"In the hands of a skillful engineer, the indicator is as the stethoscope of the physician, revealing the secret workings of the inner system, and detecting minute derangements in parts obscurely situated, and it also registers the power of the engine." -William Barnet LeVan
When a mechanic attaches an engine analyzer to the engine of your vehicle he may be using state-of-the-art equipment, but what he is doing has been done by mechanics for more than 200 years. The first instrument for analyzing the performance of an engine, and even recording the results on paper, was invented some time shortly before 1800. Most writers attribute the invention to James Watt, but others (Kalman DeJuhasz and The Victoria and Albert Museum) attribute it to John Southern, an engineer who worked for Watt. This instrument was named the steam engine indicator by its inventor, a name that continues to be used today. Bolton and Watt, the steam engine manufacturing company that employed Southern, realized the tremendous competitive value of owning such an instrument and consequently kept the existence of the invention so secret that they did not even attempt to get a patent on it. It appears that the secrecy surrounding Watt's indicator outlived Watt by nearly a hundred years. As late as 1900, Cecil Peabody, in his book Manual of the Steam Engine Indicator, stated that the exact form of the original indicator was not known and he thus proceeded to "consider it from the form ascribed to it by tradition." Most texts described the principle of the Watt indicator with sketches similar to Figure 1. The true shape of the indicator did not come to light until 1906 when the Science Division of the Victoria and Albert Museum in London (Now the Science Museum) published a photo of it in a catalog of their mechanical engineering collection. It appears from the catalog that the museum had not acquired the indicator until 1890. The only other photo that I found of a Watt indicator was in a 1934 text, The Engine Indicator, Its Design, Theory and Special Applications, written by DeJuhasz. This photo was of an indicator in the Science Museum in London but not the same indicator shown in the 1906 catalog. However, it appears to be the same indicator that I saw on display there in 1999. I checked with the Smithsonian Institution in Washington, D.C., to see if they had a Watt indicator, or if they knew of one in the U.S. They replied that they did not know of any Watt indicators in the U.S. but that they have a replica on display that was built in the museum in the 1920s. Figure 2 shows a Watt indicator.
The indicator simply records, on a piece of paper, the pressure in one end of a steam cylinder as the piston of the engine moves from one end of its stroke to the other and back. To illustrate this I will use one of the sketches of a Watt indicator shown in Figure 1. When the slide valve opens and admits steam to the cylinder, the pressure in the cylinder rises and causes the pencil on the top of the plunger to rise. If the paper is not moving, the pencil will simply draw a vertical line as the pressure rises and falls and as the steam is admitted, expands, and is exhausted from the cylinder. However, if the paper can be caused to move back and forth, as the piston moves back and forth and as the pressure rises and falls, the pencil will draw a diagram indicating the pressure in the cylinder at every point of the stroke. To accomplish this, a string is simply attached to some point on the crosshead of the engine and to the board to which the paper is attached so that the paper and the piston move together. Since it is impractical to move the paper the same distance as the stroke of the engine, a device known as a "reducing motion" is used so that the paper will move only four to six inches, even if the stroke of the engine is several feet. (More about reducing motions later.) The vertical movement of the pencil is limited to about two to three inches by selecting the appropriate spring. Figure 3 shows the location of the ports where such an indicator could be attached to my 45 HP J. I. Case engine.
