The Parsons Epicyclic Engine.

Gallery opened Sept 2005

Updated: 9 July 2020
Gerd Niephaus working model added
Back to Home PageBack to The Museum

This is the Parsons Epicyclic Engine, designed by Charles Parsons before he turned his attention to steam turbines.

It was a high-speed engine designed to drive a dynamo directly. This was highly desirable as otherwise it was necessary to use belts and pulleys, which were cumbersome and subject to slipping, or gearing, which in those days was relatively crude in design, and therefore noisy and showed high frictional losses at speed.

There were two pairs of horizontally-opposed cylinders, arranged at right angles. Both cylinders and crank rotated, the cylinders at half the speed of the crank. The geometry is arranged so that connecting-rods and crossheads are not required.

Charles Parsons took out his first patent on this engine in 1877, when he was only 23 years old. This was probably Patent 2344. He took out a second patent, No 4266, in 1878.

Left: Side elevation of the Parsons engine.

This is the version that was put into limited production.

Showing two cylinders vertical and just to the left of them, two horizontal. The centrifugal governor is to the extreme right; above and to the left is a bell-crank through which the governor controlled the throttle valve.

The inlet valve assembly is at C,D and F.

J is the oil pump for pressure lubrication.

From Engineering

Left: End view of the Parsons engine.

The handwheel H varies the steam admission cutoff.

Steam exhaust to the left, and water overflow to the right.

Some important dimensions:

Cylinder diameter
5 in
Piston stroke
6 in
Oil pump plunger diameter
1.5 in
Oil pump stroke
0.75 in
Oil pump valve diameter
1 in
Oil pump valve stroke
0.75 in

There is no suggestion here that it was a compound engine. (see below)

From Engineering

Left: Exterior view of the engine.

The dynamo is coupled at the right, through a frictional shock absorber. The governor is at extreme left; just above and to the right is the steam inlet pipe and the throttle-valve housing.

The handwheel on the side varies the steam admission cutoff.

From Engineering

Above: A diagram of the Parsons engine that attempts to show how the epicyclic action allows the piston rods to move linearly without connecting rods.

I for one find it less than clear.

Left: An article about the Parsons engine that appeared in Engineering for 1st May, 1885.

Buried in the text is a brief mention of the main (and possibly only) advantage of this engine; because the crankpins move in straight lines with respect to the cylinders, there is no need for connecting rods, crossheads or guides. The reciprocating weight is much reduced, which is a good start when designing a high-speed engine. The fact that the cylinders rotated at half crankshaft speed probably also helped in attaining high output rpm.

According to Clark, the main advantage of the engine, apart from its high speed, was the very small amount of vibration it produced.

Left: Continuation of the article from Engineering.

Here are some details of the frictional shock-absorber between the engine and dynamo.

Left: Animation of the Parsons epicyclic engine.

This shows the engine rotating about the crankshaft axis.

This superb animation is kindly provided by Bill Todd

Left: Animation of the Parsons epicyclic engine.

This animation has a different frame of reference. Here the cylinder assembly is held still. It shows the crankshaft axis describing a circle, demonstrating that in reality the cylinder assembly does not just rotate, but follows an epicyclic path as its bearing is offset from the crankshaft axis. Hope that's clear...

This superb animation is kindly provided by Bill Todd

Parsons' epicyclic engine began as a model made of paper, sealing-wax and steel wire when he was an undergraduate at Cambridge University. It became reality while he was working as a premium apprentice at Armstrong's Elswick works from 1877 to 1884. He was allowed to continue developing the engine while he was at Elswick, providing he paod for any work done by the Works. (Thi swas probably something to do with being the son of an Earl) According to Appleyard, it was a compound engine, but in the Engineering drawings there is no obvious difference in the diameters of the two pairs of cylinders. It generated 10 hp, and drove a Siemens dynamo that for a time supplied an arc lamp on the Elswick jetty. According to Appleyard it ran at 7000 rpm, but this is almost certainly a misprint for 700 rpm.

Left: The Parsons epicyclic engine as shown in the second patent, No 4266.

This shows an engine that looks rather different from the production version shown at the top of this page. It is less compact, having much longer bearings at each end, and there is no fixed casing around the rotating cylinder block.

Note the drilled oilways running diagonally through the crankshaft. These are commonplace today but were new in Parsons' time.

Patent 4266

The completed engine was offered to Armstrong's, but they turned it down. It was subsequently put to work driving machinery in the millwright's shop at the "Ordnance Works" which is believed to refer to Palmer's Ordnance Works at Jarrow-on-Tyne.
It was later sent to Easton & Anderson of Erith, where Richard Clere Parsons, Charles Parsons' brother, was on the staff. They used it to drive a high-speed pump, and made some more examples themselves, apparently with good results.
A few years later, R C Parsons moved to the engineering firm of Kitson's at Leeds (see the Kitson-Still locomotive) and they produced about forty epicyclic engines of varying power output in the range 10 - 20 horsepower, for driving dynamos. Some were used for electric lighting on steamships, some generated elecricity for lighting in the Kitsons offices, and most interestingly, some were sent to the Sudan to provide electric light for railway building at night. The last mentioned "were engulfed in sand" whatever that means, and no more were bought. Info from Clark

Left: A rare, and probably unique, picture of the the Parsons engine ready for portable use.

This is almost certainly one of the lighting systems that was sent to the Sudan. The engine is mounted on a horse-drawn carriage in front of a modestly-sized locomotive type boiler. The dynamo is at the right of the engine, and the frictional shock-absorber can be seen in the shaft between them. The governor can be seen to the left of the engine, and to the left of that is something else that is apparently being driven by the engine. Possibly a boiler feed-pump?

From Clark.

The epicyclic engine was eventually eclipsed by the Willans engine for driving electric generators, and that engine in turn was superseded by the steam turbine- largely due to the work of Parsons.

The epicyclic engine should not be confused with Murray's hypocycloidal engine (In course of preparation).

Left: The Parsons Epicyclic Engine modelled by Gerd Niephaus

Gerd Niephaus has been making model steam engines since 2007. He has made a superb model of the Parsons engine; see it on his website here. There are many high-resolution pictures.

You can see the model running on compressed air on YouTube.

Charles Parsons by Rollo Appleyard; published by Constable & Co of London in 1933.
Kitsons of Leeds by Edwin Kitson Clark; pub The Locomotive Publishing Co, date unclear but around 1938.

Back to Home PageBack to The Museum EntranceTop of this page