The Engine Indicator.
The engine indicator is a fascinating instrument that records how the pressures in the cylinders of steam and internal-combustion engines change during the operating cycle. Used properly, they can identify problems ranging from bad valve settings to constricted steam pipes.
The first attempts to register pressure generated in the cylinders of steam engines were undertaken with columns of mercury, the levels being judged by the displacement of the metal in small-bore tubes. This was unsatisfactory, as the tubes were unwieldy and prone to breakage. A better solution was proposed by James Watt (1736-1819), who devised a small cylinder, containing its own piston, which terminated in a small cone-tipped pipe that could be inserted in a hole bored in the cylinder cover. A spiral spring connected the piston rod with the frame of the indicator, and a pointer attached to the rod-tail moved against a graduated scale.
Watt’s gauge worked better than the inventor had hoped—well enough to enable the characteristics of individual engines to be assessed—but the operator had to watch the movements of the pointer during the piston-stroke and record results manually. This was comparatively easy when only a few results were needed, and the operating cycle was ponderous. As movements became quicker and more complex, and more precise results were demanded, watching the pointer became an unmitigated chore.
The solution was suggested by Watt’s draughtsman, John Southern (1758-1815), who added a board or ‘tablet’ on which a pencil could trace the rise and fall of pressure. The sliding board was arranged to reciprocate during a single cycle, allowing a closed-loop diagram to be produced.
The Watt-Southern indicator was very successful, though Watt guarded
the instruments jealously and knowledge spread only slowly. Indicators
of this type were made into the 1840s, and the idea of a moving tablet
reappeared several times in the late nineteenth century. The culmination
was a Wayne design, patented in Britain in 1894.
The first major advance to be made after the reciprocating tablet was
the oscillating drum. The credit for this is customarily given to the
Scotsman John McNaught—largely on his own testimony!—but there
is some circumstantial evidence to suggest that the idea had occurred
first to Henry Maudslay. The original McNaught indicator had the recording
drum concentric with the piston cylinder, but the perfected version had
the drum on a platform that protruded from the cylinder laterally. A swivelling
pulley, known as the fairlead, allowed the cord attached to a suitable
part of the engine to approach the indicator at an angle.
The McNaught indicator relied on a comparatively steam-tight piston sliding in a tube beneath a large coil spring. When the outward stroke of the engine piston began, the pull on the cord turned the drum through a half-revolution. The admission of steam to the cylinder raised the pencil to make its trace. When the inward stroke of the piston began, a helical spring in the recording drum rotated it back through the half-circle to its starting position. This allowed the pencil to complete its loop.
The McNaught indicator was popular, as it was comparatively simple, easily copied, and acceptable efficient. Many attempts were made to improve it, notably by Duvergier in France and Joseph Hopkinson in England, but many of these experiments sought methods of obtaining continuous diagrams instead of improvements to the basic instrument. From this period, too, came the first attempts to develop the so-called ‘lining indicator’, which constructed an average diagram from a large number of strips taken from successive engine cycles, and the first integrating ‘totalisers’, which deduced the cumulative or average values numerically.
The next great advance was made in the USA, where the first high-speed steam engines had been developed in the 1850s. The advent of the Allen engine, promoted enthusiastically by Charles Talbot Porter (1824-1910), was the catalyst. Porter realised that the McNaught-type indicators being made by the Novelty Iron Works were too prone to vibrate when used at high speed, giving tremulous diagrams that were impossible to interpret, and sought something better. The project was given to consulting engineer Charles B. Richards (1833-1919), who within a very short time had produced a workable design by combining the offset recording drum and the internal coil spring of the McNaught with a system of levers, inspired by Watt parallel motion, that amplified the movement of the piston four-fold. This kept the movement of the piston to a minimum, allowed a short stiff spring to be used to damp vibrations, and kept the instrument as compact as possible.
The Porter-Allen engine and the prototype Richards indicator (made by the Novelty Iron Works) were exhibited at the International Exhibition in London in 1862. There the engine attracted much adverse comment, but proved to work smoothly and in perfect safety—confounding the doom-laden predictions of its many detractors. The indicator was used successfully to test a variety of engines on display, though one or two well-known British engineers refused to have anything to do with it. A trial of a railway locomotive was arranged, and the existence of the indicator came to the attention of Elliott Brothers, renowned as makers of optical equipment. A licence was negotiated, and the first Elliott-Richards indicators appeared in 1863. More than ten thousand of them had been made by 1876, and work continued until the end of the nineteenth century.
