Indicators

CHAPTER N I N E
First published by
THE C ANADIAN MUSEUM OF MAKING

www.museumofmaking.org with the assistance of Nevill Publishing & Design, Hove, England Copyright John Walter, 2008, 2011
Updated to 27th May 2011
The right of John Walter to be identified as the author of this work has been asserted by him in accordance with the Copyright, Designs and Patents Act of 1988. All rights reserved
This chapter has been up-loaded to allow free access to information as work in progress. Other chapters will be made available as they are completed: please check our websites for details.
P R O D U C E D BY J O H N WALTER e-mail: [email protected] website: www.archivingindustry.com
A CONCISE GUIDE
Totalisers & Integrators

The introduction of the engine indicator was a part of a growing realisation that the efficiency of machines could be improved. The germ of the idea had been expressed in Duty, a measure of the lifting capacity (and, therefore, power) of an engine expressed in the number of pounds of water that could be lifted for each bushel of coal that was burned in the furnace-grate. The great eighteenth-century engineer John Smeaton had used Duty as the arithmetical basis for a study of the operation of the atmospheric engine, and it was taken, by James Watt and others, as a method of calculating the licensing fees that were due to the patentees, proprietors or erectors of the engines. Once the value of the bushel was standardisedit had varied from district to districtDuty became a tool of competition. This was particularly true in Cornwall, where the great single-acting Cornish Engines of ever increasing size and sophistication vied to be at the head of the published lists. The greatest Duty was customarily achieved by the largest (and usually the newest) engines, an indication more of size than efficiency. Gradually, however, the introduction of Woolf two-cylinder compound engines, which were much smaller than the Cornish variety of comparable Duty, showed that great size was not necessarily desirable. Though the success of the Woolf design was short-lived, many inventors began to see the value of higher steam pressures and ways of preventing unnecessary loss of heat. The thermal efficiency of the early steam engines was very poor, but changes for the better were easily made. The McNaught-type indicators demonstrated the ease with which diagrams could be taken, but many observers doubted if a single cycle could be representative of a protracted period of operations. This was particularly true if the engines were double-acting (unless two indicators were used simultaneously) or if they were being used under varying load. One of the first attempts to provide a Totaliser was made by the mathematician Henry Moseley (180172), then the Professor of Natural and Experimental Philosophy at Kings College, London, who fitted an accumulative recorder to a water-pumping engine in 1842. The instrument,
PAGE 9-3
THE ENGINE INDICATOR
Professor Moseleys Constant Indicator for Steam Engines was described in detail in The Practical Mechanic and Engineers Magazine (1842, pp. 4546). After crediting the basis of the idea to the Frenchmen Poncelet and Morin, in particular to the Compteur (an early form of dynamometer) built by Morin in the early 1830s, the article concentrates on the indicator The plate accompanying the description shows that the instrument, made by Holzapffel & Company some time in 1841, had a large skeletal cast-iron frame, roughly 2ft 6in high and a foot wide, to which the cylinders, springs and recording apparatus were attached. Each spring was about two feet wide, and the integrating cone and its stem were set at an angle of 45 degrees to the vertical rod joining the pistons. A separate steam cock led to each cylinder. The description read: C and D are cylinders, each four inches in length, communicating by the steam pipes A and B with the top and bottom of the cylinder of the engine In these cylinders work two solid pistons, each four square inches in area, fixed upon the extremities of the same piston-rod E F, which piston-rod (when the steam passages A and B are open, and the indicator is in action) sustains, in the direction of its length, a pressure equal to the difference between the pressures upon the two pistons fixed upon its extremities, orequal to the effective pressure of the steam on four square inches of the piston of the engine. This pressure upon the piston-rod is made to bear, by means of a shoulder Z, upon a steel spring S T, which spring is connected, by means of links at its extremities, with a second similar and equal spring Q R, supported at its centre upon a solid projection P, from the cast-iron frame of the instrument. The pressure of the piston-rod upon the lower spring borne at the extremities of the upper spring, whose centre is fixed, is thus made to separate the two springs from one another, and the separation produced isdirectly proportional to the pressure sustained, so long as the deflections are small One surface of each spring is plane, and the opposite surface of that wellknown parabolic form by which an equal strength is given to every portion of the length of the spring By this connexion of the piston-rod with the springs, its position is made to vary directly as the effective pressure upon its extremities, or as the effective pressure on four square inches of the piston of the engine; so that every additional pound in that pressure will cause the piston-rod to alter its position by the same additional distance in the direction of its length. The pulley or wheel I K (which, from the peculiar functions assigned to it in the machine may be called the integrating wheel,) turns upon the piston-rod as its axis, and traverses it at the same time in the direction of its
PAGE 9-4
A CONCISE GUIDE
Plate 91. A drawing of Professor Moseleys Constant Indicator, from The Practical Mechanic and Engineers Magazine (vol. XII no. 4, 1842). PAGE 9-5
lengthbeing prevented from moving on it in that direction by means of two shoulders fixed by adjustment screws. The arms of this wheel are pierced by apertures, through which pass three rods united at their extremitiesso as to form a rigid frame G H, which frame turns also upon the piston-rod as its axis, but does not traverse with it in the direction of its length; so that the wheel I K is made by its motion with the piston-rod to traverse the rods of the frame longitudinally, whilst it is made to sweep the frame round with it by any motion which may at the same time be communicated to it about its axis. It receives such a motion of rotation from the cone K L, which is so placed that its side may be accurately parallel to the piston-rod, and which is kept continually pressed against the wheel at K, by means of a spiral spring enclosed in a tube at M, and acting continually against the extremity of the spindle on which the cone turns. A system of bevel wheels U, Y, X, communicates to the spindle, and with it to the cone, the rotation of the pulley N, which pulley is driven by a cord carrying a weight at one extremity, and passing to the other extremity (over directing pulleys) to the piston of the engine, or to some point which moves precisely as the piston of the engine does, but through a less space. The circumference of the pulley N being thus made to move precisely as the piston of the engine, the angle described by the cone, during any exceedingly small period of time, is made to be exactly proportional to the space described during that time by the piston of the engine. Now, let it be observed, that the circumferences of the integrating wheel I K partaking of the motion of that portion of the surface of the cone with which it is at any instant in contact, the number of revolutions, or rather parts of a revolution, which it is made to describe during any exceedingly small period if time, beginning from that instant, is dependent on two causes; first, upon the dimensions of the particular circle of the cone which is at that instant driving it, and, secondly, upon the particular angle through which it is driven by that circle: that is, it depends, first, upon the distance of its point of contact K with the cone from the apex of the cone at that time, and, secondly, upon the angle which the cone describes about its axis during that small period of time. Moreover, that if either of these two elements of variation remained always the same, then the number of revolutions, or parts of a revolution, made by the wheel, would vary directly as the other; whence it followsthat when (as in the present case) both these elements vary, it varies as their product; or that the number of revolutions, or parts of a revolution, made by the integrating wheel during any exceedingly small period of time, varies [sic] directly as the product of two factors, of which one is the angle described during that time
PAGE 9-6
A CONCISE GUIDE
by the cone, and the other the distance of the point of contact K of the cone and wheel from the apex of the cone. The integrating wheel carries round with it the frame G H, on the hollow axis of which frame is fixed a pinion, running into a wheel V, whose axis turns upon bearings fixed to the frame of the instrument, and the number of whose teeth is to that of the teeth in the pinion as ten to one. This wheel carries also a pinion running into a second wheel, their numbers of teeth being in the same proportion, and so through a train of five wheels and pinions. The circumference of each of the four last wheels is divided into ten equal parts, and numbered, and the circumference of the first into 100. The number of revolutions of the integrating wheel is thus registered to five places of integers, and one place of decimals. The Moseley Indicator was duly attached to a Cornish engine installed in the East London Water Works, Old Ford, where, working continuously from 28th January to 25th February 1842, it recorded 232,617 strokes. The trial was reported in the Journal of the Institution of Civil EngineersResults of a Trial of the Constant Indicator upon the Cornish Engine at the East London Water-works. By Professor Moseley, F.R.S., &cwhere it was deduced, with totally unacceptable precision, that units of work (each representing a pound of water raised by a distance of one foot) had amounted to 21,464,067.1727. These were claimed to accord with those obtained by Thomas Wicksteed, Engineer to the East London Water Works Company, possibly on the basis of McNaught indicator cards: work per stroke upon every square inch of the piston was 120.574 from the Constant Indicator, compared with 119.338 by Wicksteeds method.[1] Much of the prolonged discussion that followed centred on the accuracy of the Constant Indicator, the most telling coming from Josiah Parkes (17931871) and John Farey (17911851). The uncertain knowledge of thermodynamics in 1842 allowed virtually any form of claim and counter-claim to be made, and the merits of the instrument were largely overlooked in favour of the simple visual aspect of the McNaught cards. Consequently, the comments made by Parkes, best known as a deep-drainage engineer, were self-important and critical; Farey, author of A Treatise on the Steam Engine (published in 1827), who had received from Mr John Taylor indicators cards of Taylors Engine
