Can we have a modelling machine? The choice between Digital and Analogue Computers in British Aeronautical Research
1 like281 views
The document discusses the evolution and significance of digital versus analog computing in British aeronautical research, particularly focusing on modeling technologies and their impact on engineering practices. It highlights the historical context, the challenges faced by engineers in adapting to digital systems, and the persistence of analog computing due to factors like trust and training. Ultimately, it underscores the critical role of these computing methods in enhancing aeronautical design and problem-solving throughout history.
![Post-war British aircraft design
and the application of computing
“There was a lack of design engineer because
so many were required for stress calculations
and it was hoped that better methods of
calculating would improve the position.”
Brig. Hinds, 1952
“The level of skill and initiative expected by the
firms of their [human] computers is not high. It
was suggested that the introduction of one or
two computers of higher calibre, would pay
handsome dividends…. computing activities
would be in the charge of a capable officer, who
would not only plan and lay out the work, but
would keep abreast of development in outside
centres, such as N.P.L. and R.A.E.”
Goodwin and Hollingdale, 1952](https://image.slidesharecdn.com/man-slides-130129111009-phpapp02/85/Can-we-have-a-modelling-machine-The-choice-between-Digital-and-Analogue-Computers-in-British-Aeronautical-Research-8-320.jpg)
![Post-war British aircraft design
and the application of computing
“There was a lack of design engineer because
so many were required for stress calculations
and it was hoped that better methods of
calculating would improve the position.”
Brig. Hinds, 1952
“The level of skill and initiative expected by the
firms of their [human] computers is not high. It
was suggested that the introduction of one or
two computers of higher calibre, would pay
handsome dividends…. computing activities
would be in the charge of a capable officer, who
would not only plan and lay out the work, but
would keep abreast of development in outside
centres, such as N.P.L. and R.A.E.”
Goodwin and Hollingdale, 1952](https://image.slidesharecdn.com/man-slides-130129111009-phpapp02/85/Can-we-have-a-modelling-machine-The-choice-between-Digital-and-Analogue-Computers-in-British-Aeronautical-Research-9-320.jpg)
![Post-war British aircraft design
and the application of computing
“There was a lack of design engineer because
so many were required for stress calculations
and it was hoped that better methods of
calculating would improve the position.”
Brig. Hinds, 1952
“The level of skill and initiative expected by the
firms of their [human] computers is not high. It
was suggested that the introduction of one or
two computers of higher calibre, would pay
handsome dividends…. computing activities
would be in the charge of a capable officer, who
would not only plan and lay out the work, but
would keep abreast of development in outside
centres, such as N.P.L. and R.A.E.”
Goodwin and Hollingdale, 1952](https://image.slidesharecdn.com/man-slides-130129111009-phpapp02/85/Can-we-have-a-modelling-machine-The-choice-between-Digital-and-Analogue-Computers-in-British-Aeronautical-Research-10-320.jpg)
![Thirty year persistence…
Mr Diprose viewed with alarm the implied tendency to
build up large programmes and so have the arithmetical
processes divorced from the physical problem.
[In response] Dr Wilkes said there was less
danger of this happening with automatic
digital computers than with a team of hand
computers. The machine would employ no
short cuts or approximations which the
programmer did not put into his coding and
in general simple repetitive methods would
be used on an automatic digital computer.
The work could be carried out to any
accuracy required by coding the arithmetic
double length or even triple length.
(Minutes of the 3rd meeting of the ARC computation panel.)](https://image.slidesharecdn.com/man-slides-130129111009-phpapp02/85/Can-we-have-a-modelling-machine-The-choice-between-Digital-and-Analogue-Computers-in-British-Aeronautical-Research-22-320.jpg)
Can we have a modelling machine? The choice between Digital and Analogue Computers in British Aeronautical Research
- 1. Can we have a modelling machine? The choice between digital and analog computer in British aeronautical research Charles Care “Computer in use” Manchester July 2006
- 2. Aeronautical technologies • Aeronautics – a field of high technology • Interesting to look at the technologies supporting high technologies… • In particular, the technology that supports engineering design work
- 3. Computer as a modelling machine • Computer as an information processor is an well established ‘history’ • This research is investigating the history of the computer as a modelling technology • The focus is on scientists and engineers • The computer (and associated technologies) assists an engineer’s problem-formulating, problem-solving activity.
