Optical type

The ever-increasing speed of steam engines, especially those that were to drive electricity-generating sets, internal-combustion engines, emphasised the limitations of the traditional indicator. Though continual improvements had been made in the design of the reciprocating drum, the amplifying linkage and the design of the piston assembly, few indicators of this type would provide satisfactory diagrams at more than 600rpm. The increasing popularity of the internal-combustion engine in the years immediately after the end of the First World War intensified the search for more accurate recording methods.

Many attempts were made to solve particular problems, particularly when pressures rose as high as they did in the chamber of a gun, with out the attendant problems of induced vibration and natural harmonics spoiling a trace. Among the most promising of the earliest designs were the optical indicators, but light had to be excluded to provide permanent records of performance. This suited them to laboratories or colleges, but not to consulting engineers who spent most of their time on-site; consequently, many other avenues were explored.

The essence of the first optical indicator seems to have originated shortly before 1880, when a letter appeared in Engineering suggesting the use of a flexible diaphragm. The Clarke & Low indicator of 1885, therefore, consisted of a hemispherical body containing an elastic diaphragm communicating with the engine cylinder through a conventional stop cock. A small concave mirror was mounted above the diaphragm on a frame that could be oscillated about its vertical axis by a connection formed with the crosshead or reducing gear. A small link or 'finger' on the diaphragm rocked the mirror on its horizontal axis as the chamber pressure changed. Combining the movements of the mirror and the supporting frame allowed a conventional diagram to be produced. A pin-point beam of light was directed onto the mirror and reflected onto a screen, where the changes of pressure during the operating cycle could be pricked-off or traced to provide a permanent record.

The system had inherent limitations — difficulties of calibrating or regulating the diaphragm, for example, or errors produced by projecting the diagram onto a flat instead of appropriately curved surface. However, the Clarke & Low system found a short-lived favour as a teaching aid, as it projected diagrams that could be several feet long.

The optical indicator designed by Dr John Perry, who subsequently became professor of mechanics and physics at the Royal College of Science, was described to the Physical Society in 1891. Clearly inspired by the work of Clarke & Low, it relied on a thin steel diaphragm contained in a circular disc that could be expanded by the admission of steam or combustion products from a union with the engine cylinder. The influx passed through a gas-tight joint into a cast-iron box or 'shoe', and thence into the chamber beneath the diaphragm. A small mirror was fixed to the diaphragm face, offset to one side so that it was tipped in relation to the rise and fall in pressure. The shoe, mounted in gimbals, was tipped laterally by a linkage attached to the reducing gear or a suitable component of the engine. The resulting bi-directional movement of the mirror could be mimicked on a screen or photographic plate by a beam of reflected light. Leaks from the admission port, which ran through one of the gimbals, could be prevented by tightening an adjuster screw at the opposing end of the shoe. However, the Perry indicator shared the weaknesses of the Clarke & Low type, even though the development of photographic recording apparatus allowed the traces to be recorded and the instrument could react rapidly to changes of load, speed or pressure.

The simplicity of the diaphragm-type optical indicator attracted many inventors. Among them was Charles Bedell of Swarthmore, Pennsylvania, who received a US Patent in January 1897 to protect one of the first to be fitted with photographic recording apparatus, the lens, the bellows and the plate becoming an integral part of the instrument. The diaphragm was flexed by steam pressure, the central push-rod being connected with the top edge of the mirror to move the trace vertically; lateral movement was effected by a rocking lever, pivoted on top of the mirror housing, which was driven by a suitable linkage attached (ultimately) to the piston rod cross-head. Though there is no evidence that the Bedell indicator was made in quantity, and it would probably have overheated during prolonged use, it did highight one particular trend in late nineteenth-century thought.

ABOVE
Left: the diaphragm-type indicator designed by Charles Bedell, showing the integral bellows-camera attachment. From the drawings accompanying the US Patent. Right: the original Hopkinson indicator of 1906, shown here in a patdent drawing, was much more compact than the 'production version' made by Dobbie McInnes.