1. Wicksteed had been sent to Cornwall in 1837 to investigate the remarkable Duty being returned by the Cornish mine engines. He was sufficiently impressed to buy an engine owned by the East Cornwall Silver Mining Company, which had been lying idle in Wheal Langford. This, which had a cylinder with a diameter of 80 inches, was re-erected in London by Harvey & Company of Hayle, its builders. By the time of Moseleys trials, a problem with the valves had been overcome and the engine was working regularly.
PAGE 9-7
at the United Mines. offered criticisms of his own. Farey then remarked that although the new instrument does not preserve any record of the minute details of any one stroke, like the ordinary Indicator, but it records a true aggregate of all the details of any number of succeeding strokes: it gives the same results as would be obtained if it were possible to have two ordinary indicator cards correctly taken, at each succeeding stroke of the engine, during the whole time of observation He added In trying the performance of a steam-vessel, alternately up and down the measured mile in the river Thames along the shore of the Dartford Marshes, it is usual to take an indicator card from each engine, at every such run; and by summing up each card, some difference will be found between them, wherefore an average of the results of several cards will give more authentic information respecting the force exerted by the engines during the whole trial, than could be obtained if one such card had alone been depended upon. The new instrument [Constant Indicator] takes cognizance of every stroke that is made by the engine during the whole time of observation; and in cases (such as in the Great Western steamer) where a considerable variation in succeeding strokes occurs frequently during such time, it is a desideratum to obtain the results which this instrument is intended to give; and which, as far as it has been tried, it seems likely to give with fidelity. The instrument, when applied as it had been at Old Ford, becomes another mode of ascertaining performance, similar to what is reported monthly respecting the engines in Cornwall Others were prepared to follow Moseleys lead, and Daniel Gooch (181689) was the first to develop a continuously-recording indicator which could be used successfully on a railway locomotive travelling at speed. This instrument owed its origin to the Battle of the Gauges, a controversy that raged in Britain in the 1840s over the ideal dimensions for railway track. One powerful faction, led by the Stephensons, claimed that the standard gauge (then 4ft 8in) was superior to the Broad Gauge (7ft 01/4in) championed by the maverick engineer Isambard Kingdom Brunel. One of the sharpest differences of opinion concerned the resistance afforded by air pressure, friction and other factors to a moving train. Results from a variety of partisan tests were published, but the most contentious were those given to the Gauge Commissioners by Daniel Gooch, which the narrowgauge faction believed to be much too low (and, by implication, fraudulent). Gooch determined to investigate the problem rigorously, building one of the first dynamometer carriages to measure the pull exerted by the locomotive.
PAGE 9-8
A CONCISE GUIDE
Plate 92. A Great Western Railway locomotive of the type first tested by Daniel Gooch, pictured in the 1850s after conversion from 222 to 422 configuration. John Walter collection.
The deflection of a massive spring balance was recorded on a paper roll, driven by a belt from one of the carriage axles, and a clock marked the paper every few seconds to allow the speed to be calculated. Gooch also developed an engine indicator to measure the changes of pressure against the passage of time. The mechanism incorporated a horizontal piston which acted against two halfelliptical springs, and an amplifying lever attached to the piston rod drew a continuous trace on a paper roll driven from the wheels of the locomotive. The resulting instrument (still in the collection of the Science Museum, London) was large, heavy and painstakingly made, but proved to work extremely well. The first trials of the dynamometer carriage were undertaken in April 1847, with locomotives Great Western and Dreadnought, though it is not known if the indicator was being used at this time. However, it had certainly been perfected by September 1847, when comprehensive experiments were undertaken with Iron Duke. Writing in his book Railway Engineering (1855), Daniel Kinnear Clark recorded that: The application of the indicator to the cylinder of a locomotivehas aided essentially in giving precision to our knowledge of the behaviour of steam in the cylinder. Mr Gooch was the first, so I am aware, who applied the indicator to the locomotive; and he is the only
PAGE 9-9
engineer who systematically applies the instrument as a sort of stethoscope. The records were obtained by Daniel Gooch and his assistants, balancing precariously on the buffer-beam of a locomotive travelling at speeds of sixty miles per hour. Clark added: Only those who have undergone the fatigue and riskwith occasionally a brisk side wind to make them conscious of how short a distance there is between time and eternitycan fully appreciate their value. Unfortunately the pencil of the Gooch indicator worked in an arc, and the trace could not be compared directly with the output of a McNaught instrument. This limited its utility exclusively to the Great Western Railway; though the dynamometer carriage survived for many years, to be rebuilt in narrow-gauge form after the demise of the Broad Gauge in 1892, the indicator had been retired in favour of Richards-type instruments many years earlier. In the 1850s, the Frenchman Arthur-Jules Morin (17951880) developed an alternative to the McNaught indicator embodying a continuous paper roll. Morin had entered the French army in 1810, rising to command an artillery unit (1817), then became lecturer (1819) and Professeur de mcanique (1830) at the school of artillery in Metz. He left the army, with the rank of general, to be the administrator and later director of the Conservatoire des arts et mtiers in Paris. He also served a term as president of the Institut de France. Well-known for his commitment to public hygiene, Morin was also an inveterate experimenter. In addition to the indicator, planimeters are known bearing his name; and among his many publications were Leon de mcanique pratique (Mathias, Paris, 1846), Notions gomtriques sur les mouvements et leurs transformations ou lments de cinmatique (Lacoix-Comon, Paris, second edition, 1857) and tudes sur la ventilation (Hachette, Paris, 1863). Morins indicator was conceived in an attempt to combat the considerable inaccuracy [that was] likely to result from the difficulty of constructing the spiral springs in Macnaughts indicator, so as at all parts of the stroke to denote equal pressures by equal divisions Scott & Jamieson, writing in The Engineer and Machinists Assistant (c. 1854), described the instrument as adapted for being fitted to the cylinder cover of the engine; it carries a stop-cock pipe G, furnished with two keys; between these is situated a small horizontal cylinder H, in which a solid piston is accurately fitted to work steam-tight. Towards the middle of the piston rod m, which is properly guided to a rectilinear course, is a square part in which is inserted the lower end of a long parabolic spring n, the other extremity of which is fixed to the summit of a standard I, forming part of the frame work of the machine, the spring being so fitted as to admit of a certain amount of travel in the piston in
PAGE 9-10
A CONCISE GUIDE
Plate 93. Drawings of one form of the Morin indicator, from Armengaude in, Publication Industrielle des Machines, Outils et Appareils (Paris, 1859). John Walter collection.
both directions. The square boss of the piston rod carries also a small pencil o, for the purpose of tracing the different degrees of tension of the steam on the opening of the lower cock G.
PAGE 9-11
Two pencils p, p, are placed in holders fixed to the framing exactly opposite to the point at which the pencil o stands when the stop-cock G is shut, and being thus immoveable [sic], serve to mark a continuous atmosphere line. A third pencil q, which is susceptible of a slight degree of vertical motion in its socket, and is destined to mark the termination of each stroke, is brought into contact with the paper by placing the instrument so that the working beam, cross-head or any other rigid part of the engine may touch lightly at the end of the stroke, the top of an upright rod u, which is connected by a system of levers r, s, t, with the top of the pencil q. A continuous band or roll of paper may be subjected to the action of this machine for an indefinite period, so as to produce diagrams representing the action of the engine during several successive strokes. The manner in which this is accomplished is as follows: the roll of paper is first wound upon the cylinder L, by means of the handle y; it is then passed over the three small rollers v, v, v, placed to oppose the pressure of the pencils, and is received upon the cylinder M, situated at the opposite end of the framing Q, Q. The axis of this latter cylinder is produced on one side so as to form also the axis of the conical pulley or fusee N, opposite to which is situated a cylindrical drum O, which receives a uniform motion from any rotating part of the engine to be operated on, by means of a worm wheel w, on its axis, geering [sic] with an endless screw on the axis of the strap pulley P. The cylindrical roller O communicates motion to the conical roller N by a cord wrapped round both and fastened at opposite extremities of each. The object of this arrangement is to compensate for the increased surface velocity due to the increased diameter of the cylinder M as the paper is wound on to it, by imparting to it a proportionally retarded motion. Scott & Jamieson added that although highly ingenious in many of its details, and capable of giving very correct indication, the Morin design was wanting in that portability and compactness which has very materially contributed to bring Macnaughts instruments into such general use. At about this time the limiting or slice indicator appeared, using the Macnaught and similar single-diagram designs to build diagrams from a series of horizontal slices in the hope that the resulting composite would allow the average performance of the engine to be gauged. This had the advantage of allowing conventional methods of analysis to be used, but the continuousfeed indicator still had its champions. The Steam-Power Meter protected by US Patent 20514, issued on 8th June 1858 to George Schuh of Madison, Indiana, was another of the roll-recorders. Schuh, or perhaps his immediate ancestor, was part of a wave of German