The Richards indicator was sturdy and reliable, and was still being used in quantity when Europe went to war in 1914. However, the success of the Porter-Allen engine had begun a quest for ever-greater pressures and ever-increasing speed. Above 250 rpm, the Richards indicator struggled to provide reliable diagrams. This was largely due to the inertia of the parts in the amplifying mechanism, which were long and comparatively heavy.
The first real successor was the work of the American engineer, Joseph W. Thompson, whose indicator was patented with the backing of the Buckeye Engine Company in 1875. The Thompson amplifier (which embodied a more mathematically correct straight-line approximator than its predecessor) was much lighter than the Richards equivalent, taking the form of the letter ‘M’, and soon proved to give good diagrams at 350 rpm or more. Consequently, Thompson-type indicators were made by many companies after Buckeye withdrew. The best known are the American Steam Gauge Company, Schaeffer & Budenberg, James Robertson & Sons and Dobbie McInnes. Dobbie-patent instruments were still being made in the 1960s substantially in their original 1898-type external-spring form.
The indicator designed by Harris Tabor, patented in the USA in 1878,
offered an alternative to the Richards and Thompson patterns. Made by
the Ashcroft Manufacturing Company from the early 1880s, the Tabor relied
on a slotted vertical standard to direct the pointer in a straight line.
The earliest Tabor was too weak to succeed, but the modified design of
1886 proved to be amply strong enough to compete on level terms with the
Thompsons.
When the Thompson patent lapsed, many other designers tried their hand.
This is particularly evident in the USA, where instruments such as the
Straight Line (1890), the Calkins (1890) and the Lippincott (1900) all
briefly prospered. There was even a place for aberrant designs such as
the British Kenyon of 1878 and its US equivalent, the Rae, which used
Bourdon-type pressure tubes instead of piston springs; the 1887-patent
Bachelder, with its adjustable leaf spring placed horizontally; and the
British Simplex of 1894, briefly promoted by Elliott Brothers, which had
a tong-like spring.
If the Thompson system was the most popular prior to 1910, then it was eclipsed by the perfected Crosby indicator. Made in Boston, Massachusetts, the 1882 Crosby and its strengthened 1895-patent successor were made in large quantities. They had the merit of exceptionally light amplifying gear, much lighter than even the Tabor, and a comparative absence of inertia effects.
By 1900, the first of the external-spring indicators were being seen. The earliest is usually acknowledged as the British McKinnell & Buchanan, patented in 1893 on the basis of the Richards mechanism, but the original Maudslay & Field indicator of the 1820s may have taken a broadly comparable form.
Many of the earliest designs were poor compromises, including the external-spring
Tabor, the first Maihak and the 1902-type Dreyer, Rosenkranz & Droop
pattern. The problem was simply that a place for the spring-support standards
or rods had been found on the cylinder cap or platform of an otherwise
conventional internal-spring design. The ideas made sense financially,
but were less acceptable where efficiency was concerned. The external-spring
Tabor, patented in 1900 by William Houghtaling, even had the amplifying
rod spindle in a separate chamber.
The ‘second generation’ of external-spring designs had the
springs beneath the platform, above an abbreviated piston cylinder, and
were also often fitted with vulcanite or similar sheathing to allow springs
to be changed when the metalwork was hot. The most successful of these
indicators were patented by John Dobbie in Britain and by William Trill
in the USA.
‘Third generation’ external spring indicators returned to springs that were mounted on top of the platform, but often relied on extended piston rods or bifurcated or duplicated amplifying links of Thompson or Crosby type to allow the spring to be concentric with the piston rod. Indicators of this type were made in Boston, Massachusetts, USA, by the Star Brass Manufacturing Company (Webster patent), the American Steam Gauge Company (Jerrauld patent), and the Crosby Steam Gauge & Valve Company (Davidson patent); in Pittsburgh, Pennsylvania, USA, by the Bacharach Industrial Instrument Company (Maihak copies); and in Hamburg, Germany, by both Maihak and Lehmann & Michels to Wilhelm Lehmann’s patent of 1909.