- 4. Computer as a modelling machine “A recent conference... attempted to map out the history of software, considering it as science, engineering, labour process, reliable artefact and industry... What the conference missed was software as model, software as experience, software as medium of thought and action… It did not consider the question of how we have put the world into computers.” Mahoney, M. “The histories of computing(s)”, Interdisciplinary Science Reviews, 30(2), 2005.
- 5. Aeronautics, aerodynamics and context of modelling technologies • Trend towards general purpose and the abstract – Introduction of formalisms (aerodynamics) – Also experimental formalisms as experiment shifted from being field-based to become a laboratory activity. • The wind tunnel (abstraction of full-scale flow) • The electrolytic tank (further abstraction, flow represented with electrical fields) • Each abstraction needs to become trusted by engineering culture…engineering trust
- 6. Analog taxonomies • Direct / indirect – Used by contemporary writers on analog computing (1950 – 1970) – Shaped by usage – Direct • Direct mapping (analogy) between systems • MIT Network Analyzer – Indirect • Relationship between two systems mediated by mathematics… e.g. differential equations • MIT Differential Analyzer
- 7. Aeronautical modelling: the electrolytic tank G. I. Taylor and C. F. Sharman, A Mechanical Method for Solving Problems of Flow in Compressible Fluids. Proceedings of the Royal Society of London,1928.
- 8. Post-war British aircraft design and the application of computing “There was a lack of design engineer because so many were required for stress calculations and it was hoped that better methods of calculating would improve the position.” Brig. Hinds, 1952 “The level of skill and initiative expected by the firms of their [human] computers is not high. It was suggested that the introduction of one or two computers of higher calibre, would pay handsome dividends…. computing activities would be in the charge of a capable officer, who would not only plan and lay out the work, but would keep abreast of development in outside centres, such as N.P.L. and R.A.E.” Goodwin and Hollingdale, 1952
- 9. Post-war British aircraft design and the application of computing “There was a lack of design engineer because so many were required for stress calculations and it was hoped that better methods of calculating would improve the position.” Brig. Hinds, 1952 “The level of skill and initiative expected by the firms of their [human] computers is not high. It was suggested that the introduction of one or two computers of higher calibre, would pay handsome dividends…. computing activities would be in the charge of a capable officer, who would not only plan and lay out the work, but would keep abreast of development in outside centres, such as N.P.L. and R.A.E.” Goodwin and Hollingdale, 1952
- 10. Post-war British aircraft design and the application of computing “There was a lack of design engineer because so many were required for stress calculations and it was hoped that better methods of calculating would improve the position.” Brig. Hinds, 1952 “The level of skill and initiative expected by the firms of their [human] computers is not high. It was suggested that the introduction of one or two computers of higher calibre, would pay handsome dividends…. computing activities would be in the charge of a capable officer, who would not only plan and lay out the work, but would keep abreast of development in outside centres, such as N.P.L. and R.A.E.” Goodwin and Hollingdale, 1952
- 11. Post-war British aircraft design and the application of computing • Aeronautical Research Council (ARC) – Established 1909, disbanded 1970 – “an ingenious mechanism… for peer review of secret work” (Nahum, 2002) – Establishment of a Computation Panel (which later grew into a sub-committee) – around 10 members including: • S. C. Redshaw (Prof. Civil engineering – Birmingham) • M. Wilkes (Cambridge), F C Williams (Manchester) • G. H. Hinds (Director of Weapons Research, MoS) • E. T. Goodwin (NPL) and S. H. Hollingdale (RAE) “…only a few people in the aircraft industry realised the need for efficiency in computation and many were content to take months over work that could, and had, been done in a few days.”
