Modern engineering demands a high degree of precision, based on accurate measurements, accurate standards of design and dimensions, and an accurate system of gauging. For example, the Imperial Standard Yard is the product of years of experiment
TESTING STANDARD GAUGES for accuracy by Johansson blocks. Invented by a Swedish engineer, Carl Edward Johansson, these metal blocks are within two parts in a million of their nominal size. So accurate are the plane surfaces that two blocks pressed together will adhere, to form a new component having a length equal exactly to the sum of the lengths of each unit. There are three blocks in the gauge illustrated, one of 2 in, one of 0.25 in, and another still smaller.
THE cornerstone of all modern engineering and of all modern manufacture is accuracy — accuracy of calculation when a new structure is being planned, accuracy of workmanship in transferring it from paper to reality. Above all, accuracy is needed in making all component parts.
Nowadays precision in work of this kind is so much of a commonplace that it tends to be more or less taken for granted. Many huge structures, such as buildings and ships, are built up from components which may have been manufactured at points hundreds of miles apart, their makers working direct from the blue prints supplied to them. Yet when the various parts have been transported to the place of assembly they go together perfectly.
In the days of hand labour there was no such thing as standardization. Every piece produced by every workman had its individual idiosyncrasies, and if it had to fit another piece the fitting had to be effected by a laborious process of trial and error — of patient scraping, filing and general adjustment.
Nor, in those days, was any great accuracy either expected or regarded as attainable. From the letters of James Watt to his partner Matthew Boulton we learn of his delight at securing a cylinder for one of his engines which had been bored to an internal diameter of 18 in, and in which “ ... at the worst place, the long diameter exceeded the short by only three-eighths of an inch”.
On another occasion we find Watt testing, and passing as highly satisfactory, a cylinder and piston which fitted so well that he could not at any point round the piston’s upper rim slip a half-crown between it and the cylinder. It is small wonder that in such days an engine, working at little above atmospheric pressure, would be kept steamtight by packing the outside of the piston with chewed paper, or old felt hats, and grease. Accuracy in modern engineering practice is based, first of all, upon accurate measurements — for which exceedingly accurate standards of length, weight and the like are required. Secondly, accuracy is based upon agreed standards of design and dimensions in such parts as screw threads; and, lastly, upon a system of gauging whereby all articles manufactured to such standards can be rapidly and accurately tested for conformity.
The system of weights and measures in general use in Great Britain seems, at first sight, to have been devised chiefly for the embarrassment of foreigners, and even of residents in different parts of the British Isles. Still, its standard of length, the yard, has an old and honourable pedigree, having survived virtually unaltered from the Middle Ages and earlier. The present Imperial Standard Yard, for example, differs by only a few hundredths of an inch from the oldest authentic standard surviving in Great
Britain. This is Henry VIII’s yard, a brass rod of octagonal section which, with a later one made during Queen Elizabeth’s reign, is preserved in the Science Museum, South Kensington.
Until 1824 Great Britain had no other official standards of length; yet the value of the standards extant was virtually negligible. For example, in 1834 Francis Baily, F.R.S., when reporting on the new standard-yard scale built to his design for the Royal Astronomical Society, pointed out that the standard yard of Queen Elizabeth had, at some unknown date, been broken into two pieces and clumsily repaired, the parts being united by a dovetail joint almost as loose as that of an ordinary pair of tongs. Yet this utter destruction of the rod’s character as a standard had in no way prevented the widespread circulation of copies of it throughout Europe and America, accompanied by a stamped parchment certificate attesting that they were true copies of the standard yard.
Faulty Act of Parliament
In 1824 a well-meant but ill-considered Act of Parliament enacted that in future the Standard Yard was to be “ . . . that straight brass rod now in the custody of the Clerk of the House of Commons, whereon the words and figures ‘STANDARD YARD, 1760’ are engraved”, and no other. It also laid down that if this standard were ever to be destroyed, a successor to it was to be obtained by comparison with the effective length of a pendulum swinging exact seconds, in a vacuum, in the latitude of London. The length of such a pendulum was taken by the Act to be 39.1393 in.