The first optical indicator to achieve real success was developed by Bertram Hopkinson (1874-1918) of Cambridge University, Professor of Mechanism and Applied Mechanics from 1903 until his untimely death in an air crash. Its development history is obscure, though the engineer Harry Ricardo testified that he had been given the prototype were when he left Cambridge in 1906. An indicator pictured in Engineering on 25th October 1907 is identical with the perfected "Hopkinson's Flashlight Engine Indicator" marketed in quantity by Dobbie McInnes Ltd of Glasgow from 1908 onward.

Hopkinson indicators were supplied with two differing beam-type springs and three pistons (their surface areas were customarily one unit, a half-unit and a quarter-unit), which, with cylinder liners if necessary, allowed the operator sufficient choice to suit most circumstances. They proved to be successful teaching and experimental aids, as the spring calibration was found to remain remarkably constant over long periods of time — an error less than 2 per cent being customary.

However, the requirement either to plot the trace manually or employ comparatively cumbersome photographic recorders did little to commend Hopkinson indicators to consulting engineers. Even the manufacturer had lost interest by the early 1920s, concentrating instead on the Farnboro electrical engine indicator. Dobbie McInnes may have made as many as 250 Hopkinson optical indicators, but survivors are rarely seen...even though an illustration was still appearing in the company's promotional booklet, The Engine Indicator. Its Commercial Value and Instructions for Use, in the days of Dobbie McInnes & Clyde Ltd.

ABOVE
Hopkinson optical indicator no. 155, made about 1911/12 by Dobbie McInnes & Co. Ltd of Glasgow. Virtually all of the survivors reported to date (four of them) seem to have been used by colleges and reaching establishments. Museum of Making collection.

A modification of the Hopkinson 'beam spring' system was developed in Japan by Fujio Nakanishi, eventually working in collusion with Masaharu Ito and Kikuo Kitamura. A short-stroke pistol (eventually superseded by a pressure diaphragm) conveyed movement to a light beam-like spring with an angled mirror at the mid-point of each arm. Intended to minimise the effects of vibration, this construction allowed a beam of light from a source placed vertically above one of the mirrors to pass from mirror to mirror and then back to the recording medium (e.g., a photographic plate). Like the Hopkinson design, the Nakanishi oscillated around its vertical axis to provide the 'time' part of the diagram. The Japanese indicator progressed through several improvements in 1928–35, and was certainly made in small numbers—though never as widely distributed as the original Hopkinson design had been.

In 1910, Frederick Purdy, a 'Mechanical Expert' of Kenosha, Wisconsin, patented an indicator specifically for use with multi-cylinder internal combustion engines. The essence was a series of tubes leading to a block containing the valves that allowed the pressure in any single cylinder to reach a diaphragm. A rod touching the front of the diaphragm transmitted movement to a spring-loaded mirror pivoted on the front of a bevel gear which received its motion from the engine crankshaft. As the mirror rotated, fluctuations in pressure moved the point of light generated by an arc lamp away from the perfect circle that sufficed as the atmospheric line to trace an irregular closed loop on a glass screen or a photographic plate. Though the form of this diagram was unconventional, it was nonetheless possible to interpret the change of pressure with time.

The indicator patented in Britain in June 1913 by William Dalby and William Watson was another of the many diaphragm-and-mirror designs, the thickness of the diaphragm varying from 0.015 to 0.06 inches depending on the test-pressure range. The inventors were respectively Professor of Engineering at the City and Guilds Enginering College and Assistant Professor of Physics at the Imperial College of Science and Technology in London, showing clearly that the instrument was intended more for laboratory use than a commercial proposition. And, like virtually all of its type, it was heavy and cumbersome.

The apparatus consisted of a sturdy base, which was to be directly attached to the engine, beneath a detachable camera-box. When the connection between the cylinder and the indicator cock was opened, steam pressed on the diaphragm. The diaphragm in turn acted through a light spring-loaded rod to turn a small mirror on a vertical axis, the amount of deflection being directly proportional to the pressure applied. This gave a continuous record of the changes within the cylinder in the same way as the pencil arm of a mechanical indicator. Simultaneously, a mirror set with its axis horizontally was oscillated by a cam and rod driven by a chain from the engine being tested. The ratio of the throw of the eccentric to the length of the eccentric rod had to be the same as that of the crank radius of the engine to the length of its connecting rod, to ensure that the angular displacement of the horizontal (or 'stroke') mirror was proportional to the linear displacement of the engine piston. It also provided accurate timings for the rise and fall of pressure in the engine cylinder, effectively replicating the drum motion of a conventional mechanical indicator.