PAGE 9-12
A CONCISE GUIDE
Plate 94. Drawings (from the relevant U.S. Patent) of the continously recording indicator invented by George Schuh in the 1850s. Note that it is labelled as a dynamometer. PAGE 9-13
immigration to this particular part of the USA in the middle of the nineteenth century. Very little is known about his background, but he was clearly aware of the existence of other indicators. His improvement, he stated, has reference to a meter which has before been applied to a steam engine for ascertaining and registering the amount of work done by the engine during any period of time, the principle of action of which has consisted in the production of two motions, one of which was equal to the motion of the engine pistonand the other equal or proportional to the pressure of the steam on each square inch of the pistons surface, and, by combining these two motions, a third motion was produced under which all possible variations of the two former, was always equal to the first multiplied by the second and represented and indicated by means of a counter the amount of work done Misleadingly labelled Dynamometer on the US patent drawings, the instrument consisted of a vertical standard supporting a bed-plate to which the recording mechanism was anchored. A short piston with a double-ended rod was contained in a long, small-diameter cylinder held to the bed plate by the threaded plugs in which the steam pipes terminated. Steam could be applied to either end of the cylinder, moving the piston laterally. This in turn moved the pendulum arm against the resistance of an inverted V-spring, and moved the friction wheel from its rest position at the centre of the disclike friction plate (where it did not move) out across the plate surface in direct proportion to the changes in steam pressure. The rotation of the friction wheel was recorded by a counter mechanism, relying on the pendulum arm, which had moved the carriage to the opposite side of the bed-plate, to add the results of the return stroke of the engine to the results of the outward stroke. The friction disc was driven by a cord, wrapped around its drum, led around intermediate pulleys, and attached at each end to a suitable reciprocating point on the engine. Small spiral compensating springs were interposed between the ends of the cord and the anchor point to minimise inaccuracy arising from the shock of reversal of motion or irregularities in the action of the friction disc. The Steam-Power Meter was a fascinating attempt to provide a calculating indicator, without ever encountering long-term success. No surviving example is known, and it is possible that only a single working prototype was ever made. Until more is known about George Schuh, the mystery will remain. Another attempt to provide a recording indicator was made by S. Lloyd Wiegand and William Barnet Le Van, the appropriate issue of US Patent 41182 dating from 5th January 1864. The indicator was one of the first to embody the spring-type pressure diaphragm that had been pioneered in Europe twenty
PAGE 9-14
A CONCISE GUIDE
years earlier in pressure gauges. The pressure of steam within the U-shape diaphragm tilted a lever tipped with a tracer, which acted in concert with a roll of paper, driven from the engine across a flat bed, to record continuously as paper transferred to a take-up spool. A pawl-and-ratchet mechanism was included in the drive to ensure that the paper moved only in one direction. Supplementary drawings showed a variety of diaphragms, including corrugated discs and bellows, and Richards-like methods of amplifying the comparatively small movement of the diaphragm. The Multi-Meter, or Steam Telegraph proposed by George Clarke of New York and Marmont Edson of Brooklyn, and protected by US Patent 77584 of 5th May 1868, was another continuous recorder. However, though it relied on a trace created by a simple spring-and-piston system, the drum was rotated continuously (if imperceptibly) on a time basis to provide a daily record of performance. It did not produce an individually analysable diagram.
Plate 95. Multi-view drawings of the diaphragm type continuously-recording indicator patented in the USA during the Civil War by Wiegand and Le Van. By courtesy of the US Government Patent Office, Washington DC.
PAGE 9-15
Plate 96. A typical Ashton & Storey recorder, showing the layout of the principal components. Note that this particular example has a McNaught-type direct-reading pointer, but others, undoubtedly later, used a Richards-type amplifying system. From Max de Nansouty, Chaudires et Machines Vapeur (1911).
William Ashton and John Henry Storey of Manchester, centre of the English textile-manufacturing industry, patented a variety of recorders in the 1860s and 1870s. Typical of these was the design patented in Britain in 1869 and in the USA on 11th June 1872 (no. 127729). The drawings show another variation of the shifting friction-disc drive, relying on splines on a rod, placed vertically,
PAGE 9-16
A CONCISE GUIDE
engaging a pinion to drive a series of counter wheels. Displacement of the axis of the horizontal disc, elevated by the steam pressing on the piston, from the centre of the vertical discdriven by a belt from the engineprovided a cumulative indication of power. The illustrations also show how a method of amplifying the movement of the piston, and how a conventional reciprocating cylinder indicator could be included to allow conventional single-cycle diagrams to be taken when required. The possible position of an integrating cone is also sketched. Ashton & Storey recorders were made in small numbers, usually with the mechanism entirely encased. Later examples may include a Richardstype amplifying indicator instead of the direct-reading type shown in the 1872 specifications. Work had ceased by 1880. Attempts were made to fit McNaught indicators with continuousroll apparatus, most notably by the French clockmaker Paul Garnier in 1857 65. However, the success of the Richards design held immediate attention. Continuousrecord instruments of this type often relied on three drums one holding the paper roll, one recording, and the other receiving the used rollthough others had only two and a few attempts were made to contain the entire assembly in the recording drum. Elliott Brothers of London made Darke and Improved Richards indicators of this type, but the universal acceptance of pillartype torsion springs at the expense of flat spiral springs in the drumbase subsequently favoured external paper feed systems. Paul Garnier (180169), who had been apprenticed to the great clockmaker Louis Breguet, opened his own workshop in Paris in 1825, invented the coupballe (chaff cutter) escapement in 1830, and was responsible for the first electric clock to be made in France (1849). The first Garnier MacNaughts resembled British prototypes, but had two springs and two drums. Arnal and Gauthier, in Trait de Mecanique (published c. 1894), described the construction of these instruments in detail. The body, a brass cylinder, consisted of two chambers screwed together; the upper cylinder accepted the two-part spring assembly, with collars at the base and lower mid point and a threaded cap. The ring-tip piston rod passed through a small hole in the centre of the spring chamber and up through the spring to emerge from the cap. Its lower end was anchored in a hollow-face piston by a hexagonal-head nut. The twin drums were supported on a comparatively fragile platform formed integrally with the split-jaw collar that clamped to the indicator body with the aid of a small screw. One drum contained a roll of paper, which was wound on to the other as the cord attached to the engine reciprocated. The hinged trace arm and
PAGE 9-17
Plate 97. Drawings of the perfected single-drum version of the McNaught-type indicator made in France by Paul Garnier of Paris. One version was made with a roll of paper inside the drum. From Max de Nansouty, Chaudires et Machines Vapeur (Paris, 1911).
PAGE 9-18
A CONCISE GUIDE
the detachable spring scales were conventional, but the fairlead of the British instruments was replaced by a reducing pulley attached to an axle mounted transversely beneath the platform. A dial-type speed indicator and a detent to hold the recording mechanism out of gear were also fitted. The original Garnier indicator, judged to be fragile and much too complicated, was soon replaced by a simplified version (described in Chapter One) with a single spring and a single drum. He also made a distinctive MacNaught-type indicator with a magazine drum with steadying gear driven by a lightweight ratchet bar engaging a pinion on the transverse axle. A second pinion on the midpoint of the axle meshed with a toothed ring on the drum, operating against resistance afforded by a helical spring in a chamber at the axle tip. Most examples held a roll of paper within the drum and had a clutch, actuated by a lockable pivoting lever, which could put the mechanism in train when diagrams were to be taken. The Martin-Garnier indicator seems to have been similar, but prevented steam impinging directly on the piston head by interposing a flap valve. The goal was to reduce the shock of impact, with the attendant risks of inaccuracy in the recording, but required a separate cock to drain condensation from the steam chamber. The Clair design was also based on the MacNaught. Its principal claim to novelty concerned the recording drum, which relied on double helical bevel gearing and two special ratchets (deux encliquetages Dobo) to translate reciprocating motion into continuous rotation. The so-called limiting or slice indicators dated from much the same period. The earliest of these was probably the work of an Alsatian engineer, Gustave-Adolphe Hirn (181590). Alterations had been made in the 1850s to a co-axial McNaught-type indicator that may originally have been the work of Hopkinson. Hirn preferred the direct-reading instrument where absolute accuracy was concerned, but acknowledged that the oscillations that affected the pencil posed a real problem when the engine-speed rose above 2025rpm. An answer was found by limiting the recording to a fraction of the cycle, then repeating it section-by-section until an overall picture was created. Pre-setting the spring by confining movement within narrow limits ensured that the piston could not rise until the lower pre-set position had been reached and stopped the movement when the upper pre-set had been reached. Hirn had first applied the method crudely, simply by pressing down on the pointer with his finger, but then progressed to mechanical stops. Owing to the rotation of the drum, this had the effect on the power stroke of drawing a straight line at the lower limit, then a small ascending curve as pressure rose and another straight line at the upper limit; the exhaust stroke
