Internal-spring type
The continuously-recording machine proposed by Daniel Gooch (1816–89) was not only the first indicator to be used successfully on a railway locomotive at speed, but also the first to incorporate an amplifying recorder: a pencil bar, attached at one end to the piston rod was mounted so that the movement of the piston was magnified approximately fourfold.
Richards, 1862
The next great advance occurred in the USA, owing to the interest shown in engine design by Charles Talbot Porter (1824–1910). The success of the Porter Governor, which improved the performance of many a steam engine from the 1850s onwards, led to an association with a mechanic named John Allen. Allen had designed what he felt was a better way of controlling a steam engine than the classical adjustable slide valve. Porter not only persuaded Allen that his design was worthwhile, but also suggested that a high-speed engine could result from a marriage of the Allen 'liberating gear' and Porter's flyball governor.
The first high-speed Porter-Allen engine was built in 1861, ran successfully, and inspired the creation of a larger machine shown at the International Exhibition in London in 1862. There it aroused great controversy, as many observers opined that high speeds posed a real threat to life and limb. The running speed was so great, indeed, that only the indicator supplied with the machine could produce useful pressure/time diagrams.
The
designer of the indicator was Charles Brinkerhoff Richards (1833–1919),
a successful Brooklyn- born consulting engineer with whom Porter had previously
had contact. Richards declined to patent the indicator, leaving the task to
Porter in return for a $100 fee and a ten per cent royalty on each instrument.
Porter subsequently obtained patents in many countries in 1862–3, including
British Patent 1450/62 and US Patent 37980. The principal claim to novelty concerned
the parallel-motion type amplifying mechanism, which ensured that the pencil
that traced the diagram moved in a straight line.
The Richards Indicator amazed onlookers in the Crystal Palace by recording perfect diagrams at an unprecedented 150rpm. No other indicator of the day would have coped in such circumstances, owing to the inertia of their heavy moving parts and the effects of vibration on their feeble springs. The Record of the International Exhibition, 1862 described the improvements as the substitution of "very light moving parts, attached to a short, and therefore stiff spring, instead of the comparatively heavy moving parts, attached to a long, weak, and therefore tremulous spring. The motion is multiplied by a lever of the third order, and a parallel motion is employed to guide the recoil spring in a straight line... It is very little more complex than the [common] M'Naught instrument". It attracted so much attention that it was used to test all but two of the British-made engines that were present.
Porter stayed in England after the 1862 International Exhibition finished, keen to promote not just the Porter-Allen engine but also the Richards indicator. His experiences with the engine were unhappy, as the initial contractor (Oswald, Grierson & Company of Manchester) failed in 1866 and a subsequent liaison with the autocratic Joseph Whitworth was equally unproductive. Conversely, the indicator was successfully licensed to in the winter of 1862/3 to Elliott Brothers of London.
Elliott-Richards
indicators had improved recording units, with the parallel-motion components
made in the form of two thin bars instead of a single flimsy rod. A better method
of connecting the piston rod to the parallel motion with a swivel is usually
attributed to an employee of Elliott, Edward Darke, and the provision of exchangeable
springs was also a helpful novelty: in 1874, the range of springs extended from
15lb/sq.in to 175lb/sq.in. To allow such flexibility in a single instrument,
Elliott instruments had a piston-area of 1/2sq.in, double that of the Richards
prototype.
Elliott
Brothers made 'Improved Richards' indicators into the early years of the twentieth
century. Though production was slowed by the introduction of better designs
(Elliott also made Simplex and Wayne indicators), at least 23,000 Richards-type
instruments had been made in a variety of sizes. Most of the 'Improved Richards'
indicators had detents of a type patented in 1875 by Edward Darke, which allowed
the drum to be stopped without disconnecting the driving cord.
ABOVE
The perfected post-1875 Elliott-made Richards
indicator, fitted with a Darke-patent detent. This particular example dates
from about 1890. Note the full complement of six springs, three of which match
the surviving rulers. There is little doubt that this instrument was the property
of a consulting engineer or an insurance inspector. Museum of Making collection.