The trace was created by allowing light from an external bulb to enter a pinhole in the casing of the indicator. The light-ray ran the length of the body to strike the vertical (pressure) mirror and was then turned back to strike the horizontal (time) mirror before leaving at approximately ninety degrees to its line of entry. The reflected light-ray struck a ground-glass screen or a photographic plate in a wooden box attached to the indicator body to create a visual record of the events that had occurred within the engine cylinder.

Dalby and Watson claimed novelty not only in the way in which the pressure mirror could be adjusted to change the position of the reflected light beam, but also in the adjustable stop for the diaphragm. Provision was also made for a fixed concave mirror (in the wall of the tubular housing alongside the moving mirror) to provide an additional datum line. The ease with which the pressure-mirror unit could be removed was also noteworthy, as it was only necessary to detach the camera, release a lock screw, and withdraw the tubular housing containing the mirror.

The Briton Archibald Low, one of the best-known scientific writers of his day but then an army officer serving with the 'Royal Flying Corps Experimental Works', was still promoting a diaphragm-type indicator during the First World War. Low's British Patent of January 1918 shows two methods of obtaining a movement corresponding to pressure: by variations in the resistance of carbon granules or by a coil moving within a coil, each method varying the current in an electrical circuit in which a moving-coil galvanometer provided deflection to a small mirror. This mirror supplied the vertical (pressure) component of the pressure/time trace on light-sensitive paper, the horizontal (time) component being provided by moving a second mirror driven by any suitable method—an electromagnet, for example—as the crankshaft revolved.

Low's design was in some ways ahead of its time, and may not have been employed outside the military establishment to which he had been seconded. More successful, though with a similar genesis, was an indicator created in 1919 by Leonard Thring, an engineer once associated at Cambridge University with Bertram Hopkinson. This was developed specifically to obtain pressure/time information in situations where very high pressures rose exceptionally quickly. Though this applied to high compression engines in some respects, a much more obvious example was a gun—effectively a 'one stroke' internal-combustion engine.

ABOVE
Left: the Dalby-Watson optical indicator patented in Britain in 1913. Right: the original type of Thring indicator, with traverse table, from the drawings accompanying the 1919 British Patent.

Thring proposed to harness high pressures to give a very small movement by attaching an inner tube securely to the top of an outer tube that had been screwed into a base plate. The lower end of the inner tube ended in a plunger communicating with the source of pressure, and a strut or plate extending vertically upward inside the hollow head of the plunger rocked a rigidly-mounted mirror by twisting either a flat spring or spring-loaded trunnions. When the gun was fired, pressure generated by the combustion of propellant forced the plunger upward. This had the dual effects of compressing the inner tube against its joint with the outer tube and extending the outer tube from its joint with the base plate. The resulting two-stage movement rotated the mirror, and allowed changes in pressure to be seen in a pinpoint beam of reflected light. Thring also proposed a 'traverse table' which, responding either to recoil or similar movement, rotated the mirror laterally to allow a conventional pressure/time diagram to be obtained. A later patent, sought in Britain in 1928, protected a simplified instrument in which the mirror strut was constrained at both ends to reduce the degrading effects of vibrations in the trace.

In May 1920, Engineering reported that an optical instrument designed by Professor Frederic W. Burstall of the University of Liverpool, representing 'we understand, the latest form which [the] indicator has taken', had been exhibited at the Royal Society soirée. The subject of a patent application made in Britain in October 1920, this was another of the mirror types. The two-piece body, with a diagonal joint, contained the mirrors in the upper half and the image plate in the lower part. Changes in pressure were registered by the lateral movement of a hollow-bodied piston against a bar spring terminating in a mirror, relying on a combination of a ball joint, a spring-locating screw and a 'V'-shape spring channel to ensure accuracy. A water-cooled cock and piston housing (and an optional forced lubrication system) kept the piston moving freely. The passage of time during an individual cycle was shown by a mirror, oscillated by the crankshaft or other part of an engine to divert the light path vertically, to complete the bi-directional movement required to trace a diagram on a photographic plate.