PAGE 9-19
was similar, excepting for a descending curve as the pressure decayed. The resulting diagram was composed of a number of piston strokes, instead of recording a single stroke. This had certain advantages, as it allowed the performance of an engine to be assessed on the basis of an average instead of a single (possibly atypical) cycle. It also allowed a picture to be constructed without the pencil following the same track over and over again, with the resultant difficulty of deducing a mean. The later MacNaught-type limiting indicators used by Hirn were made in Paris by Paul Garnier. One method of adapting the indicator consisted of attaching a T-bar to the tip of the piston rod. The bar was allowed to move in a short vertical slot in a bifurcated block attached to a long threaded rod. This rod passed through a boss on a top-cap held to the indicator-body cap by two slender standards. Semi-circular cuts on the edges of the bifurcated block engaged the rod-like standards to ensure precise vertical movement, and the height of the block could be varied merely by turning the knurled disc-like head of the threaded rod. In this way, the pressure at which the resistance of the spring was overcome (and the tracer began to move vertically) could be minutely controlled. Another modification of the MacNaught indicator, also made by Paul Garnier, was developed by Marcel Deprez (18431918). Born in the rural commune of Aillant-sur-Milleron, Deprez, keen to pursue experiments of his own, left the cole des Mines in Paris without completing his course. Yet his gifts were such that he was almost immediately appointed as secretary to the director of the college, Charles Combes, and began a remarkable series of investigation into the pressures developed by gunpowder. Deprez subsequently became fascinated by electricityhe was among the first to demonstrate that electric current could be conveyed over long distancesbut not before he had made his mark on the development of the indicator. Like Hirn before him, Deprez sought to create a single diagram from a series of horizontal slices or tranches minces. A review of the perfected design in Engineering, on 22nd September 1871, after crediting Deprez with several mathematical investigations on the distribution of steam, who has invented an expansion gear and a simple and ingenious calculation for ascertaining the principles of distribution, drew attention to the introduction for engineers and contractors of a new type of Watts dynamometric indicator, which shows an real improvement on the well-known apparatus. Claiming that it was well known that when the speed of the engine exceeds 60 revolutions per minute, the piston of the indicator is thrown beyond the point corresponding to the maximum pressure of the steam
PAGE 9-20
A CONCISE GUIDE
Plate 98. Two typical McNaught-type limiting indicators: by Hirn (left) and Deprez (centre). From Jacques Buchetti, Guide pour lEssai des Machines Vapeur (Paris, c. 1886). PAGE 9-21
owing to the inertia of the moving parts, piston, piston-rod and spiral spring, Deprez reduced the weight of the components by making the piston and its rod of aluminium. The review describes the indicator: The spiral spring is enclosed as usual in a cylindrical tube having a vertical slot for the passage of the pencil-holder. Below the spring is placed a horizontal forked piece which enters the tube through a slot. The piston rod, which passes through both arms of the fork, carries between them [sic] an enlarged part, the height of which is a little less than the distance between the two arms of the fork. By the aid of an outside screw the position of the fork can be regulated in such a way as to compress the spring up to a point corresponding to a certain pressure of steam, for example, 40 lb. per square inch. It will be easily understood that if such an indicator is put in communication with a steam cylinder, the piston cannot move until the pressure reaches 40 lb., and the pencil will describe a straight line on the paper rolled around the drum of the indicator. If in one point of the stroke the pressure exceeds 40 lb., the piston yields, and will begin to move, but as soon as the enlargement of the piston rod reaches the upper branch of the fork, this movement will also stop. The pencil will then trace , after having described a horizontal line, a small portion of a curve, then another part of the horizontal line, and the point of departure of the portion of the curve will mark the point in the stoke when the steam reached the pressure of 40 lb. It is sufficient, while the engine is in motion, to turn the screw, and thus to give movement to the fork in the vertical slot of the tube, to obtain as many points as are wished On joining these points by a continuous trace, the curve of expansion is given without any error or regularity. Deprez also made what must be considered among the first of the maximum-pressure indicators that were to become popular in the twentieth century, using a series of individually encased pistons with pre-loaded springs to indicate the exceptionally high peak pressures associated with the use of gunpowder in firearms. The illustration accompanying the Engineering review, though crude, shows the lining mechanism to have been actuated by a fine-pitch screw supported in a collar around the body of the indicator. The actuator was a small horizontal handle (which would have fouled the casing if constructed as drawn!), and the piston was contained in a separate chamber beneath the body. An illustration in Jacques Buchetti, Guide pour lEssai des Machines Vapeur et la Production conomique de la Vapeur (c. 1886), shows the drive with right-angle bevel gears allowing the handle to move in a vertical plane.
PAGE 9-22
A CONCISE GUIDE
Slice diagrams were subsequently employed in Britain both by Farnum & Bodley and by Moses Wayne (see Chapter One), and in the USA by John Burkitt Webb of Ithaca, New York State, who proposed an alternative form of reciprocating-drum indicator. This was the subject of U.S. Patent 265198, Diagram-instrument for Engines, granted on 26th September 1882. Webb had been appointed Professor and head of the department of civil engineering in Talbot College (subsequently incorporated in the University of Illinois) in November 1871, remaining until departing for Cornell University in 1878. He was also active in the promotion of the American Association for the Advancement of Science, serving as Vice-President of Section D (Mechanical Science). Webbs goal was similar to that of Deprez: a system of indication which would be suited to locomotive and other engines running at high speeds. According to the patent specification, the improvements were: first and most important, an improved form of diagram and of the apparatus for describing the same, which enables me to render the accuracy of the former
Plate 99. The larger of the two Burkitt Webb indicators, combining a tablet recorder with a limiting system. By courtesy of the US Government Patent Office, Washington DC.
PAGE 9-23
independent of the speed and of the friction of the moving parts; second, an attachment for automatically measuring and recording the area of the diagram during its formation; third, an apparatus for the systematic reduction of the stroke of the engine to the exact stroke required of the indicator; fourth, arrangements for conveniently stopping, starting, and adjusting the instruments. The patent shows two forms of indicator. The larger of the two, intended for laboratory work, assembled diagrams with the assistance of more than thirty individual strokes scribed onto a card attached to a curved plate, the location of the ends of each of the strokes replicating the trace of a conventional indicatorthough, Webb claimed, without showing any of the degrading effects of vibration. The simpler version of Webbs design, presumably intended for series production, had an oscillating drum placed on an upright frame containing the piston cylinder. The piston rod extended upward through a short sliding sleeve until its threaded tip could be screwed into a four flanged collar held in the jaws of a forked arm forged integrally with the frame. A parallelogramtype pointer mechanism was pivoted on the sliding sleeve, and attached to an anchor point on the frame by a short link. The diagram was constructed from 1520 engine strokes instead of a single out and back cycle. It was operated by screwing the piston rod manually down as far as it would go, where the spring was at its weakest and the recording pointer was at the bottom of the diagram. The first outward stroke of the engine piston drew a straight line on the drum of the Webb machine; on the return movement, however, a sprung pawl attached to the drum rotated a collar flange through 90 degrees. Acting on the rapid pitch thread on the piston rod, this effectively lifted the piston and increased the pressure applied by the spring. The process continued until the diagram had been completed, each band showing where the steam pressure had overcome the resistance of the piston spring, simultaneously raising both the piston rod sleeve and the pointer. Like the Deprez patterns before it, the Webb instrument gave surprisingly good results in circumstances that would have defeated a Richards indicator.
The later integrators