- 12. Post-war British aircraft design and the application of computing • Aeronautical Research Council (ARC) – Established 1909, disbanded 1970 – “an ingenious mechanism… for peer review of secret work” (Nahum, 2002) – Establishment of a Computation Panel (which later grew into a sub-committee) – around 10 members including: • S. C. Redshaw (Prof. Civil engineering – Birmingham) • M. Wilkes (Cambridge), F C Williams (Manchester) • G. H. Hinds (Director of Weapons Research, MoS) • E. T. Goodwin (NPL) and S. H. Hollingdale (RAE) “…only a few people in the aircraft industry realised the need for efficiency in computation and many were content to take months over work that could, and had, been done in a few days.”
- 13. Deciding between analog and digital: the case of flutter • One of the major calculations that aircraft designers had to make • By 1955, it had “become an increasingly serious problem due to the combination of higher aircraft speeds and thinner wings and tail surfaces” “…calculations have to cover more degrees of freedom, and the effects of variation in the aerodynamic and structural parameters need to be investigated to a greater extent… and high speed computational aids have become a necessary adjunct to flutter problems.” (Templeton, 1955)
- 14. Deciding between analog and digital: the case of flutter • One of the major calculations that aircraft designers had to make • By 1955, it had “become an increasingly serious problem due to the combination of higher aircraft speeds and thinner wings and tail surfaces” “…calculations have to cover more degrees of freedom, and the effects of variation in the aerodynamic and structural parameters need to be investigated to a greater extent… and high speed computational aids have become a necessary adjunct to flutter prediction.” (Templeton, 1955)
- 15. Deciding between analog and digital: the case of flutter • Flutter simulators: analog devices to solve the ‘flutter equations’ – FS I, a prototype machine with 2 degrees of freedom – FS II, 6 degrees of freedom – FS III, 12 degrees of freedom The RAE FS I
- 16. Deciding between analog and digital: the case of flutter The RAE FS II Also other special purpose analog devices to mechanise the other stages of the flutter problem (NOMAD, INCA, MAYA)
- 17. Deciding between analog and digital: the case of flutter “...an analogue flutter simulator would be preferred by the people working on flutter because an all-purpose machine could be used for other computations and therefore would not be for exclusive use.” Diprose, 1952 “… the popularity of analogue machines was due to the fact that firms already employed staff trained in electronics and servo-mechanisms who could be used to service such machines. Digital machines required more specialised servicing teams and some training schemes would be required to provide the necessary staff. Prof. Pugsey, 1952
- 18. Deciding between analog and digital: the case of flutter “...an analogue flutter simulator would be preferred by the people working on flutter because an all-purpose machine could be used for other computations and therefore would not be for exclusive use.” Diprose, 1952 “… the popularity of analogue machines was due to the fact that firms already employed staff trained in electronics and servo-mechanisms who could be used to service such machines. Digital machines required more specialised servicing teams and some training schemes would be required to provide the necessary staff. Prof. Pugsey, 1952
- 19. Thirty year persistence, four shortcomings of digitalisation • Why did analog persist? • Four major problems, three relating to programming and one to engineering trust. 1. Engineers were not trained to program – either they needed to learn how, or alternatively out-source to a programmer. 2. Computing was not necessarily separable from the design process. 3. Issue over whether the design process (and the engineers) should be adapted to fit the technology. Should computing be close-shop or open-shop? 1. Great engineering tradition in communicating knowledge through physical analogies
- 20. Thirty year persistence, four shortcomings of digitalisation • Why did analog persist? • Four major problems, three relating to programming and one to engineering trust. 1. Engineers were not trained to program – either they needed to learn how, or alternatively out-source to a programmer. 2. Managing computations was not necessarily separable from the design process. 3. Issue over whether the design process (and the engineers) should be adapted to fit the technology. Should computing be close-shop or open-shop? 1. Great engineering tradition in communicating knowledge through physical analogies
- 21. Thirty year persistence, four shortcomings of digitalisation • Why did analog persist? • Four major problems, three relating to programming and one to engineering trust. 1. Engineers were not trained to program – either they needed to learn how, or alternatively out-source to a programmer. 2. Managing computations was not necessarily separable from the design process. 3. Issue over whether the design process (and the engineers) should be adapted to fit the technology. Should computing be close-shop or open-shop? 1. Great engineering tradition in communicating knowledge through physical analogies
- 22. Thirty year persistence… Mr Diprose viewed with alarm the implied tendency to build up large programmes and so have the arithmetical processes divorced from the physical problem. [In response] Dr Wilkes said there was less danger of this happening with automatic digital computers than with a team of hand computers. The machine would employ no short cuts or approximations which the programmer did not put into his coding and in general simple repetitive methods would be used on an automatic digital computer. The work could be carried out to any accuracy required by coding the arithmetic double length or even triple length. (Minutes of the 3rd meeting of the ARC computation panel.)
