Also made by Dobbie McInnes, and can be used on slow-speed diesels as well.  The area under the graph is the power developed per stroke.

On 12 March 2024 at 22:11 Nigel Graham 2 Said:

How slow, slow-speed? Large marine engines?

The example I used was on a single cylinder horizontal gas engine running at a few hundred RPM (from a very dim memory). It wasn’t the same make as the photos above.

There is lots of information on indicators on Google etc. if one searches for “engine indicator” or “steam engine indicator, there are even manuals for these instruments on the web.

My understanding is that they were used to measure power output for most sizes and speeds of steam engine and also early IC engines, but that the inertia of the reciprocating parts of the indicator made indicators unsuitable for use on the high speed IC engines developed later.

A ” planimeter”, mentioned in one of the posts above, is an instrument for measuring the area of an irregular shaped “thing” on a drawing or map.

Rod

Mechanical cylinder pressure indicators cannot cope with high engine speeds.

For high speeds, (above 500 rpm) electronic means have to be used.

At Rolls Royce, in the late 1950s, we used a capacitive method, feeding into a FM unit for a display on a Cossor oscilloscope. Balancing the pressure on a brass disc until it ceased to move, and then the pressure was read off a calibrated pressure gauge.

Crank angle was measured using a Tufnol disc, with small magnets inset at 5 degree intervals,fixed to the flywheel.  tdc was indicated by the magnet being inserted in the opposite sense all the others.

A pressure transducer incorporating a calibrated strain gauge, on a disc, could provide the means of pressure sensing.

Howard

Let us see once again what Grundy (1942) [op.cit.] tells us…

The early steam-engine indicators used a piston balanced by a spring but as the speeds of steam-engines for driving plant like dynamos, then i.c. engines, increased such indicators were soon out of their ability.

Reducing the piston size and stroke, and using fully-rotating rather than oscillating indicator-card drums, reduced inertia problems but the limit was around 1000rpm.

The first example of replacing the piston with a spring-steel diaphragm was probably the Clarke & Lowe Optical Indicator described in Engineering, July 10, 1885. A ray of light reflected from the surface of the diaphragm created the trace photographically.

These cope with higher speeds but can have an inherent problem of the diaphragm’s own resonance.

It was soon found that a large passage from cylinder to diaphragm affects the compression-ratio of an i.c. engine, making the test inaccurate. That was solved by making the diaphragm unit fit as a screwed plug, like a sparking-plug or injector.

The next step was an electrical transducer: piezo-electric (a quartz in Grundy’s day – it would be an artificial ceramic material now), electro-magnetic coupling (as in an electric guitar), capacitance variation (as Howard states) or by resistance variation in carbon discs (analogous to a carbon microphone).

The varying voltage with pressure was fed to the Y-axis of an oscilloscope whose X-axis modelled the crank-angle by a timing unit.

These systems became quite sophisticated, as in the ‘Standard Cathode Ray Engine Indicator’ produced by Standard telephones & Cables, Ltd., of London. Made for testing c.i.-engines, this detected the fuel-valve movement as well as the gas pressure; the latter via rate of change.

….

The “Farnborough”… JA’s friend had bought….

Dr. Grundy tells us all about the “Farnboro” [sic] Electric Indicator, made by Dobbie McInnes Ltd. of Galsgo – perhaps more familiar to most of us for their piston-type steam-engine indicators.

This, Dr. Grundy’s book says in a footnote on p.687, is also known as the “R.A.E. Electric Indicator”. I.e. Royal Aircraft Establishment, Farnborough.

This large, complex machine used a spark-point to record a diagram approximately 12″ long by up to 6″ high, on a drum revolved at half or actual engine speed. The author is not clear on this, but I assume the trace was on a (sensitised?) paper pro-forma wrapped round the drum.

The transducer is a water-cooled disc-valve, or diaphragm, in a unit screwed into a test-point in the cylinder. It balances the engine’s internal pressure against an applied pressure, and it forms the contact-breaker in the recording-spark circuit.

The valve’s travel is <0.01″ , between two seats that are the circuit’s earth-point, so it moves only at the instant the cylinder and balance pressures pass each other. The balance supply, whose maximum pressure has to exceed the engine’s, was compressed air or for better unit life, nitrogen.