The Richards Indicator required careful manufacture and accurate assembly to work efficiently and was also surprisingly expensive. However, as steam pressures and engine speeds rose dramatically in the 1860s and 1870s, so the need for its sophisticated analytical capabilities grew commensurately. Richards-type indicators were still being advertised by many suppliers of steam accessories (e.g., W.F. Stanley & Co. Ltd of London) in the 1920s. For slow-speed installations, they were as reliable as any of the later, lighter and usually more delicate designs, and were eminently suited to the comparatively ill-trained engineers to whom many manufactories, waterworks and similar installations entrusted them. They were also commonly used at sea.
Charles
Richards made one final, if short-lived contribution to the design of the indicator.
A pantograph-type instrument was apparently made for exhibition in Paris in
1878 on the general lines of an Elliott-Richards instrument, but in steel instead
of brass. The new amplifying mechanism was inspired, no doubt, by the emergence
in the USA of the Thompson and Tabor indicators.
This particular instrument is still in the museum collection in Paris, inventory no. 09416. It was donated after the exhibition had ended and is marked appropriately 'Presented/to the/Conservatoire/des/Arts et Métiers' and 'C.B. Richards' over '1878' on the body. The amplifying mechanism is a simple pantograph system comprising a link, pivoted on the serpentine standard protruding from the body, which is connected at its tip with the elongated pencil arm. A link from the top of the piston rod to the pencil arm is connected to the rear link, at approximately its mid-point, by a short double-pivoting bar. This maintains a proper straight-line motion, but the undoubted efficiency of the design was never matched by commercial exploitation; indeed, perhaps only the prototype was ever made. Jacques Buchetti, in his book Guide pour l'Essai des Machines à Vapeur et la Production Économique de la Vapeur (1887?), gives a diagram.
RICHARDS
COPIES. Many Richards-type indicators were made prior to the First World
War, most of them being advertised as 'improved'. The principal alterations
concerned the drum spring, which was often changed from a coil in the drum-base
to a coil around a vertical post, and in the design of the piston. Novelty was
sometimes introduced into the cord pulley or the indicating-pointer linkage,
though the basis of the instrument almost always remained unchanged.
Classifying
these instruments is complicated by the involvement of a variety of manufacturers
in the USA, Britain and Europe. These included Casartelli of Market Street,
Manchester, who made sold standard Richards-style indicators in Britain before
proceeding to a modified version with a lighter double-bar parallel motion system
'somewhat like the pentagraph'; Hannan and Hannan & Buchanan made them in
Glasgow; and Martin-Garnier of Paris and Duvergier of Lyon made them in France.
Many of the Duvergier-made examples had a patented integral reducing wheel,
which was fitted to a bracket beneath the drum-support part of the frame.
The status of the Richards indicator in the USA has often been questioned, and it is widely believed that the initial needs were satisfied with Elliott-made instruments imported from Britain—indeed, several indicators of this type, with low serial numbers suggesting production in 1863–4, have been found in the USA. It is now known that the American Steam Gauge Company of Boston, Massachusetts, made Richards indicators before progressing to the better-known Thompson design. One of these Richards instruments (pictured below) is owned by the Knox County Historical Society of Mount Vernon, Ohio, and may be seen on display in the local museum. It appears to be numbered '599' and also bears an acknowledgement of the design—C. B. RICHARDS’ PAT. MAR. 24, 63 in two lines on the rear link bracket. This suggests that it dates prior to the expiry of the patent, placing manufacture in the mid 1870s.
Though
Richards-style instruments have been identified as the work of Rosenkranz of
Hannover, the principal German manufacturer was Schaeffer & Budenberg of
Buckau bei Magdeburg — though production (possibly never excessive) had
ceased in favour of Thompson indicators by the late 1870s.
Thompson,
1875
The Richards indicator reigned supreme until the master patent expired in the mid 1870s, allowing a stampede to develop a better design. Though the Richards-type instruments were robust and reliable, tests often showed that oscillations began to affect the accuracy of its diagrams above 200 rpm, and rival promoters considered that many details of the basic pattern—e.g., the pointer linkage—could be improved to reduce both friction and inertia. This left the field open to speculation, and a variety of 'improved' designs appeared.