Figure 4 shows a rather idealized diagram similar to one created by the pencil on an indicator. This diagram shows the pressure in the cylinder at various points of the stroke. If the engine is running under a full load, the vertical line at the left shows how the pressure rises in the cylinder when the inlet valve opens and the piston is at that end of its stroke. As the piston moves to the right and the valve stays open, the pressure remains at its highest along the steam line. If both the throttle valve and the governor valve are wide open, this pressure is only slightly lower than the boiler pressure because of pressure-drop in the pipes, fittings, governor and throttle valve. At the point marked "cut-off" the inlet valve closes, and no more steam is admitted to the cylinder. (If you have ever wondered why a steam engine will not start unless the crank is in certain quadrants, it is because, in two quadrants, the piston is beyond this cut-off point.) Steam engines operate most efficiently when they effectively use the curved portion of the diagram marked "Expansion." In this area, no more steam is admitted, and the engine simply extracts power from the steam as it expands. (When you "hook-up" the reverse lever on a steam engine, you are moving the cut-off point to the left so that less steam is admitted and more advantage is taken from the expansion of the steam. Later cut-off gives more power when it is needed, but you will use much more fuel to get that power.) At the point marked "release" the exhaust valve opens and releases any remaining steam. This steam is swept from the cylinder as the piston moves back toward the end of the cylinder where we began. However, before it reaches the end of the cylinder, the exhaust valve closes prior to the inlet valve opening so that some of the steam in the cylinder is compressed, cushioning the piston as it comes to a stop at about the time the inlet valve opens. This compression serves a dual purpose in that it also raises the temperature of the trapped steam and of the cylinder, thus reducing the condensation of the freshly admitted steam. The cycle now begins all over again.
It is not too hard to imagine how extremely valuable this information was to the Bolton and Watt engineers, when, for the first time, they were able to see what was happening inside of the cylinders of their engines. And it is not hard to imagine the advantage that they enjoyed while their competitors were still guessing as to how to design and set their valves.
However, in the 1820s or 1830s an Englishman by the name of John Farey, while visiting Russia, discovered an indicator attached to one of Bolton and Watt's engines and immediately recognized its value to engineers. When he returned to England, he arranged with a man named William McNaught from Glasgow, Scotland, to build indicators for him. According to DeJuhasz, McNaught not only built the indicators, he also recognized some of the shortcomings of the Watt indicator and made major improvements in the design before beginning production sometime between 1825 and 1830. These improvements were necessary because in the years since the indicator was invented, the design of steam engines had advanced to where they were using higher and higher steam pressures and were running at ever-faster speeds. The same conditions that made the Watt indicator obsolete were the factors that eventually doomed every mechanical indicator that was designed to overcome them.
McNaught's big contribution to the design of the indicator was the replacement of the flat board that held the paper, or "card" as it was called, with a rotating drum. A McNaught indicator is shown in Figure 5, and a similar indicator made about the same time by Maudslay and Field is shown in Figure 6. As the engine speeds increased, the inertia of all components of the indicator became increasingly important. The rotating drum could be made much lighter and consequently could be accelerated, decelerated and reversed much more easily than the bulky board. The rotating drum was a feature of every mechanical indicator that was built in the following century or more until optical and electrical instruments made the mechanical devices obsolete. As when McNaught's device replaced the Watt indicator because it could deal with the higher pressure and speeds, the electronic and optical devices eventually replaced the mechanical devices for the same reasons, Charles Porter reported that in 1862, "the McNaught and Hopkinson indicators were in common use in England: that one or both of these were to be found in the engine-rooms of most mills and manufacturing establishments" (This is the only reference that I found to a Hopkinson indicator.) Porter also stated that in the U.S., "The Novelty Iron Works made a very few McNaught indicators, almost the only users of which were the Navy Department and a few men like Mr. Ericsson, Mr. Stevens , Mr. Sickels, and Mr. Corliss." McNaught's indicator was capable of handling the ever-increasing speeds and pressures for some years, but it seems that the indicator began to fall into disuse on newer engines around the 1850s because of its inability to provide accurate cards at the speeds and pressures that were coming into play. (The Victoria and Albert catalog gives an early example of how increased pressures affected the design of indicators when Watt and Co. had to convert the mounting of their indicators to threaded connections rather than simple tapered plugs and sockets because the tapered fittings would tend to blow out of the mating fittings.) The biggest problem with the McNaught indicator was that, at high speeds, the spring would tend to continue to bounce up and down rather than following the pressure changes in the cylinder.