The oscillating-drum indicator rapidly proved to be a surprisingly good recorder of single strokes, and would operate continuously if a roll of paper
PAGE 9-24
A CONCISE GUIDE
could be wound from spool to spool after each stroke. Unfortunately, each diagram required individual assessment before an idea of average performance could be obtained. The obvious answer to this problem was a return to the idea, promoted by Professor Moseley, of a calculating indicator which would automatically analyse the results of multiple engine strokes. Counting strokes was easily accomplished, but it was much more difficult to design a recorder that could combine the rise and fall in steam pressure with the elapsed time of each stroke. THE LEA INDICATOR The advent of the Amsler polar planimeter and the Richards indicator, with the ability to amplify the trace movement fourfold, encouraged Henry Lea to combine these two instruments in an integrating indicator. This was the subject of British Patent no. 785/77, sought on 26th February 1877 and sealed on 13th July, though, oddly, the complete specification was not deposited with the Great Seal Patent Office until 21st August. Lea was born in 1839 in Birmingham, Warwickshire, the son of prosperous merchant Nathaniel Lea, whom the 1871 census of England and Wales describes grandly as Member of the Birmingam Stock Exchange. In 1862, the younger man had circulated details of his new office in Birmingham: Henry Lea begs leave respectfully to announce that by the advice of many gentlemen well acquainted with his qualifications and experience, he has commenced practice as a Consulting Mechanical Engineer. Henry Lea is now remembered as a champion of the incandescent light, in 1882 supervising not only the installation of electrical lighting in Birmingham Town Hall in 1882the first in a public building in Britainbut also the pioneering street-lighting system in the London borough of Chelsea. He subsequently specialised in central heating and air-conditioning systems, the Royal Victoria Hospital, Belfast, being among the first buildings in the world to be air conditioned.[2] The Lea indicator was a fascinating, if short-lived departure from the practice established by Watt, Southern, McNaught, Richards and Thompson. After briefly describing the established forms of indicator in his patent application, Lea ventured his opinion that the work done by the engine is sometimes so regular that a single diagram will represent with sufficient
2. Henry Lea died in 1912, control of his business passing to his eldest son Frederick McKenzie Lea (1877 1939). Shortly after Frederick Lea died, his younger brother Donald Lea amalgamated the Henry Lea & Company Practice with Edwin S. Hoare & Partners of Bristol. Hoare, Lea and Partners is now known simply as Hoare Lee.
PAGE 9-25
PAGE 9-26
A CONCISE GUIDE
Plates 910 and 9-11. Two illustrations of the Lea Planimetring Indicator. Plate 910 (left) is taken from the patent granted in 1877; plate 911 (right) is taken from The Engineer for 15th December 1882. Note that though these are practically identical, the patent drawing shows a counting mechanism in the drum-top to record the number of strokes made by the engine. That this is omitted from the later engraving may suggest that Lea indicators were made without it. Plate 810 by courtesy of the UK Intellectual Property Office, London; plate 811, John Walter collection.
accuracy the power exerted by the steam during a period of several minutes or hours, but in most cases the action of the steam is very irregular, and in such cases the diagrams may vary so rapidly that it becomes impossible to ascertain correctly what total power the steam has exerted for any period other that that during which any particular diagram or diagrams was or were taken. One method of ascertaining the mean pressure per square inch from a diagram drawn by the indicator is to measure by means of a planimeter that area in square inches contained within the outline of the diagram, and then to divide the area thus found by the length of the diagram Leas design, therefore, simply combined the indicator and a planimeter to give an easilyread assessment. Built on the general lines of a Richards indicator, Leas design was easily recognisable. The trace arm had a tiny wheel where the pencil-point would normally lie, and an abnormally long rearward overhang to accommodate the drive to the gear train of the recorder (usually with five individual dials).
PAGE 9-27
The recording mechanism was carried in a special bvox-like housing, with a hinged lid, which replaced the standard that normally supported the links and levers of an amplifying mechanism. The remainder of the instrument was conventional. It is assumed that Henry Lea, from a wealthy background, commissioned a few indicators from a well-known instrument maker: Elliott Brothers or, perhaps, Casartelli. The drawings accompanying his patent are sufficiently detailed to suggest that the design had been finalised by early 1877. Engravings published in the British weekly periodical The Engineer on 15th December 1882 replicate the patent drawings, except for the counter in the drum-top that was originally intended to record the number of strokes. The Lea planimetring indicator was theoretically capable of recording a single stroke, though it only gave a final reading and gave no clue to the steam distribution that would have been obvious from a pencil trace. Its true role was to record and calculate the mean performance of an engine over a large number of strokes. However, The Engineer also recorded in December 1882 that under certain conditions the instrument yielded extremely accurate results, but not so under all conditions, hence the invention was abandoned [and is]placed on record as a link in the history of the indicator. No survivor has been found. The accuracy of integrating indicators of Lea type obviously depend greatly on the precision of the integrating wheel and the accuracy with which the counting-gears are cut. Friction, backlash and similar losses in the gear train could be minimised, if not entirely overcone; but vibration, rapid reversal of motion in the drum-drive, the effects of moisture (or, worse, oilmist) on the movement of the integrator over the surface of the drum would all conspire to reduce the value of the readings. The Lea indicator would have been far more suited to a laboratory than an engine-house. THE BOYE INTEGRATOR The subject of British Patent 2409/80, granted to Lykke Boye, of Bergen, Norway, but at present of Dumbarton, in the County of Dumbarton, Scotland on 14th June 1880, this was another of the indicators embodying a Bourdon tube to register pressure. Boye worried that the design of the near-contemporaneous Kenyon and similar Bordon-tube patterns, with lever-type amplifiers and reciprocating drums, was very seriously flawed. This arrangement is faulty, he claimed, because the inertia and momentum of the piston, levers and cord seriously affect the sensibility of the action of the instrument, which can thus only
PAGE 9-28
A CONCISE GUIDE
give a very incorrect application of the work done in the cylinder His solution, described in greater detail in Chapter One, amalgamated a vertically moving carriage with a horizontal Bourdon tube, which moved the drum segment according to changes in pressure. The carriage had a rod-like base, sliding in a tube in the indicator body, and a bracket, projecting horizontally, to support a tracer attached to the tip of a slender pillar. The carriage was actuated mechanically by attaching it to a drive-rod connecting with any suitable reciprocating part of the engine.
Plate 912. The Boye Planimetring Indicator, from the British patent granted in 1880. Note the integrating wheel on top of the drum segment E. By courtesy of the UK Intellectual Property Office, London.
PAGE 9-29
Plates 913 and 914. Two views of the Boys Power Meter, which had a tablet-type recorder in addition to the integrating mechanism that recorded results on the dials. The drawing is taken from Buchettis book (c. 1886).
The Boye integrator or mean pressure recorder was attached to the carriage rod by a collar. Splines in the collar engaged helical slots in the carriage-rod extension, giving the integrating wheel an additional reciprocating motion as changes in pressure moved it across the flat top of the drum segment. The results were shown numercially on a counter. THE BOYS INTEGRATOR The planimetring indicator designed by Charles Vernon Boys was protected by British Patent 2449/81 of 14th June 1880. Among the most renowned physicists of his day, Boys, the son of a clergyman, was born in 1855 and educated at Marlborough College. He studied at the Royal School of Mines in 18736, then joined the Royal College of Science in 1881; by 1889, Boys had risen from the post of Demonstrator to Assistant Professor of Physics.
PAGE 9-30
A CONCISE GUIDE
Charles Boys was best known for his work with quartz fibres, soap bubbles and high-speed photography. Elected Fellow of the Royal Society in 1888, Boys was the recipient of many awards; these included the Rumford Medal of the Royal Society (1924) and the Elliott Cresson Medal of the Franklin Institute, Philadelphia (1939). He was knighted in 1935 and died in St Mary Bourne, Andover, in 1944. A Boys Medal was subsequently awarded annually by the Institute of Physics, until it was renamed the Moseley Medal in 2008. The Boys indicator, known as the Engine Power Meter, sought much the same goal as the Lea pattern. However, it was also capable of simultaneously providing the trace of a single stroke or an accumulation of strokes produced at the same engine-setting. Its construction included a cylindrical brass piston housing, flaring into three standards to support the platform that held the integrating arm and a drum mounted on a reciprocating carriage driven from the engine cross-head. A drive-rod from the drum to a gear train allowed readings to be taken from a large dial at the front of the platform.
Plate 915. A photograph of the platform of the Boys Power Meter, showing the way in which a single diagram could be taken even as the instrument was working as a totaliser. Canadian Museum of Making collection.
PAGE 9-31
PAGE 9-32
A CONCISE GUIDE
Plate 916, previous page. A US Patent granted in 1886 to Frederick Low protected an integrating indicator. This drawing suggests that the prototype was converted from an American Steam Gauge-type Thompson, though the patent was assigned to Ashcroft, maker of the Tabor. By courtesy of the US Government Patent Office, Washington DC.
The instruments were made only in very small numbers, and were rarely seen outside colleges. The example in the Museum of Making collectionthe only known survivoris marked BOYS PATENT ENGINE POWER METER over No. 3 over ELECTRICAL TRADING COMPANY LIMITED. above 2 VICTORIA MANSIONS, WESTMINSTER, LONDON and SOLE MANUFACTURERS. in four lines. Perhaps fewer than a dozen indicators of this type were ever made. THE LOW INDICATOR Among the first of the integrating instruments to be developed in North America, this was the subject of US Patent no. 350069 of 28th September 1886. The inventor, Frederick Low of Chelsea, Massachusetts, assigned the patent to the Ashcroft Manufacturing Companymaker of the Taboreven though the drawings suggest that an American Steam Gauge-type Thompson provided the basis for the prototype. Low claimed that providing an indicator with a planimeter-wheel would allow it to measure the work of the engine during its movement, and shall immediately indicate it, and without the use of a card. In other words, instead of using a pencil to mark upon the card the indicating-diagram, I employ a planimeter-wheel so mounted and operated that it shall indicate in proper terms the variations noted by the pencil and while the engine is in motion In carrying my invention into practical effectI use very largely the mechanism of the ordinary steam-engine indicator, simply substituting for the pointer and parallel-motion levers or arms the planimeter-wheel and holding device and suitable connections The Low indicator, consequently, followed the standard concealed-spring design in all respects other than the amplifying mechanism. A lever attached to the tip of the piston rod pivoted in a forked standard attached to the collar. As pressure rose, propelling the piston upward, the lever rotated in the fork; this motion turned the yoke or counter-bracket assembly (secured at right angles to the operating lever) to alter the angle made between the integrating wheel and the surface of the drum. The combination of the rise and fall of the operating arm and the reciprocation of the recording drum, to which a blank card could be fixed, allowed the average horsepower to be read directly from the counter mechanism.
PAGE 9-33