This Act contained at least two glaring blunders. First, the bar specified had been made in that year as a copy of a copy, although the original still existed.
The Act’s second blunder was in specifying that the standard yard, if lost, should be recovered by means of pendulum experiments. In theory, there was nothing to be said against some plan of the kind. But in practice most attempts to obtain a natural standard of length have been failures. The metre was intended by the devisers of the metric system in 1792 to be exactly one ten-millionth part of the distance, on the Earth’s surface, from the North Pole to the Equator. It is now known, however, that the length of such a quadrant of the Earth’s circumference varies slightly according to the meridian along which it is measured. No natural unit of adequate accuracy was known until 1892-93, when the American physicist Albert A. Michelson succeeded in expressing the wavelength of light in terms of the standard metre.
Within ten years, the contingency provided for by the Act of 1824 came about. The Standard Yard was accidentally destroyed in the fire which consumed the Houses of Parliament (October 16, 1834). The bar was recovered from among the debris but it was badly distorted. A new Standard Yard had to be determined. It was ultimately decided that the length of this should be obtained by comparison with various existing scales which at different times had been compared with the “STANDARD YARD, 1760”. The work of doing this was entrusted to Baily — but he died (in 1845) before he had made much progress.
His successor, the Rev. Richard Sheepshanks, was a man of great ability and enormous energy. He erected, in the cellars of the Royal Astronomical Society (then accommodated in Somerset House) a huge apparatus of comparator-microscopes, with elaborate arrangements for maintaining a constant temperature. Here he plodded on for ten years, taking and recording some 200,000 micrometrical measurements — until, just when the end of his labours seemed in sight, he was struck down by apoplexy (July 29, 1855) and died a few days later.
THE STANDARD CUBIC FOOT is represented by a special bottle of that capacity in the Standards Department of the Board of Trade. One cubic foot equals 1,728 cubic inches, or 0-028317 cubic metre.
Sheepshanks had realized, and it was generally admitted, that any standard yard which his work produced was bound to be a compromise. No one could ever be certain that it agreed exactly with the 1824 Parliamentary Standard. What was really needed was not that the new standard yard should agree to the millionth part of an inch with the old one, but that, once fixed, it should be invariable, and that a number of exact copies of it should speedily be made and distributed, so that the new standard should never again be in any danger of being lost.
Sheepshanks’s work was completed and published by his friend George B. Airy, the Astronomer Royal. Summarized, what he had accomplished was this. He had had cast a number of bars of “Baily’s metal” (16 parts copper, 2½ parts tin, 1 part zinc). Five selected bars were then compared, accurately, with the existing scales, one of them (No. 1) being found to be as nearly as possible a standard yard in length at the prescribed legal temperature of 62° Fahr. This was chosen to be the new Imperial Standard Yard, and was allotted this title by Act of Parliament (1855). Originally kept in the Exchequer Office, it was afterwards transferred to the Standards Department of the Board of Trade, where it now is. It constitutes the standard legal measure of length throughout the British Empire.
It is a rectangular bronze bar 38 in. long, with a 1-in. square section. One inch from either end is a ½ -in. Hole sunk to the centre line of the bar, and in the centre of each hole is a gold plug, set in flush with the bottom. On the face of each plug are inscribed three transverse lines about 0.01 in. apart, and two longitudinal lines about 0.02 in. apart, the whole resembling a three-barred gridiron in miniature. The lines are not more than 0.0005 in. wide. The legal yard is defined as being, when the bar is at a temperature of 62° Fahr., the interval between those portions of the central transverse lines included between the two longitudinal lines.