In 1860 Porter, the designer of the Porter-Allen steam engine and the Porter governor engaged Charles T. Richards to design a new indicator
that would be capable of giving accurate results on the high-speed engines that he was designing and building. The first indicator of Richards design was built by the Novelty Iron Works and on September 13, 1861, was first used on an Allen engine running at 160 RPM. A Richards indicator is shown in Figure 7. The big breakthrough on the Richards indicator was the use of a precision set of levers to magnify the movement of the pencil. With this arrangement the travel of the piston could be reduced to of the travel of the pencil. Because the piston is the heaviest moving piece in an indicator it is important to limit its travel as much as possible. The combination of sort piston travel and a much stiffer spring eliminated the bouncing of the pencil as the pressure fluctuated. One major difficulty in arriving at this design was the requirement that the pencil must travel in a straight line, not in an arc, as the pressure fluctuated. If the pencil was simply on the end of a lever, its the path would be an arc and would produce a distorted diagram. This problem was solved through the use of an arrangement called a lemniscoid movement. In Figure 7 the pencil is located behind the screw next to the "0" on the scale. The Richards indicator was the prototype of every mechanical indicator that was built, until their manufacture ceased in the twentieth century. Porter purchased Richards' rights to the indicator for $100 plus 10 percent of the future proceeds. Richards did not wish to patent the device himself, saying, "If I patent everything I think of I shall soon be in the poorhouse."
In the winter of 1862-63 Porter arranged a seven-year contract with the instrument builder, Elliott Brothers, to manufacture the new indicator, and by 1876 they had produced over 10,000. About 1870 Porter sold the patent rights, and, much to his dismay, the new owner, upon realizing that Porter had no written agreement with Richards regarding the 10% agreement, never paid Richards another dollar for his invention.
Based on the markings on my Richards indicator it appears that after the patent rights expired, Porter had the Liverpool optician and scientific instrument maker, Louis Casartelli, manufacture Richards indicators. My Richards indicator is marked "Casartelli and Porter Patent, L. Casartelli, Liverpool." I have reason to believe that this indicator was made in the 1890s.
Even though other manufacturers offered improved indicators for use on the newer, faster, engines, the Victoria and Albert catalog stated in 1907 that the Richards indicator, "except for high-speed engines, is still in general use."
In the following years a number of companies, in both the U.S. and Europe, brought forward variations on Richards' design. In 1889, William LeVan stated that, "The best forms of indicator as made and sold are commercially known as the 'Thompson,' 'Crosby' and 'Tabor.'" At about the same time, Otto Stephenson listed the Thompson indicator as the "Standard" but mentioned the McNaught's, Richards, Tabor and "others."
J. W. Thompson of Salem, Ohio, patented an indicator in 1875 that, according to Steven Roper in his Engineer's Handy-Book, published in 1892, would, "take cards at a very high speed, say three hundred revolutions per minute or even more," and at pressures as high as five hundred pounds to the square inch. One of the improvements introduced by Thompson was the change from the lemniscoid to an ellipse type of magnifying mechanism for the pencil movement. This change reduced the inertia of the mechanism, making the device more sensitive to variations in pressure. (DeJuhasz attributes this improvement to Crosby, but it appears that his patents were not issued until four years after Thompson's.) This indicator was sold by the American Steam Gauge Company as the American Thompson Improved Indicator. It is considered by many to have been the Cadillac of steam engine indicators during the heyday of steam.
In their 1896 catalog, the American Steam Gauge Company warned buyers against being misled by others who were using the Thompson name. This may have been a reference to the Schaeffer & Budenberg Company who offered an Improved Thompson indicator in their 1893 catalog and the James L. Robertson & Sons, Inc. who offered a Robertson-Thompson indicator. It appears that the American Thompson Improved indicator, the Improved Thompson Indicator, and the Robertson-Thompson Indicator may have been three almost identical devices sold by three different companies. To further complicate things, a company by the name of Thompson and Bushnell also sold indicators. Figure 8 shows the Improved-Thompson Indicator.