The first improvement is credited to an American, Joseph W. Thompson of Salem, Ohio, whose US Patent was granted in August 1875. The Watt-type or lemniscoid parallel motion used by Richards did not entirely ensure that the pen moved in a straight line. Once the central range of the lemniscoid curve had been exceeded, the pen began to trace an arc, and although the error was comparatively small, reckoned to be about 0.8 per cent for a standard 'large' example, many reasoned that it was avoidable. The Thompson system made use of the mathematically superior "Evans' Parallelogram", also known as the 'Scott Russell', 'Grasshopper' or ellipsoid method.
The instruments are readily identified by the design of the recording mechanism, as the serpentine linkage and robust supports of the Richards design were replaced by much lighter components taking the general form of the letter 'M'. The piston-rod extension often incorporated a universal joint, not only to allow proper movement of the parallel motion but also to combat wear. The piston-rod extension pushed upward on the pointer bar, which was attached to one link attached to a fixed post and another pivoted on the cylinder body.
 |
LEFT |
Thompson
indicators generally performed better with high-speed engines than the Richards
pattern. The first instruments were probably made on behalf of the assignee
of the patent, the Buckeye Engine Company of Salem, Ohio, by the American Steam
Gauge Company. Later, once licenses had been obtained or after the patent had
expired, Thompson-type indicators were made by Schaeffer & Budenberg of
Buckau bei Magdeburg, Rosenkranz of Hannover in Germany and Kraft & Sohn
of Vienna in Austria-Hungary; Victor Lefebvre in France; and Hall-Brown, Buttery
& Company of Glasgow, Scotland.
Changes
made as a result of intensive scrutiny, largely undertaken in Germany in the
1880s, led to the 'Improved Thompson' instruments made by the American Steam
Gauge Company, Schaeffer & Budenberg, and Dreyer, Rosenkranz & Droop.
Essentially similar examples were made by Star, Lippincott and others.
The indicator designed by Ebenezer Hall-Brown of Hartlepool, patented in Britain in 1889, was among the most interesting Thompson derivatives. Though the distinctive linkage was retained, Hall-Brown devised a method of detaching the pointer mechanism, cylinder cap and piston/spring assembly simply by aligning the bayonet joint and a supplementary locking collar running around the cylinder body. The spring in this design is exceptionally easy to change, but the Hall-Brown system does not seem to have survived into the twentieth century; the readily detachable cylinder cap would have contributed to accidents if the engineer forgot to rotate the collar back to its locked position before opening the steam cock. However, a substantial number of indicators were made by Hall-Brown, Buttery & Company of Glasgow. There is no doubt that they were intended for use at sea, something the original indicator cards made clear.
Another
of the interesting variations on the Thompson theme was that of Joseph Bachelder
of Manchester, New York, who received a US Patent in 1887 to protect a Thompson-type
indicator linkage working in conjunction with a unique adjustable-pressure riband
spring placed horizontally beneath the drum. This was intended to avoid changing
springs to suit differing engine characteristics and exploited the ease with
which the effective length of the spring could be changed to provide an 'adjustable'
indicator. The Bachelder instruments were made in small quantities by Thompson
& Bushnell of New York City, but may not have been successful; they are
now exceptionally rare.
Other
American-made indicators included the Buffalo; the Calkins, perhaps made only
in small numbers; the Robertson-Thompson, which became the Trill early in the
twentieth century; the Lippincott; the Lyne; and the Star. (See 'Sources of
information')
The most popular variation of the Thompson linkage to be found in Britain in 1900 was based on a patent granted in November 1887 to Thomas McInnes (or M'Innes) of Glasgow. An instrument maker by profession, McInnes eschewed the minimalist approach and accepted that his new indicator would inevitably contain more parts than some of his better-established rivals. McInnes sought to reduce friction whilst simultaneously improving the response of the pointer linkage to changing pressures, but the greatest claim to novelty, however, lay in the adoption of vulcanite sheathing to protect the operator's hand from heat absorbed by the piston-cylinder body.
The earliest instruments may have been made by a short-lived partnership of McInnes & Cairns (allegedly in Edinburgh). By 1889, however, McInnes was trading as 'T.S. McInnes & Co.' from 56 Waterloo Street, Glasgow. A limited-liability company had been formed by 1894, manufacturing facilities moving from Waterloo Street to Clyde Place.
These
first-pattern McInnes indicators were made until the end of the nineteenth century.
