body could provide, might well serve not only to widen the scope of the centre's work but also to bring it into the public eye.
Whether or not a research organisation like INRIA is created the question what to do with the NCC remains. Here the best chance of success for the centre's future development would seem to be in a far greater identification with the interests of the computer user. And in this context closer links with the Central Computer Agency which represents Britain's largest computer user - the government - would seem to be essential.
Whatever the outcome, it is high time that some of the official bodies which exist in the computer field become imbued with a greater sense of purpose. The expertise, some of it unrivalled in world terms, is there. If ever there was a time for making up our minds what to do with it, it is now.