Porter tells in his reminisces of being visited by Harris Tabor who wished to demonstrate a new indicator that he had recently invented. The unique feature of Tabor's indicator was that it used a curved guide rather than intricate levers to create the straight-line motion for the pencil. Porter and Tabor installed the indicator on one of Porter's engines that ran at 450 RPM. Porter reported that they took a number of diagrams with this indicator and they proved to be quite free from the oscillations that were produced by the Richards indicator at the same speed. The Richards indicator was thus made obsolete by the same factors that had made Watt's and McNaught's devices obsolete.
Armed with the results from the test on Porter's engine, Tabor proceeded to Boston to meet with a manufacturer named Ashcroft. For many years thereafter, the Ashcroft Manufacturing Company of Bridgeport, Connecticut, sold the Ashcroft-Tabor indicator. Porter does not indicate the date of his meeting with Tabor. However, a friend of mine owns a very early Ashcroft-Tabor indicator (serial No. 443) with an internal spring and a December 10, 1878 patent date. This date would seem to indicate that the meeting took place in 1877 or 1878.
The top of Figure 9 shows how the Tabor indicators appeared before 1900. The lower portion of the figure shows how they appeared after the pencil guidance mechanism was modified in 1900. This indicator was also available with an external spring.
In 1879 and 1882 Mr. G. H. Crosby of Massachusetts obtained patents on an indicator that eventually rivaled the Thompson in popularity. One problem that appeared as steam pressures were increased was the effect of the higher temperatures on the springs of the indicators. In most of the early indicators the spring was enclosed in the steam cylinder and was subjected to the same temperature as the piston and cylinder. Crosby and other manufacturers overcame this problem by constructing indicators with the springs outside of the cylinders. Figure 10 shows a Crosby indicator with an external spring and a reducing motion (to be described later).
The Bachelder indicator differs from most other indicators in that the spring is a straight piece of steel rather than a coil spring. Instead of changing springs for various pressures, it is necessary only to adjust a support to the appropriate point along the spring.
The Bacharach Company of Pittsburgh manufactured a line of mechanical indicators. One of these was designed specifically for the higher temperatures, pressures, and speeds that are characteristic of internal combustion engines. This indicator uses a solid bar instead of a coiled spring, and the paper drum is much smaller in both diameter and length than on steam engine indicators. The German version of this indicator, (known there as Maihak rather than Bacharach), is shown in Figure 11. Note the massive spring #12. These were the only mechanical indicators mentioned in the 1952 book The Engine Indicator for Performance Evaluation, and may have been the last of the mechanical indicators to be manufactured.
Some lesser-known names, some of which were by European manufacturers, included:
Lehmann & Michels
Dreyer, Rosenkranz & Droop
About 1919 General Motors' Dayton Wright Division introduced an entirely different type of indicator, The Midgley Gas Engine Indicator. This instrument employed mirrors, a light, and photographic film and was specifically suited to internal combustion engines (GM, 1920). The internal combustion engine and the availability of electric power together marked the beginning of the end for both the steam engine and the mechanical indicator.
Most, if not all, indicators were packaged in very nice boxes that also contained many of the items required for their use. These items often included:
An assortment of springs for use with various steam pressures
A corresponding assortment of small wooden scales calibrated for use with each of the springs.*
An oil can, sometimes containing seal oil
Spare pencil leads
String, with special hooks for connecting and adjusting
A supply of blank cards
An instruction book
A valve to isolate the indicator from the cylinder so that the cards could be changed
*A spring marked #100 will allow the pencil to move one inch for each 100 pounds of pressure. Wooden scales marked #100 will have 100 divisions to the inch with each mark representing one PSI. If a diagram is completed with a #100 spring in the indicator, the pressure at any point on the diagram can be determined using the #100 scale. Likewise, a #50 spring will allow the pencil to move one inch for each 50 pounds of pressure and the #50 scale will have 50 divisions to the inch.
