Portal de Periódicos da CAPES

Automatic Computation as an Aid in Aeronautical Engineering

1951; Taylor & Francis; Volume: 25; Issue: 1 Linguagem: Inglês

10.2307/3029572

1930-0980

John Lowe,

Aerospace Engineering and Control Systems

The entire field of application for automatic computing machines is very large. In fact, if we admit machine accounting to this field, it affects all of us in a very personal way every day. However, my remarks are to be restricted to only the application of automatic computation in aeronautical engineering, and further restricted to consider only digital computers. I shall first mention something of the reasons for automatic computation, its value and methiods; discuss briefly its future, and lastly mention some of its difficulties. First of all, what do we mean by automatic computation? As the experts would probably not agree exactly with my definition, let me beg the question and resort to example. The definition can be simplified by excluding analog computers, some of which are not truly computers at all but rather devices which simulate physical problems. By digital computer we mean a divice which solves problems digitally, or, literally, counts on its fingers. However from automatic computers we exclude desk calculators of the Friden or Marchant type. Specifically, then, this discussion is concerned with the use of IB M punched card equipment and such more sophisticated machines as the SWAC, the SEAC, BINAC, IBM's Selective Sequence Electronic Calculator, ENIAC, and certain similar machines presently being developed. These latter machines are spectacular and have been popularized in newspapers and magazines. However, practical digital computing may be said to have started about eight years ago with the widespread use of IB M punched card equipment for this purpose. Even today, by far the largest share of the actual computing is being done with standard IB M equipment. These are essentially accounting machines; indeed, their proper namre is Electric Accounting Machines. In the early days, we had a lot of trouble with such things as persuading IBM to change some machines to print minus signs instead of credit symbols to indicate -negative numbers. The proper handling of algebraic signs presented considerable difficulty, and the first multiplier which provided a reasonably direct solution for this problem was regarded with awe, much as early automobile owners viewed the self-starter. However, today's machines have come a long way and are being improved at an ever increasing rate. There must be some rather compelling reasons for the use of automatic computation. It is not cheap and provides plenty of headaches. Even a small-scale operation may cost $10,000 a month, and on a bigger scale we find that the cost of a single machine is in the order of magnitude