Walled Up for 70 Years
The other four bars were used to form copies of the new standard. The exceedingly small amounts by which they differed in length from it were determined, micrometrically, as accurately as possible. Then, the coefficient of expansion of the metal being known, the temperatures were computed at which each bar would be exactly equal in length to the standard, supposing this to be at 62° Fahr. All four of these determinations ranged within ½° of 62°. Each bar was then engraved with its correct respective temperature, and the four bars became “Parliamentary Copies of the Standard Yard”. One was deposited at the Royal Mint, one at the Royal Observatory, Greenwich, one at the Royal Society’s rooms, and one in the House of Commons. This last bar was walled up and, apart from a casual inspection in 1892, remained undisturbed for seventy years, after which it was taken out, compared with the other copies (it was in satisfactory agreement with them) and replaced.
STANDARD POUND WEIGHT and end of Imperial Standard Yard in the Standards Department of the Board of Trade. The standards rest in a special coffer. The pound is made of platinum and the yard of Baily’s metal (16 parts copper, parts tin and 1 part zinc). The yard is a bar, 1 in. square, with a gold plug (photograph on right) set at either end. Lines marked on the gold plugs mark the limits of the yard measurement. The gold plugs are protected by metal caps.
The corresponding original standard of the metric system used in France and other countries, the Standard Metre, has been safely preserved since its manufacture in 1792. This standard bar, preserved in Paris, and known as the Metre des Archives, is a bar of platinum 25mm by 4mm in section, its length (as measured by the distance between the flattened faces of its ends) being 1 metre exactly at 0° C. In 1877, however, this standard was supplanted, for practical purposes, by a new one known as the “International Prototype Metre”, of which many copies are extant. This is made of a platino-iridium alloy (90% platinum, 10% iridium) and resembles, in section, a flattened letter H, this form combining rigidity and lightness to a high degree. In the same way as the Imperial Standard Yard, the Metre des Archives uses line-measurement as against end-measurement; that is to say, it indicates the standard metre as the distance between two defined points on its surface, and not by the interval between its ends.
As the result of comparing one of the International Prototype Metre copies with the Imperial Standard Yard, an Order in Council of May 19, 1898, defined the length of the metre, for all legal purposes in the British Empire, as 39.370113 in. A redetermination made, on an improved plan, by the National Physical Laboratory in 1922 gave 39.370147 in. This afforded satisfactory proof that the two standards had undergone no sensible alteration of length.
For most practical purposes, copies of length-standards are now generally made of nickel — or, sometimes, of invar. This alloy of nickel and steel, invented by Dr. Charles E. Guillaume, for many years Director of the International Weights and Measures Bureau, Sevres, near Paris, is unique in having a virtually negligible coefficient of expansion. As, in addition, it takes a high polish, and can readily be graduated, it might seem at first sight to be the ideal material for all length-standards; but, unfortunately, it is comparatively soft and, what is far more serious, its dimensions are not altogether stable. Its microscopic coefficient of expansion depends upon an internal molecular stress which, apparently, causes its length to vary in the course of years. For example, a metre bar of invar kept under observation at the National Physical Laboratory for thirty years has exhibited a slow but steady growth during the whole period.
One-Millionth of an Inch
It would have greatly surprised Sheepshanks in the early years of his work on the standard yard if someone had told him that a toolmaker in Manchester had already produced, and was using, apparatus by which he could measure accurately and rapidly to one-millionth part of an inch, and that this man’s workmen, using simple gauges devised by their master, were in the habit of working to hundredths, and even to thousandths of an inch. The man was Joseph Whitworth.
The basis of all accurate machine-tool construction is the formation of a metal surface which is a true plane. Before Whitworth’s time this was done, after a fashion, by planing two surfaces as well as possible and then grinding them together. This produced a smooth-looking surface, but one far from true. The process merely confused the errors of the two surfaces. If one were originally flat and the other concave, grinding made the latter less concave and the other slightly convex.
Whitworth was so struck by the defects of this plan that one day, while engaged in grinding-in two plates, he remarked to his benchmate, a stolid Yorkshireman named John Hampson, “If these plates were true, one ought to lift the other”. In a short time Whitworth was able to show Hampson a pair of plates, made in his own time and by his own methods, which were true planes, and fitted so accurately that if pressed together so as to expel the film of air between them, the lower one adhered strongly to the upper and could be lifted by it without difficulty.