In May 1898, however, a British Patent had been granted to John Clark Dobbie,
a partner in A. Dobbie & Son. Dobbie was making clocks and watches in premises
virtually adjoining McInnes' premises in Clyde Place, and it was a logical step
for McInnes to make the new Dobbie-patent indicator.
ABOVE
McInnes-patent 'Large Size' indicator no. 637,
made by 'T.S. McInnes & Co.' of Glasgow, c. 1893. Note the trademark
moulded into the vulcanite sheath of the threaded union (at the bottom of the
body) that couples the indicator to the steam cock. The label pasted in the
lid of the case replaced the original at some time after 1928. Museum of
Making collection.
The
new instrument offered an external-spring system and a modified mechanism with
the pointer directly above the supporting bar instead of alongside. These changes
were incorporated on the McInnes-type closed-cylinder indicators made after
c. 1900, though production was slow. All Dobbie-patent instruments had 'D'-prefix
serial numbers, and most post-1903 examples bore the marks of Dobbie McInnes
Ltd. Perfected McInnes-Dobbie internal spring indicators were still being made
in the 1920s, but were eventually superseded by efficient outside-spring types
patented in 1898-9.
Thompson-type indicators were also made by Whyte, Thomson & Company of Glasgow. Their history is remains obscure; though their existence is mentioned in several pre-1914 sources, only one could be located for examination—'Model I' no. 305, which is somewhat similar to the contemporaneous McInnes pattern. However, comparing the design and construction of individual components reveals a total lack of similarity, and any theories that Whyte, Thomson indicators were simply McInnes or Dobbie McInnes instruments being marketed under another name can be discounted.
Schaeffer
& Budenberg made several versions of the Thompson indicator, with internal
and external springs. There was also an instrument that allowed a simultaneous
combined diagram to be taken from a double action engine. It was otherwise necessary
to use two indicators, or throw a three-way cock to indicate first one side
of the piston and then the other—but not on the same stroke.
Tabor,
1878
The indicator patented in the USA by Harris Tabor in 1878 was the result of a prototype shown to Charles Porter that could run efficiently at 450rpm. This had interested the Ashcroft Mfg Co., which was soon making Tabor indicators in quantity. The first few hundred of these, despite the statements that have been made to the contrary, were made in close accord with the patent drawings. The essence of these Tabors — which were being made in quantity by 1883—is a small cam-plate elevator attached directly to the piston-rod extension. In 1886, however, after at least 413 of the first-type indicators had been made, Ashcroft replaced the flimsy elevator with the more familiar amplifying mechanism: a sturdy standard, attached to the cylinder cap, containing a curved track to ensure that the pencil arm moved vertically. Links pivoted at both ends joined the pointer bar to the piston-rod extension and the cylinder body. Series production of the new-pattern Tabor indicator began in 1887, and continued for many years.
 |
ABOVE
A boxed Tabor indicator, no. 1951 dating from
1889 or 1890. Note the design of the guide pulley for the drum cord; later examples
replaced the knurled fairlead retainer with a large hexagonal-headed nut. This
particular set is lacking the steam cocks. John Walter collection.
Tests
conducted by the U.S. Navy in Brooklyn Navy Yard, in 1886, with Richards, Thompson,
Tabor and other indicators, revealed that the moving parts of the Tabor instrument
were thirty per cent lighter than the nearest rival and had 24·3 per
cent less friction. One
of the most popular systems of indication in the 1890s, the basic Tabor pattern
was extensively marketed in Europe. Among the British distributors were the
Globe Engineering Works and John Musgrave & Sons Ltd of Bolton, who began
work in 1889. Engravings accompanying a short descriptive article published
in Engineering in June 1889 depict an Ashcroft-made indicator no. 915,
but the printing blocks may have been several years old.
Darke, 1879
The
simplicity of the Tabor system influenced Edward Darke of London, whose minimalist
design was patented in Britain in 1879. Made exclusively by Elliott Brothers
in 'small pattern', the Darke Patent High Speed Indicator had only one fixed
pivot and a swivelling link connecting the piston-rod extension to the pointer
bar. The role of the other fixed pivot was taken by a vertical slot cut in a
plate on the cylinder body.