There was one other major accessory that was essential for use of an indicator, and that was known as a reducing motion. A reducing motion was used to duplicate the movement of the piston of the engine but at a greatly reduced scale. Most engines had strokes of one foot to fifteen feet or more, but indicators were capable of working only with strokes of six inches or less. On early beam engines the stroke reduction was achieved by simply connecting the string to a point on the beam close to the pivot point where the travel was only a few inches. However, on engines with cross-heads and no beam this was not possible. Devices such as pendulums, pendulums with brumbo pulleys (See Figure 12), and pantographs often had to be constructed by the mechanic to achieve the required reduction. A special, highly universal, form of pantograph was known as "Lazy Tongs". Examples of all of these can be seen in Figures 12 and 13.
Eventually, most indicator manufacturers began to offer more refined and more universal reducing motions with their indicators. An example of a reducing motion on a Crosby indicator can be seen in Figure 10. It appears that some indicators were built with special reducing motions that were designed for use on one specific engine. For example, I have a Bacharach (Maihak) indicator that was designed solely for use on an engine with a stroke of thirteen feet!
To really get the most information from an indicator card it was necessary to determine the area of the diagram in square inches. This was one of the vital pieces of information that was necessary to calculate horsepower, efficiency, steam usage, etc. This calculation could be made manually by dividing the diagram into 10 separate rectangles, using the small wooden scales to determine the area of each, and then adding them all up. A much simpler way was to use a polar planimeter, a mechanical device that would automatically calculate the area by simply tracing the outline of the diagram. Some of these devices would even calculate the average steam pressure in the cylinder. Two planimeters are shown in Figure 14. A very ingenious device for this same purpose, though probably somewhat less accurate, was the hatchet planimeter, invented by a man named Prytz in 1888 (Science Museum). This device is simply a piece of steel rod approximately 5/16" in diameter that is pointed on one end, flattened like a hatchet on the other, and bent into a "U." A hatchet planimeter is shown in Figure 15. In 1893 a polar planimeter from Schaeffer and Budenberg cost $30, but a mechanic, at almost no cost, could make a Hatchet planimeter.
Several writers relate the exciting experience of taking indicator diagrams from the engines on steam locomotives. In his book A Treatise on the Richards Steam-Engine Indicator, Porter dedicates the last chapter to "A Ride on A Buffer-Beam" on an engine of the Great Eastern Railway making a trip from London to Yarmouth and back. According to Porter, the ride was on an engine with 16 inch bore, 24 inch stroke, with wheels 7'-1" in diameter. The wheels made 237 revolutions per mile and sometimes reached 250 and even 260 RPM for a speed of 66 miles per hour. Porter said, "Comfortable seats were provided, on one of which each operator sat, quite secure, with his back to the wind, and the indicator between his knees." In this manner they were able to take one diagram per minute at low speed and one every minute and a half at high speeds. For each card, they recorded at the instant each diagram was taken, the pressure of the steam, the point of cut-off, the number of revolutions per minute, the gradient of the tracks, the curve of the tracks, and the weight of the train. He concludes by stating that, if no accident happens, there is no more danger than riding in the carriages!
McShane, in the 1899 edition of his book The Locomotive Up to Date, acknowledges some of the challenges involved in taking diagrams from locomotives:
"It is necessary that the operator should have the free use of both hands when taking diagrams. Considering the position of the indicator and the rocking motion of a locomotive when running at a high rate of speed, an enclosed platform is usually attached to the bumper beam for the operator's safety."
In the 1920 edition of the same book he makes the remarkable observation that it was possible to take four cards a minute under these conditions using an American Thompson Improved indicator similar to the Improved Thompson Indicator shown in Figure 8.
I would like to mention another hazard. The body of the indicator can be as hot as 365 degrees when an engine is running on 150-PSI steam.