Whitworth had not ground them together at all. After a preliminary planing, he had covered one plate with a thin film of rouge and pressed the other on to it. The resulting spots of rouge on the upper plate were carefully brought down by scraping, and the process was repeated, with either plate alternately, until contact distributed the colouring matter evenly over both their faces. To guard against the possibility of a slight hollow in one plate coinciding with a similar elevation in the other, Whitworth made a third plate, and brought it to a perfect fit with first one and then the other of the original pair.
THE COMPARATOR at the International Weights and Measures Bureau at Sevres, France, comparator is a machine for comparing lengths with standards. The International Prototype Metre serves a purpose similar to that of the Imperial Standard Yard; it measures 39-370147 in. In the photograph is Doctor C. E. Guillaume, the inventor of invar, a nickel and steel alloy, in which many copies of length-standards have been made.
He then knew that all three were truly plane. He next devised a planing machine which used one of his planes as its fixed base, and by means of which his true surfaces could readily and accurately be produced in quantity. The plan of scraping, used by Whitworth in preference to grinding or filing, is the best known method for the original production of a true surface.
An outstanding example of the accuracy of mechanical construction which could be reached on occasion, even in mid-Victorian times, is the Peters writing machine. So extraordinary are the performances accomplished on this machine that, to most people, they will sound altogether incredible. Yet they are perfectly well authenticated. The machine belongs to the Royal Microscopical Society and is still in working order.
A London banker named Peters began working in the early 1850s on a machine for producing, by a system of compound levers on the lines of a reducing pantograph, exceedingly minute writing in pen and ink. His idea was that endorsements and so forth made by such a machine could not be forged so long as the machine remained unique. In this attempt he failed, as the machine’s writing, if reduced beyond the limit of hand forging, became totally illegible. But he then modified the machine to write, or rather scratch, with a diamond point on a glass plate. In such circumstances there is virtually no limit to the fineness of the writing which can be produced, except the backlash in the bearings of the levers. Peters reduced this as far as possible by forming them of true-turned steel balls, encircled by collars through which fine-pointed screws were tapped at three points.
In one of his early experiments he wrote the words “Matthew Marshall, Bank of England” in the space of 1/356000 of a square inch, the writing being quite invisible to the naked eye, but perfectly legible, and appearing well-formed, under a magnification of 2,000 or over. He succeeded, later, in writing the whole text of the Lord’s Prayer in 1/356000 of a square inch. The Lord’s Prayer (short form) contains 223 letters, the whole text of the Bible 3,566,480. On this scale, then, the whole text of the Bible could be written in the space of 1/22 of a square inch. In other words, the whole text of the Bible could be written six times on a glass disk the size of a silver threepenny bit, and there would still be a little blank space left over.
Even more wonderful things have been done in recent years by an improved form of the Peters machine built by A. McEwen, of New York, and the Rev. J. G. Crawford. It has produced legible writing on the scale of forty-nine Bibles to the square inch, rather more than twice the reduction obtained by Peters. Even this is not the machine’s limit.
THE MICROPANTOGRAPH is a machine for reproducing microscopically writing in pen and ink. This machine was invented about 1852 by N. Peters, of London, a Fellow of the Royal Microscopical Society. It can reproduce the entire text of the Bible, legibly written, forty-nine times on a square inch.
In 1833 Whitworth left London for Manchester, and set up in business for himself, renting a single room furnished with steam power and putting up a modest sign, “Joseph Whitworth, Toolmaker, from London”. In the intervals of business he turned his attention to the standardization of screw threads.
At that date there was no such thing. Every maker had his own type of threads and his own sizes of bolts. Thus, whenever any machine needed repair, it was necessary to make special screws, copied as nearly as possible from those used in it.
Whitworth began by making a collection of all the bolts used by the more prominent engineering firms throughout the country. Using these as a basis he evolved, as a compromise between their varying diameters and threads, a standard type of bolt. This had the pitch of its thread definitely related to its diameter, with the sides of the thread inclined at the most suitable angle, and with its edges rounded off at a fixed proportionate distance from their point of meeting. Whitworth circulated copies of his new standardized bolts, and their advantages soon brought them into general use. So thoroughly had his pioneer work been done that the standard Whitworth screw thread has since been adopted throughout the world.