The pointer was allowed to slide back along its bar as a guide pin rose in the
slot. However, though this was claimed to promote accuracy—the Darke and
a properly regulated Tabor were among the few indicators to be mathematically
perfect—it was also susceptible to excessive friction in the slider unit.
Darke claimed the improvements in his system as 'an invention for giving motion
to the paper drum by means of rods, &c, in rigid communication with the
engine, with means of stopping and start-ing the paper drum in a very simple
and effective way': a sprung pawl-type detent (patented in 1874) could be activated
at will by engaging a spring-steel plate that slid horizontally around the cylinder
body. This allowed the operator to stop the movement of the drum whilst the
indicator card was removed.
 |
ABOVE
A typical small-size Darke indicator, made by Elliott Bros.of London. This particular
example, no. 565, dates from the mid 1880s. Many rival promoters were scathing
about the accuracy of the Darke mechanism; though defensible mathematically,
there is little doubt that there was far too much friction between the slide
and its track. Museum of Making collection.
Darke-patent
indicators generally had a coil-type spring in the base of the drum, freeing
the interior for a continuous roll of paper. The paper clip was often a hinged
two-bar pattern instead of the customary Richards spring fingers. Instruments
of this type were made in large numbers, but had been overtaken by better designs
by 1900.
Crosby, 1882
Very few indicators have embodied a pantograph amplifying mechanism, excepting a prototype exhibited in Paris in 1878 by Charles Richards and the Simplex, made in quantity in the late 1890s by Elliott Bros. Ltd of London. The only other pantograph-like indicator was the Crosby, though sufficient change had been made to the geometry of the links to compromise the mathematical certainty of 'straight line' rivals. Perhaps the most successful of the late nineteenth-century designs, made in Boston, Massachusetts, by the Crosby Steam Gage & Valve Company, the Crosby Indicator was the subject of patents granted in the USA in 1879, 1882 and 1885. The perfected pencil motion, an ultra-light pseudo-pantograph, was just one of several alternatives proposed in the master patent of 1882. This mechanism bore some resemblance to the 1875-vintage Thompson type, though the short anchor link was pivoted to the piston-rod extension link instead of the pointer arm.
The Crosby indicator was exceptionally successful, particularly after changes in the geometry of the links were made after intensive scrutiny by experimenters in Germany suggested that the original design was not as efficient as it should be. Post-1895 Crosby indicators, therefore, have a sturdier straight rear link than their predecessors, which relied on a very slender link with a noticeable curve.
ABOVE
A matched pair of Crosby enclosed-spring indicators,
one left-hand and one right-hand (note the positions of the trace arm). These
instruments are believed to have been the property of the Municipal Technical
College in Brighton (England), opened in 1897, but were already several years
old and may have come from School of Art & Science. They represent the first
type of Crosby, with a curved rear link in the amplifying mechanism. The quality
of the nickel plating is extremely good, and the blueing of the steel links
is exemplary.
The
perfected Crosby Indicator may have been the first to incorporate a spiral rotating
spring within the drum body, which was soon accepted as an improvement over
the conventional coil spring within the drum base. This feature was widely copied
once Crosby's patents lapsed. Crosby was also known for the counter-wound spring,
but was responsible for a variety of other relevant innovations.
The Crosby instrument was very popular, allowing its manufacturer to maintain
offices in Boston, New York and Chicago. A branch was also operated for many
years at 147 Queen Victoria Street, London. These indicators were made in several
styles, and were often supplied in a single box as matched pairs, triples or
even quadruples. Crosby-type linkages were used by a variety of external-spring
rivals, including Maihak in Germany, and are still evident on the Leutert instruments
being made today.
Casartelli, c. 1883
Another
Tabor-like design was the Casartelli High-Speed Steam Engine Indicator, made
in Manchester, which was exhibited at the Engineering and Metal Trades Exhibition
staged in London in 1883. This offered a vertically slotted plate to guide the
pointer, but had an additional moving pivot. The cylinder body had a screwed
joint at its mid-point, while the piston rod (in the form of a tube) was attached
to the piston with a universal joint. The total weight of the working parts
was claimed to be less than one ounce, which may explain why indicators of this
type are rarely seen; they may well have been too lightly built to withstand
hard use.