The Burstall indicator exhibited in 1920 had a water-cooled cylinder containing a tiny piston with a diameter of 0.4in and a stroke of just one-tenth of an inch. A stiff steel cantilever arm acted as a spring, reducing the effects of vibration until engines running at 2500rpm and pressures as great as 600lb/in² could be tested satisfactorily. A modified version appeared in the late 1920s, with the layout of the components refined to produce a more compact design, but the Burstall indicator was never made in quantity. Like the earlier Dalby-Watson design, it could never break free of the constraints of the laboratory.

ABOVE
Left: the original or 'straight line' type of Burstall optical indicator. The later pattern had the lamp housing offset to one side and the mirrors angled to make the design more compact. Right: the Midgely indicator, from the US Patent granted in 1923. Note the faceted mirror, part no. 63.

Thomas Midgely, Jr, of Dayton in Ohio, was a prolific designer. Among his many patents, usually assigned to the Fisk Rubber Company or the General Motors Research Corporation, was an optical indicator designed specifically to investigate the performance of high speed multi-cylinder internal combustion engines. Like some of the preceding designs. Midgely relied on two mirrors—one to show the changes in pressure by moving the trace vertically and the other to denote the length of cycle by deflecting the trace laterally. The vertical movement was undertaken by a tipping mirror controlled by a rod attached to the spring-loaded piston; and the horizontal movement was controlled by a faceted vertical mirror within the wood body, in plan virtually a quadrant, that supported the curved viewing screen. The vertical mirror was driven by a motor so controlled by the drive mechanism carried on a separate base plate that it rotated in phase with the engine crankshaft. Typically, the motor rotated at one-eighth of the engine speed; each of the eight facets of the mirror, therefore, recorded one complete revolution of the engine. Other features included the ability to alter the rotation of the mirror in relation to the engine crankshaft; to obtain a photographic record of even a single engine cycle when necessary; and to alter the instrument to become a pressure/volume recorder simply by allowing the faceted mirror to oscillate instead of rotate.

The Midgely indicator was a very sophisticated tool, and several patents of addition were filed by other employees of the General Motors Research Corporation (e.g., to J.H. Sheats and Harvey Geyer in 1923-4). These were usually concerned with adapting the indicator to other purposes: one of Geyer's designs allows an opposing pressure to be applied to the piston, and the other incorporates a spark plug to allow the piston unit to be inserted directly into the combustion chamber.

It is not known if Midgely indicators were made in quantity. One survivor dating from the early 1920s is marked 'Dayton-Wright Division' (which General Motors sold in February 1923) and numbered 'A-160', but it has been suggested that the number refers more to the laboratory inventory of the General Motors Research Corporation than a large-scale manufacturing operation. Hopefully, more of these fascinating instruments may now be found.

Among the many other designs of optical indicator were the 'Manograph' of the Frenchmen Hospitalier & Charpentier. Diaphragm-type instruments included the designs of van Dijck & Broeze of Proefstation 'Delft' (a research establishment owned by a subsidiary of the Royal Dutch Shell petroleum company), made by Kipp en Zonen of Delft, and that of la Société Genevoise d'Instruments Physiques of Geneva, Switzerland. The indicators promoted in Germany by OTA-Apparate GmbH of Frankfurt am Main and Maihak AG of Hamburg-Altona—to the designs of Otto Schulze and Alfred von Gehlen respectively—were both based on cantilever springs. The Maihak indicator, which incorporated three mirrors, was a particularly compact unit which could be substituted for a spark plug.

Yet optical indicators began to lose favour in the 1930s. The incorporation of moving mirrors and mechanical drive introduced unwanted friction and inertia, and attention turned instead to measurement that could be obtained electrically. Though these often created a conventional trace on the screen of a cathode-ray tube, they are not regarded as 'autographic'-defined fore the purposes of this site as having overtly physical characteristics.