- 23. Conclusion • No direct barrier to digital computing • Engineering practice adapted to fit the technology. • Analog computing was part of their trustworthiness and professional credibility. • 30 year persistence corresponds to length of a working life
- 24. Conclusion • No direct barrier to digital computing • Engineering practice adapted to fit the technology. • Analog computing was part of their trustworthiness and professional credibility. • 30 year persistence corresponds to length of a working life Why research this, where’s it heading? • Trying out the modelling machine theme • Exploring the shaping of professional status and knowledge – engineering trust • Developing a ‘what works’, or ‘if it ain’t broke’, historiography to study analog persistence.
- 25. Conclusion • No direct barrier to digital computing • Engineering practice adapted to fit the technology. • Analog computing was part of their trustworthiness and professional credibility. • 30 year persistence corresponds to length of a working life Why research this, where’s it heading? • Trying out the modelling machine theme • Exploring the shaping of professional status and knowledge – engineering trust • Developing a ‘what works’, or ‘if it ain’t broke’, historiography to study analog persistence.
- 26. Selected References • Allan G. Bromley. Analog computing devices. In William Aspray, editor, Computing before Computers, pages 159–199. Iowa State University Press, 1990. • Robert Bud and Philip Gummett, editors. Cold War, Hot Science: Applied Research in Britain's Defence Laboratories 1945–1990. Science Museum, London, 2 edition, 2002. • Martin Campbell-Kelly and William Aspray. Computer, A History of the Information Machine. BasicBooks, New York, 1996. • David Edgerton. Warfare state : Britain, 1920-1970. Cambridge University Press, 2006. • S. H. Hollingdale and K. V. Diprose. The role of analogue computing in the aircraft industry. Typeset report of the Computation Panel of the ARC. Dated 7 January. National Archives: DSIR 23/21372, 1953. • S. H. Hollingdale and G. C. Toothill. Electronic Computers. Penguin Books, 1970. 2nd edition. • Michael S. Mahoney. The histories of computing(s). Interdisciplinary Science Reviews, 30(2), 2005. • Andrew Nahum. The royal aeronautical establishment from 1945 to concorde. In Bud and Gummett (2002).
- 27. Selected References (cont.) • Edward Pyatt. The National Physical Laboratory : a history. Adam Hilger Ltd., Bristol, 1982. • James S. Small. The Analogue Alternative : The Electric Analogue Computer in Britain and the USA, 1930–1975. Routledge, London, 2001. • H. Templeton. Computational aids or the solution of flutter problems. National Archives: DSIR 23/23788, 1955. • Stephen Robert Twigge. The early development of guided weapons in the United Kingdom : • 1940-1960. Harwood Academic, 1993. • Stephen Robert Twigge. Ground-based air defence and abm systems. In Bud and Gummett (2002). • Aristotle Tympas. From digital to analog and back: The ideology of intelligent machines in the history of the electrical analyser. IEEE Annals of the History of Computing, 18(4):42–48, 1996. • R. Keil-Slawik U. Hashagen and A. Norberg, editors. History of Computing: Software Issues, Berlin, 2002. Springer. • Walter G. Vincenti. What Engineers Know and How They Know it : analytical studies from aeronautical history. The John Hopkins University Press, Baltimore, 1990.
- 28. Major sources for understanding military science research
- 29. Previous work on analog computing history