Whitworth also broke new ground in another direction. When he set up for himself, a workman would consider that he had done well if he produced a piece of turning which, when measured, was “right to the thirty-second of an inch”. Whitworth evolved a plan by which, without wasting any time in precise measurement, any handy workman could keep within much finer limits of accuracy. If, for example, a shaft had to be turned down, with great precision, to a certain diameter, Whitworth provided the turner with a steel plate in which two accurately cylindrical holes had been drilled, differing in diameter by, say, one-thousandth of an inch (or, in exceptional instances, even as little as one ten-thousandth), one being larger and the other smaller than the required size of the shaft. All that the workman now had to do was to ensure that the finished shaft could be passed completely through the larger of the two holes, and would not enter the smaller.
An inversion of this plan was followed in drilling accurately-sized holes, the gauges then becoming a pair of small solid cylinders, accurately turned and differing slightly in size. This simple gauge soon came into widespread use, and to-day it is the foundation of all modern repetition work, ensuring complete accuracy within any assigned limits and avoiding the delay which would be incurred by measurement.
In connexion with the manufacture of his gauges, Whitworth devised a simple and marvellously accurate measuring machine. Essentially, it consisted of a cast-iron bed, formed into a true plane, along which two sliding-blocks could be traversed by rotating truly-threaded shafts passing through them. The inner face of each block also was formed into a small Whitworth plane, exactly at right angles with the bed and with the common axis of the two shafts. The outer end of each shaft carried a large and carefully divided wheel. The pitch of the screw threads and the radius of the wheels being known, it was a simple matter to calculate how much the inner faces of the blocks would approach and diverge if either wheel were turned the amount of one division and, in this way, to ascertain the exact length of a gripped between the two faces.
Whitworth Measuring Machine
To ensure uniformity,a “feeler piece” of known thickness was first placed at one end of the bar, and the two gripped. One wheel was then cautiously turned back until the “feeler” dropped out by its own weight. The readings of the wheel-graduations, less the thickness of the feeler, then gave the exact length of the bar. A machine of this type, for which Whitworth received a Council Medal at the Exhibition of 1851, was capable of measuring any bar not exceeding a yard in length to the nearest millionth of an inch.
The main features of Whitworth’s work, his true planes, his measuring machine, and his system of gauges, set a standard which accurate machine-shop methods have followed ever since. Much has been done to develop and extend them. For example, the few original standard sizes of Whitworth bolts have multiplied into hundreds of different styles and types, whose specifications are available to engineers all over the world. His measuring machine appears in many shapes and sizes, down to tiny portable micrometer-gauges no bigger than an ordinary watch; and his system of gauges has been applied to the most diverse uses, and in a great variety of forms. Virtually the only outstanding innovation, in principle, which has appeared in modern gauge methods is the Johansson type of block gauge, first produced about 1917.
Invented by Carl Edward Johansson, a Swedish engineer, these gauges consist essentially of a series of small metal blocks, two opposite surfaces of which have been formed into a pair of Whitworth planes which are accurately parallel (a most difficult matter to ensure) and a given distance apart. This distance, as stamped on each block, is correct to within two parts in a million.
A remarkable feature of the Johansson blocks, recalling the adhesion of the ordinary Whitworth planes, is that a series of them can be “wrung” together — that is, caused to adhere by bringing their surfaces together by a combined push and twist — to form a built-up rod whose length is the exact sum of the component blocks. The method by which the original Swedish gauges never similar gauges are now produced in Great Britain and in the United States. They were initially regarded merely as an expensive luxury, but their value is now fully acknowledged, and they are in considerable demand.
COMPARING AND TESTING a yard measure with a microscope. The yard is the standard of length in the British Empire and the Imperial Standard Yard was established by Act of Parliament in 1855.