According to DeJuhasz, by the 1930s this process had evolved to where all of the work could be done from the cab using a special electrically controlled Maihak indicator built specifically for locomotive use. (Just as a curiosity, I'll mention that De-Juhasz also included an example of an indicator diagram that was taken from an airplane engine while in flight.)
Even though many traction engines had their cylinders bored to accommodate the installation of an indicator (See Fig. 3), there is little evidence that indicators were ever used once an engine left the factory. There were two reasons for this.
One reason was the high cost of an indicator and the accessory equipment. The following prices are from the 1896 catalog of the American Steam Gauge Company:
American Thompson Improved Indicator ... $85
Steam Cock ... $2.75
Reducing Wheel (or Pantograph) ... $15
Pack of blank cards ... $2.50
Indicator cord ... $15
Boxwood scale ... $15
Spring ... $5
Total ... $110.55
The indicators in my collection have up to seven springs and scales included with them. Six additional springs and scales would be another $30.90.
I am not sure what a day's wages were at that time, but if they were around $5.00, an indicator (with only one spring and one boxwood scale) would have cost approximately the equivalent of a month's wages.
The second reason for the limited use of indicators on traction engines was the fact that a skilled engineer was needed to interpret the cards once they were taken from the indicator.
The mechanical indicator was not limited to improving the design and operation of steam and internal combustion engines. They were also used extensively with pumps and compressors.
The Crosby Steam Gauge and Valve Company even made one indicator specifically to assist in the design and operation of heavy artillery. In use, the pencil recorded the oil pressure in the recoil chamber as the recoil movement of the barrel rotated the paper drum. The book Elements of Ordnance, by Colonel Thomas Hayes, published in 1938, shows a Tabor indicator that was used for the same purpose.
I recently had to take an indicator diagram from a hydraulic pump that was operating at 20 strokes per minute and 10,500 PSI. Using an electronic transducer and a process-control computer I simply collected 33 pressure readings per second and fed them into an Excel spreadsheet. Using a desktop computer, I manipulated the data and printed out the "Indicator Card" on a laser printer. We have it so easy today!
I want to thank Dr. Robert T. Rhode for his valued assistance in the preparation of this article.
Contact Bruce L. Babcock at 11155 Stout Road, Amanda, Ohio 43102
American Steam Gauge Company, American Thompson Improved Indicator, 1896?
Catalog of the Mechanical Engineering Collection in the Science Division of the Victoria and Albert Museum, (now the Science Museum) South Kensington, England, 1907
Croft, Terrell, Steam Engine Principles and Practices, 1922
Crosby Steam Gauge & Valve Company Practical Instructions Relating to the Construction and Use of the Steam Engine Indicator, 1896
DeJuhasz, Kalman J, The Engine Indicator, Its Design, Theory and Special Applications. 1934
Frohwein, Paul Steam Engine Testing, 1931
General Motors, Dayton Wright Division, The Midgley Gas Engine Indicator, 1920
Hawkins, N. Hawkins' Indicator Catechism, 1901
Hayes, Colonel Thomas J., Elements of Ordnance, 1938
Hines, J. D., The Engine Indicator for Performance Evaluation, 1952
Houghtaling, William Steam Indicator and its Appliances, 1899
Le Van, William Barnet, Steam Engine and the Indicator, 1889
McShane, Charles, The Locomotive Up To Date, 1899
McShane, Charles, The Locomotive Up To Date, 1920
Low, F. R., The Steam Engine Indicator, 1910
Peabody, Cecil H., Manual of the Steam-Engine Indicator, 1900.
Porter, Charles, Engineering Reminiscences, 1908
Porter, Charles, A Treatise on the Richards Steam-Engine Indicator, 1874
Pray, Thomas, Jr., Twenty Years with the Indicator, 1902
Rankine, William John MaCquorn, The Steam Engine, 1870
Roper, Stephen Engineer's Handy-Book, 1892
Schaeffer & Budenberg Co., Illustrated Catalog, 1893
Stephenson, Otto, Engineers' Practical Test, 1896