The Low indicator may have been made in small numbers, but never attained commercial success. In its simplicity lay an important weakness: the small size of the planimetring wheel and the counters restricted its accuracy. THE LITTLE INDICATOR The instruments protected by British Patents 17575/94 of 15th September 1894 and 29199/97 of 17th September 1897 (issued to William Godson Little of Blendon Grove, Bexley, in the County of Kent, Engineer, and Charles William Godson Little, of Heckington, in the County of Lincoln, Engineer), and equivalent US Patents 546897 and 622504 of 25th September 1895 and 4th April 1899 (to William G. Little, of Bexley, and Charles W.G. Little of Heckington, England), pursued similar goals. They were all intended to provide a continuous record of the output of an engine working under varying load, keeping an accurate tally as work proceeded. William Godson Little was born in St Pancras (London) in September 1852, son of William Little of Eye in Northumberland and Elizabeth Godson of Heckington in Lincolnshire; Charles William Godson Little was born in
Plate 917. Taken from the 1895 US Patent, these drawings show the original form of the Little integrating indicator, built on the basis of an internal-spring Crosby. By courtesy of the U.S. Government Patent Office, Washington D.C.
PAGE 9-34
A CONCISE GUIDE
Plate 918. The perfected form of the Little integrator, shown here, was an impressive and efficient design. Unfortunately, this example is completely unmarked. Though obviously built on a Crosby body (with considerable modifications), it is not as obvious where the alterations were made. Canadian Museum of Making collection.
Sleaford, Lincolnshire, in March 1866, son of Charles H. Little and Ann E. Little (ne Godson?).[3] Both inventors came from prosperous backgrounds.
3. The 1851 English census reveals that Charles Little, then aged 20, was living with farmer Josiah Godson in Heckingon. It is assumed that William Little and Charles Little were brothers, and that they each married into the Godson family (though the 1851 census returns do not reveal the identity of another Godson daughter).
PAGE 9-35
The Little indicator The operation of the instrument was explained in Engineering, 10th December 1897: A is the cylinder of the indicator containing the piston and spring, which are of the usual standard type. B is a carriage, sliding on to the upper part of the cylinder, and securely held in position upon the bracket of the instrument by the conical screw G which, on being screwed home, brings the axis of the shaft C into a position to cut the axis of the drum O at right angles. C is a shaft provided with a rigid arm P carrying the counter F and one pivot of the axis of the wheel E, the other end of which is pivoted on the shaft itself. The wheel E is dished to allow of the forward bearing H being brought close up to the contact slide H of the drum O, and in line with the shaft C. The counter F consists of two toothed wheels, one of 100 teeth and the other of 99 teeth, both engaging with a worm on the axis of the planimeter wheel, the upper one being graduated in 100 divisions, a simple and well known device for recording a large number of revolutions. D is a crank carrying a pin I working in the slot of the crosshead R, which pin is kept in contact with the lower surface of the slot by means of the small spring S. The spring I bears upon the pointed end of the shaft, keeping the wheel in contact with the drum at fixed pressure. N is a rocker enabling the wheel to be placed in contact with the drum or not at will. K is a frame carrying the bearings of the drum O, the latter being provided with a hollow spindle which can be filled with oil for the lubrication of the upper and lower bearings, which are conical centres. The lower part of the frame is split, and fitting accurately over the spring box on the bracket, can be tightened up in any desired position by means of a single screw. The milled head xxx is a removable stop to the drum O. M is a device for accurately measuring the stroke of the drum while in motion, and which, by means of vernier, can be read to show 1/200in L is a hard brass sleeve which can be slid up and down, allowing a change of contact surface if necessary; but as the pressure of the wheel upon the slide is very slight, a great number of revolutions can be made upon one place before the line of contact can be detected even by the finger nail
PAGE 9-36
A CONCISE GUIDE
Plate 919. A top view of the Little integrating indicator, showing the planimetring wheel, the counters, and the drum-stroke measure. Canadian Museum of Making collection.
William Little had been described in the 1861 British census as a newspaper proprietor, living in Queens Road in the fashionable Regents Park district of London; by 1881, he had retired to Heckington to farm four hundred acres, perhaps inherited from his father-in-law. The 1891 census records William G. Little living on his own means at Blendon Cottage, Bexley, Kent, with his wife Ada and a young daughter; by 1901 he was a chemical manufacturer living in Blendon Grove, Bexley, with his family. The scale of his success is implicit in the employment of domestic servants, gardeners and a coachman. Charles W.G. Little was listed in the same census as an electrical engineer and employer, living in the Camberwell district of London with his wife and small daughter. Though the Little indicator could not analyse individual strokes in the traditional manner, tests showed it to be accurate. One such experiment was undertaken in June 1895, by Professor William Unwin of the Engineering Laboratory of the Central Institution, London, in which the original 1894type Little was matched against an internal-spring Crosby indicator. Reported in Engineering on 10th December 1897, the tests suggested that comparable results were obtained from both instruments. Readings from the low-pressure cylinder of the steam engine under test, taken simultaneously over a period of 25 minutes, gave indicated horsepower of 15.55 (front end) and 14.72 (back end) with the Crosby compared with 15.56 and 14.87 for the Little. A Willans engine running at 450 rpm gave
PAGE 9-37
PAGE 9-38
A CONCISE GUIDE
Plate 920. A page of drawings from the patent granted in the USA in 1902 to protect the integrating indicator designed by Augustus Nagel. By courtesy of the US Government Patent Office, Washington DC.
mean pressures of 14.22 lb/sq.in (Crosby) and 14.41 lb/sq.in (Little), but more impressive was a test undertaken on a single-cylinder Marshall horizontal engine. This revealed that the results obtained under brake load, 26.4 and 26.1 horsepower for the Crosby, 26.3 and 26.03 horsepower for the Little, were comparable. Figures obtained under variable load showed much greater difference: 82.1 and 84.6 horsepower for the Crosby, 86.86 and 87.98 horsepower for the Little. Unwin suggested that, in his view, the continuous indicator has recorded the correct horsepower, and...the diagram indicator has failed to do so, owing to changes of load. He also offered the opinion, in July 1897, that the improved Little integrating indicator was likely to be considerably more accurate than the form I tested, [as]some defects of the original instrument have been removed. The Little indicators were not robust enough to withstand run-of-the-mill engineering work, and were confined to laboratories, schools, colleges and educational establishments. Production, never large, came to an end with the rise of the optical indicator in the early twentieth century. Very few survive; currently, only two have been found. Unfortunately, the 1897-type instrument in the Canadian Museum of Making collection yields no clues to its origin, though the basis of the body is quite clearly a Crosby internal-spring pattern that has been altered to support the new recording mechanism. The absence of markings (not even a number) suggests that the alterations were made privately, perhaps by a clockmaker; the quality and accuracy of the work are exceptionally good. THE NAGLE INDICATOR Patented in the USA on 11th March 1902, no. 695227, this was the work of Augustus Nagle of Montclair, New Jersey. A combination of a conventional internal-spring indicator and a counter mechanism, it relied on a frictionwheel carriage, attached to the piston rod, which rose vertically between two slender standards supporting an enclosed multi-wheel counter. Reciprocating motion was imparted to the small horizontal friction wheel by contact with a vertical disc driven by a pulley connecting with a suitable part of the engine. As the pressure varied, the recording wheel rose or fell to an appropriate point, relying on its position relative to the engine-driven disc to permit an assessment of average power to be made.
PAGE 9-39
Plates 921 and 922. The original pivoting carriage MaihakBttcher indicator (below), from Stausss 1911 book; and (bottom) the 1919-type recorder in its case. Museum of Making collection.
PAGE 9-40
A CONCISE GUIDE
THE MAIHAK INDICATORS The Lea, Boys and Little integrators were made only in small numbers, and then, perhaps, only used for experimentation. Interest in instruments of this type seems to have lapsed for several decades, only to reappear in Germany. Among the earliest was a design patented in Germany on 12th November 1908 by Anton Bttcher of Hamburg (DRP 212628, Zhlapparat fr die indizierte Leistung von Kolbenmaschinen). Comparable protection for an Improvement in or relating to Integrating apparatus for Steam engine and like Indicators was accepted in Britain on 4th August 1910 as no. 26015/09 of 1909. US Patent 1039250 of 24th September 1912, Counting Apparatus for the indicated output of Piston-engines (which includes an additional claim
Plate 923. The 1924-model Maihak-Bttcher integrating indicator was an improvement on the 1919 instrument, with the integrating unit separated from the paper drum. The new design could record a single-cycle diagram while simultaneously averaging results accrued over a period of time. By courtesy of Bruce Babcock, Amanda, Ohio, USA.
PAGE 9-41
Plate 924. A top view of the 1924 Maihak-Bottcher indicator, showing how the recording dials were separated from the paper drum. The two parts of the system could be engaged independently. By courtesy of Bruce Babcock, Amanda, Ohio, USA.
for electrical control and digital recording), names the inventors as Anton Friedrich Claus Bttcher and Ludwig Wilhelm Lehmann, of Hamburg, Germany, Assignors to the Firm of H. Maihak, of Hamburg, Germany. The original or 1908 design could be identified by an integrating wheel set in a carriage on top of the recording drum and a weight projecting from the carriage to keep the wheel in place. Bttcher claimed that his intention was to improve on the construction of apparatus[by providing] a combination of parts comprising an indicator drum upon the upper surface of which a friction wheel is adapted to bear and when revolving, to operate the counter mechanism. The friction wheel is journalled in a framingand means are provided consisting of a sliding weight adjustable on an arm connected to the said framingfor adjusting the pressure of the friction wheel on the indicator
PAGE 9-42
A CONCISE GUIDE
drum. The friction wheel is adapted to be radially displaced on the upper face of the indicator drum by means of the pressure in the indicator cylinder [acting] through a double bell-crank lever, one end of which is connected with the piston rod of the indicator and the other end to the framing in which the friction wheel is journalled... A few of these instruments were made by Maihak of Hamburg prior to 1914, recognisable by traditional construction, with a prominent pillar rising from the platform between the drum and the amplifying gear to receive the bell-crank lever pivot. A conventional card could be indicated on the drum, though the integrating-wheel carriage had to be disengaged (by detaching the tensioning spring, then pivoting it back and upward) before cards could be changed. A few 1919-type instruments were made immediately after the end of the First World War, with a simpler sheet-steel carriage and links between the piston-rod extension and the integrator. Finally, in the mid 1920s, a muchmodified Bttchers Leistungszhler Maihak Modell 1924 appeared, driven by a combination of gearing at the base of the drum and a rocking rod engaging a collar on the piston rod. This was a much better design: the integrating mechanism and the counter unit were much more compact, and single-cycle cards could be taken without moving any of the other components. Yet the Maihak/Bttcher planimetring indicators were still too complicated for most applications and were usually confined to research or teaching facilities. Few survive, and so it is difficult to judge the true extent of production. Another integrating design was offered by Heinrich Eicke of Hamburg. Protected by German Patent 341366, granted on 7th March 1916, this may also have been made by Maihak. The basis of the instrumentPlanimetriender Indikator oder Leistungszhler, planimetring indicator or pressure meter was a standard Lehmann-Maihak external spring indicator, but the drum was replaced by a reciprocating conical drum (set an at angle to the platform) and a planimetring wheel/counter unit was attached directly to the piston-rod extension beneath the spring. The wartime origins of the Eicke design may have contributed to its failure to reach production status.
C0ntinuous diagram recorders