Since this equal point occurs twice in each engine cycle, a set of pressure-points w.r.t. to time over the whole cycle, is obtained in steps by adjusting the balance pressure. As this records each point pair over very many cycles, it averages the normal slight variations between cycles. The time difference from step to step is given by the spark-point being directly related to the crankshaft’s, hence drum’s, angle of rotation.

The pressure-displacement travel of the indicator’s tracing point was achieved by a separate assembly, a spring-balanced piston gauging the balance-gas pressure.

So the system relayed sets of many averages from the electro-mechanical transducer to the recorder itself. Rather than the recorder reading the engine pressure directly; although the conventional indicator will also produce an average as the mid-line of a slightly blurred trace.

The diagram shows a 3-cylinder engine but presumably Dobbie-McInnes could supply these sets with additional transducers and associated, extra pipes and electrical fittings.

The “Farnboro” Indicator could also test the induction and exhaust pressures, balanced against a depression drawn by a hand-pump.

Obviously all these systems rely on either the transducer unit also holding the spark-pug or injector, or a separate test-point in the cylinder-head, normally fully filled by a plug.

………….

The idea of electrical-discharge recording on sensitised paper was still current (oh dear) in the 1980s.

At the time I worked for a local electronics company that designed and sold its own make of Sidescan Sonar for sea-bed imaging. It formed the image on a continual web of special paper fed past a revolving drum carrying a pair of single-turn helical electrodes rather like model lawnmower blades, one each port and starboard.

Each electrode passed a current through the paper to a return platen, inversely proportional to the echo from each point in the swathe of ultrasonic pulses insonnifying the sea-bed, to create a false-colour line picture resembling an old TV image in fetching purplish-brown. The stronger the return the lighter the image, to create the contrast between sediment, objects and their acoustic shadows.

I helped only assemble the recorders but I would guess the system created its own time-base to follow the delay between near and far echoes. (Speed of sound in sea-water averages 1500m/s.) I don’t know if there was also a distance-proportional amplifier to compensate for the range attenuation.

You had to be careful to store the plot properly because the paper was also slowly light-sensitive.

I do not know if the ‘Farnboro’ also used similar paper. Even the book’s illustration does not tell us.

Thankyou for that, Duncan.

A comprehensive history, and I had never thought they were still made in the early 1970s!

.

That company for whom I worked was Waverley Electronics, and most of its work was contract so the Sidescan Sonar was quite a departure.

The description of paper, oscilloscope and photographic recording of indicator-diagrams reminds me of one unusual contract WE fulfilled, though I think the customer made most of it in his laboratories.

The customer needed record unwanted, irregular, occasional, very short transients, I think in HV electricity transmission equipment, to help trace their source.

So between the two companies we made a “light-funnel” that was a truncated brass pyramid, nickel or chrome-plated internally, completely filled with a polished acrylic cast. The large end was against an oscilloscope screen, the small end against a camera; the assembly was presumably sealed against stray light. I don’t know if the shutter was open all the time or triggered by the transient, but the screen’s remanence would have given time for the photograph.

I gather it worked very well!

I am sure the Farnborough indicator did not use photosensitive paper. It probably used an electric spark since this was a proven and robbust technology.

I first met UV photosensitive paper a couple of years later at tech on a project concerning water hammers. Later in life UV traces were used almost daily. The paper came in well made heavy cardboard boxes ideal for storing bits in the workshop. Although UV paper disappeared about 30 years ago I still have some boxes.

JA

This has drifted well off the original topic, one of the joys of this forum. It occurs to me that one could have a modernised version of the Farnborough Indicator by having one pressure sensor on the engine cylinder and another on the reference pressure, push their outputs into a comparator and when the difference changes sign log crank angle and reference pressure. Why? Well you only get 2 readings per cycle, so plenty of time for logging the results in between. You could probably do away with the reference pressure by having a slowly declining voltage calibrated against the engine cylinder sensor output

Once again, nothing to do with indicators, but in the very early 1980s our company moved away distant, bulky, line printers to stand alone printers local to the teleprinters used to access the mainframe. These printers used photosensitive paper that came as rolls and we used a lot of it. I believe silver was the photosensitive base. Once again two rolls came in a nice, very valuable, stiff cardboard box. The paper became unavailable when the Falklands war started. I guess they all went with the RN to the South Atlantic.

The printers disappeared soon afterwards along with the teleprinters and the old mainframe computers.