The development of the high compression internal-combustion engine during the First World War, particularly for aerial use, presented new problems. Conventional continuous-recording indicators such as the Mathot or the Cippollina (both made by Dobbie McInnes) and the Lanza (made by Crosby) could not be used satisfactorily in the air, owing to the excessive stresses involved, and optical indicators developed for laboratory work—such as the Hopkinson and Birstall designs—were equally inappropriate. The first to be used successfully in the air was the Farnboro indicator, developed by the British Royal Aeronautical Establishment in 1919-20 and eventually licensed, in an improved form, to Dobbie McInnes. Relying on a spark generator to mark the trace on paper stretched over a large horizontal drum, Farnboro instruments were still being touted in the 1960s.
Another idea to see success in the 1920s was the cantilever-spring indicator, a descendant of the Bachelder of 1887 patented in Germany in 1924 by Alfred Adolf von Gehlen of Hamburg and made by Maihak AG. This substituted a robust spring-steel bar for the spring, the comparatively minimal deflection under load suiting the instrument to high speed/high pressure recording. The amplifying mechanism and the pointer were essentially the same as the standard Maihaks, however.
Mechanical indicators such as the Maihak Typ 30 and Type 50 and the Dobbie McInnes Design No. 4 were still popular in the 1960s. Though their distribution declined as first electric and then electronic analysers became available, they were regularly used to indicate marine diesel engines. Indeed, the two German manufactures, Lehmag (formerly Lehmann & Michels) and Leutert (successors to Maihak) still offer external-spring and bar-spring instruments for this particular purpose.
Engine indicators were customarily supplied in sturdy boxes made of beech, mahogany, oak or walnut, though Bacharach of Pittsburgh preferred galvanised sheet-steel. These boxes were all carefully fitted to accept additional springs, scale rules, steam cocks, screwdrivers, wipers and oil bottles. It is widely believed that empty spring chambers or forlorn posts signify that something is missing, but most of the manufacturers had standard box patterns that would hold at least four springs. However, only a single spring was offered with each indicator: additional springs brought in more money. Purchasers who were interested only in the performance of one single-cylinder engine needed no more than the basic spring, but, at the other end of the scale, consulting engineers and the inspectors employed by insurance companies could require a dozen.
Jointed pantographs simplified interpretation of the diagrams by splitting them into narrow vertical strips, and spare pistons allowed an indicator to be used with low-pressure steam or high-pressure internal combustion engines at will.
Reducing wheels enabled indicators to be used with engines of widely differing strokes simply by altering the diameter of a particular pulley. Their chronology is still uncertain, but it seems likely that the earliest designs emanated from central Europe—a Bohemian inventor, Ladislav Stanek of Prague, filed a patent in Germany as early as 1878. The greatest distribution of reducing wheels was ultimately to be found in the U.S.A., where the Lippincott-patent ‘Victor’ (1897) promoted by Robertson & Sons and the ‘Faultless’ by Trill(1903) were just two of the many that may be found in the indicator boxes. Ironically, reducing wheels were uncommon in twentieth-century Europe, though German manufacturers offered increasingly sophisticated designs in the 1920s and 1930s before briefly reviving them after the Second World War. In Britain, wheels of this type are exceptionally rare—Dobbie McInnes made a few prior to 1914, but most Britons preferred to use levers and bars attached to the reciprocating parts of the engine.
Analysing pressure/time diagrams was greatly helped by the use of a polar planimeter, a mathematical instrument patented in Switzerland in 1853 by Jacob Amsler. This allowed any area bounded by a single continuous line to be quickly and accurately computed, and was an ideal, if expensive adjunct to the engine indicator. A few planimeters were even boxed with indicators and reducing gear. Among the best-known of the pre-1914 designs are the Swiss-made Amsler and Corradi, which can be found with the names of distributors or manufacturers such as Elliott Brothers or the Crosby Steam Gauge & Valve Company. Keuffel & Esser planimeters were popular in the U.S.A., though they were originally made in Germany prior to the First World War; far less common were the distinctive Lippincott, Trill and Willis patterns, and the earlier ‘Coffin Averager’.
Oscillating drum-indicators were among the most useful tools available to the steam engineer prior to the Second World War, ensuring that countless engines ran efficiently throughout the world. Yet their story is rarely told, and it is this omission that the Canadian Museum of Making collection seeks to rectify not only in words and pictures, but also by sponsoring the work of our specialist curator John Walter (www.archivingindustry.com/Indicator).
Text by John
Walter
johnwalter883@btinternet.com