Indicators of this type come in three classes: those in which paper is fed manually before each diagram is taken, those in which the paper is advanced automatically before each diagram is taken, and those in which paper is fed
PAGE 9-43
continuously as the diagrams are taken. The first two give the characteristic closed loops, displaced from each other by the amount of feed; the third gives a diagram in the form of an unbroken line with peaks corresponding to maximum pressures. EARLY NORTH AMERICAN DESIGNS Patent records testify that the idea of continuously monitoring the performance of a piston engine was not lost on inventors in the USA. One of the most interesting of the earliest attempts was made by Reuben Collins of Dollar Bay, Michigan. Protected by US Patent 495148 of 11th April 1893 (sought in May 1892), this relied on a combination of electromagnets and the duplicated multi-drum feed, between opposing sector plates, to record the out and back strokes successively on separate rolls of paper. A central amplifying mechanism with a double-headed tracer marked diagrams of each set of strokes alternately. Collins was well aware of the value of his invention. So far as I am familiar with indicators of this class, he claimed, none has been designed for producing continuously recorded, separate diagrams; all such prior devices having been designed for the taking of single diagrams or readings and when it has heretofore been desired to obtain several readings, representing successive strokes of the engine, the diagrams have so conflicted withy each other (because of the fact of their being practically on the same field) as to make it difficult if not altogether impossible, to trace out the lines of each individual diagram and even if it were possible to pick out and measure the lines separately, it would be impossible to determine the relation each diagram bore to the strokes of the engine This invention seeks to overcome this defect by presenting to the pencil point or stylus a new surface of paper for each stroke of the engine. The drawings accompanying the patent specification suggest that the basis of the Collins indicator was an Arc (see Chapter Five), and that the movement of the tracer was radial instead of a truly vertical line. The continuouslyrecording instrument was undoubtedly capable of working effectively, but a combination of large size, complexity and the use of electromagnets doomed it to become little more than a footnote to the history of the indicator in the 1890s. Based on the Thompson indicator, the continuous recorder proposed by Englishman Thomas Gray, living in Terre Haute, Indiana, marked a paper strip which could be fed from the supply drum to the storage drum by means of an idler and a friction-drive roller. The recording mechanism was driven
PAGE 9-44
A CONCISE GUIDE
by a chain or cord from the engine passing over a segmental wheel, with the return stroke controlled by a cord (with two inset springs) running over a twopulley bracket attached to the vertical arm of the body-platform extension. The Gray indicator was protected by US Patent 615151 of 29th November 1898. Though complicated and clumsy, the system had the merit of providing diagrams of constant lengththough Gray acknowledged that the feed roller could be omitted and the paper wound directly into the storage drum if precise comparisons between successive diagrams were unnecessary. The indicator patented by Alva Hill of Detroit, Michigan (US no. 687827 of 3rd December 1901), was remarkable for the drive system contained in a box below a frame or bed piece attached to the platform of an otherwise standard internal-spring indicator: the patent drawings show a Crosby. Two bevel gears and two clutches in the drive allowed the Hill indicator to feed a roll of paper from a storage drum, between the feed cylinder and a tensioning roller, and out into a suitable receptacle. Unlike the Gray instrument, in which the drum reciprocated, Hills design relied on the clutches to turn the feed cylinder continuously in the same direction even though the engine piston reversed its motion at the end of each stroke. THE MATHOT INDICATOR The advent of high-speed steam- and internal-combustion engines, bringing a need to indicate progressively higher pressures, has been discussed in Chapter Seven. Among the first inventors to offer anything other than a small-diameter piston and strengthened amplifying gear was Rodolphe Mathot of Brussels, author of the influential Gas-Engines and Producer-Gas Plants (originally published in French in 1903). Born in 1866, soon after the Lenoir engine had entered production, Mathot was granted British Patent 14896/00 of 1900 to protect an auxiliary clockwork drum that could record the pressures obtained from successive piston strokes. This enabled the researcher to tell not whether each individual stroke was flawless, but if the engine was running consistently. In his Provisional Specification, Mathot observed that: indicators are at present used in connection with the testing and verifying of the work of thermal engines, but they generally prove insufficient for controlling the said work in explosion engines, such as gas and petroleum engines. The diagrams taken on these afford only indications or data on each explosion taken separately and from the stand-point of the initial pressure at the time of same, etc. Now, all the explosions which succeed each other, whether successive or
PAGE 9-45
distant on account of misses, differ from one another in their importance, their efficiency, the time of ignition of the mixture, etc. For controlling the complex phenomenons [sic] which take place during the motion in explosion engines, it is accordingly important to note during the proper time, all the explosions and misses which succeed each other under the various conditions of work or regulation The patent drawings show the drum mounted on a Schaeffer & Budenberg indicator, but the best-known use of the Mathot patent was made in Glasgow, where T.S. McInnes & Co. Ltd of Glasgow (and the successor, Dobbie McInnes Ltd) made a few Mathot-type instruments combining a conventional vulcanite-clad external-spring body with a clockwork-driven paper drum. Intended for use with explosion engines (gas, oil or petrol), the springdriven paper mechanism did not give a closed diagram, nor could power be measured directly (the rotation of the drum was not proportional to the movement of the piston), but the trace did give record the pressures reached in the cylinder during each cycle. The original Mathot recorder, containing a roll of paper about twelve inches (30 cm) long, gave a trace consisting of a large number of peaks representing the pressures generated during each power stroke. A review published in Engineering on 30th May 1902 noted that Many explosion enginesrun too fast for it to be feasible to use a reciprocating drum on an indicator. The Mathot recorder servesto furnish a continuous diagram, from which much may be learned as to what is happening in their cylinders The effects of different proportions of inlet and exhaust passages, and of various sizes of carburetters, can be seen very clearly It also stated that there were: three paper drums, the band travelling from the first, across the second, and being coiled on the third. The second drum is driven by clockwork contained in the box below it. This clockwork can be started and stopped, at will, at either of two speeds. When other speeds are required, the driving wheels at the top of the drum are changed for another pair of different proportions. Diagrams ranging from 1/8in to 3in. or more in length can be obtained as a result. The roll of paper is dropped on the left-hand drum, and is drawn out over a fixed surface acting as a brake to keep it taut. It then passes round a considerable proportion of the centre drum, and is finally wound on the right-hand drum. There is a fixed, but
Plate 925, next page. Taken from a catalogue published in 1930 by Dobbie McInnes & Clyde Ltd of Glasgow, this shows the Mathot-type continuous-roll recorder and its accessories. By courtesy of Peter Forbes. PAGE 9-46
A CONCISE GUIDE
PAGE 9-47
The Cipollina-Dobbie recorder

The diagram, taken from a volume in the ICS Reference Library series, shows how the two pencil arms a work against the solitary drum b to draw a diagram for the performance of each end of a double acting cylinder simultaneously. A continuous roll of metallic paper, contained in drum b, is taken from spindle c, between the guide rollers d and around the exterior of the recording drum to a second spindle e. One of the indicator cylinders is connected with the head end of the engine cylinder, and the other with the crank end. Two independent cocks f and g are placed between the indicator and the engine. The cord h drives the recording drum from a suitable moving part on the engine, usually by way of reducing gear. The ratchet wheel i is propelled forward by one tooth for every revolution of the drum. A cam wheel j rotates with the ratchet, to which it is attached, and carries a pair of small spigots on its surface. A pawl k resting on the cam wheel is raised by the spigots as j revolves. Acting in conjunction with the connecting bar l, the pawl k brings the pencil points m into contact with the drum. Each time the pencil points m are lifted from the recording drum b at the end of each stroke, paper is unrolled automatically from spindle c and a corresponding amount is wound onto spindle e. Altering the cam wheel j allows the number of diagrams taken per minute to be changed from 25 to 50 or 100. The wheels i and j share their spindle with drum n, which receives a band of paper travelling from spindle o under the guide pulley p. The toothed wheel q is moved forward by pawl r through a distance of one tooth each time the recording drum b turns, allowing pencil s to mark the paper strip appropriately. Every tenth tooth was customarily removed to simplify counting. The terminals t t could also be connected with a clock capable of completing the circuit at pre-determined intervals. Each time this happened, the pencil u was drawn toward the coil v by means of armature w and marked the same roll of paper which was recording the drum revolutions. This allowed the running speed of the engine to be calculated.
PAGE 9-48
A CONCISE GUIDE
adjustable, pencil, by which the atmospheric line is drawn along the whole length of the paper Improved indicators of Mathot type were still being offered by Dobbie McInnes & Clyde Ltd and then Dobbie McInnes Ltd until the Second World War began, though demand had never been large: production is unlikely to have exceeded a few hundred. At least a few were fitted with an improved drive system protected by British Patent 225906 of 11th August 1923, granted jointly to Walter Provan Clyde and Dobbie McInnes Ltd. This included a slip-friction clutch mechanism to allow uniform winding with the increase in diameter of the take-up roll. THE CIPPOLINA RECORDER Dobbie McInnes Ltd, after making a few No. 3 external spring indicators with an optional paper-roll feed and an auxiliary drum-spring unit beneath the platform, then progressed to the unique Cippolina Continuous Double Diagram Recorder. Production of this complicated two-drum unit began about 1905. Protected by British Patent 25366/02, sought on 18th November 1902 by a naval engineer, Giuseppe Cipollina of Via Prione, Spezia, Italy, this indicator was capable of recording continuously from one or two cylinders simultaneously. Basically two Dobbie McInnes No. 1 indicators sharing a common platform, it incorporated unidirectional feed, an electromagnetic clutch, and time markers driven by clockwork; and relied on friction drive to move paper contunuously from the spool to the take-up drum. The diameter of one of the drums could be altered by adjustung a cone, as the four segments of the drum were held together by an elastic ring. A patent to protect an improved form of the Cipollina indicator was sought in Britain on 30th September 1907. British Patent 21653/07 was issed to Gaspare Bernati and Luigi Cecchi of Genoa, but US Patent 956680 of 3rd May 1910 (comparable with the 1907 British protection) names Giuseppe Cipollina, of Genoa, Italy as the inventor and Gaspare Bernati and Luigi Cecchi, of Genoa, Italy as the assignors of the patent. Complexity and high price, and the emergence of efficient electrically powered rivals, kept production to a minimum; and the Cippolina instrument is now rarely encountered, even though production continued into Dobbie McInnes days. HOGLE INDICATOR Among the earliest of the roll-feed designs, this was protected by US Patent 898474, issued on 15th September 1908 to Milton W. Hogle of Pittsburgh,
PAGE 9-49
PAGE 9-50
A CONCISE GUIDE
Plate 926, previous page. Dating from the mid 1930s, this Dobbie McInnes & Clyde clockwork-drive continuous recorder is an improved form of the Mathot design of the early 1900s. It is also fitted with electrical control gear. Canadian Museum of Making collection.
Pennsylvania. An application had been made as early as February 1906. The patent drawings show two Crosby-type indicators placed side-by0side at the front of a large platform supporting the continuous-recording system in such a way that two recording pencils may move in closely adjacent parallel lines without interference. Drums were used to record the performance of each end of a cylinder in the form of separate traces on a roll of paper passing across an idle platen roll, relying on a ratchet and pinion to translate reciprocating motion into movement in a single direction. This method gave characteristically open diagrams. In the words of the patent, ... CROSBY-LANZA INDICATOR This large, comparatively expensive but sophisticated roll-feed system was based on US Patent 902586, issued on 3rd November 1908 to Gaetano Lanza, Professor Emeritus of Theoretical and Applied Mechanics at the Massachusetts Institute of Technology (but sought as early as October 1906). The 1911 handbook, The Lanza Continuous Diagram Appliance with Crosby Indicators states that the instrument ...is not in itself an indicator, but displaces the ordinary drum as a means of supplying the paper for taking indicator cards, and any indicator may be combined or adapted for use with it. It is assembled upon a bracket or frame, which is designed to support also the indicator and its connections, and that these parts may be rigidly fixed in proper mutual relation. Upon this bracket are mounted the spindle for receiving the new roll of paper, the drum which feeds the paper forward and upon which the pencil point bears in making the record, and the spool on which the paper is afterward wound. The drum is rotated continuously in one direction by the alternate engagement of two effective clutches controlled by a cord passing over the pulley at the extreme end of the bracket arm and actuated by a crosshead block which is positively connected to the to the crosshead of the engine or some other convenient portion of the machinery which moves in exact accordance with the piston. This connection between the the crosshead block of the Continuous Diagram Appliance and the engine crossheadmay be by any reducing motion which will drive the Appliance positively in both directions on the forward and backward strokes, as a spring is not depended upon for the return stroke as in ordinary indicator drums and reducing motions. Fom
PAGE 9-51
this peculiar and desirable mode of connection there must result accuracy and positiveness of action in making continuous records. The Lanza attachment seems to have been abandoned in the 1920s. TRILL AND SHEPERDSON INDICATORS On 4th June 1912, William Trill of Corry, Pennsylvania, received two US Patents1028154 and 1028155to protect the design of a sophisticated rollfeed indicator. One of the most interesting features of the patents, which had been sought in November 1908 and March 1911, was the skill with which the drawings were executed. The illustrations are also noteworthy as the first appearance of the perfected Trill indicator, with the spring suppported externally between the cylinder and the platform (which were separated by three rod-like standards). Patent 1028154 protected a two-drum design, driven through a clutch to allow open diagrams to be taken continuously or conventional closed-loop
Plate 927. The Lanza-patent indicator was adapted for series production by Crosby. This photograph shows clearly how the standard New Model No. 2 indicator could be used either to produce single-cycle diagrams or a continuous series. By courtesy of Bruce Babcock, Amanda, Ohio, USA.
PAGE 9-52
A CONCISE GUIDE
Plate 928. This continuous-feed indicator was made to patents granted to William Trill in 1912. The mark on the side of the platform could suggest that it was the 211th of its type to be made, perhaps in 1936. By courtesy of Bruce Babcock, Amanda, Ohio, USA.
diagrams to be taken from a reciprocating drum. There were many other improvements in the paper-handling system, and the Faultless reducing gear was also standardised. Patent 1028155 showed a modified continuousrecording system, with a large-diameter paper drum separated by guide rollers from the take-up spool. The movements of the drum and the tracer were controlled electromagnetically. Trill was also the assignee of US Patent 1056950, sought in March 1912 and issued on 25th March 1913 to John Sheperdson of Johnstown, Pennsylvania. The
PAGE 9-53
specification reveals that Sheperdson was a British citizen. His indicator was intended to be used with rolling-mill and similar engines to simultaneously form open records on coinciding ordinals of the variations in pressure in both ends of the cylinder [and to] provide for the formation of records of time intervals, dead centers, the imposition of loads, etc. The drawings show two external-spring Trill indicators, one placed at each end of the platform supporting the paper-handling drums and guides to indicate both ends of the cylinder continuously. WALLACE INDICATOR The subject of a patent sought in Britain on 31st December 1909, no. 30541/09, by James Wallace of West Didsbury, Manchester (and comparable US Patent 1030216 of 18th June 1912, sought in November 1910), this consisted of two Crosby internal-spring indicators attached to a curved bedplate in such a way that the tracersone left-hand, one right-handmarked diagrams from alternate ends of the cylinder on a marking drum placed centrally. One of the Crosby drums was replaced by a drum containing the paper roll, and clutches on the pulleys ensured that the feed was unidirectional. A special clockworkdriven time-recording mechanism could be attached to the platform edge when desired.
PAGE 9-54

Indicators

Uploaded by

Copyright:

Available Formats

Indicators

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Indicators

Uploaded by

Copyright:

Available Formats

What is the purpose of the engine indicator?

What is the purpose of the engine indicator?

How did the engine indicator evolve over time?

How did the engine indicator evolve over time?

CHAPTER N I N E

THE C ANADIAN MUSEUM OF MAKING

P R O D U C E D BY J O H N WALTER e-mail: [email protected] website: www.archivingindustry.com

Totalisers & Integrators

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

The later integrators

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

C0ntinuous diagram recorders

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

The Cipollina-Dobbie recorder

THE ENGINE INDICATOR

THE ENGINE INDICATOR

THE ENGINE INDICATOR

You might also like