Rotary Steam Engines: Page 10

Updated: 16 Mar 2015

'A Thought: Rotary Steam Engines & Reciprocating Electrical Engines' added
More on 'The Market For Rotary Engines'
Back to Home PageBack to The Museum EntranceBack to The Rotary Steam Engine Wing

As noted in several places above, there are great similarities between rotary compressors (which work and are very common) and rotary steam engines. (which did not work and are vanishingly rare) It is a little puzzling why this should be so. Single stage rotary compressors can deal with significant pressures- typically up to 80 psi, so that's not the problem. Could it be the steam temperature? It seems unlikely, as compressing air releases large amounts of heat, and cooling is often an issue.
I suspect the answer may be that inefficiencies in compressors are more acceptable than in engines; compressing air is inherently an inefficient way of converting energy because of the heat loss mentioned above.

There are such things as rotary air motors: see (external link)

While there may seem to be a confusing number of types of rotary engine, they can be divided into relatively few categories. Here is my attempt:

Please note that this section is in course of arrangement and does not currently include all the rotary steam engines in the gallery.

  • Cooke 1787
  • Chapman 1810
  • Poole 1817
  • Wright 1825
  • Boardman 1899
  • Holmes 1860


  • Cartwright 1797
  • Hodson 1882
  • Brown 1912

    SLIDING DOORS (to let rotor pass)

  • Bramah & Dickenson 1790
  • Rumsey 1791
  • Turner 1816
  • Galloway 1834
  • Taylor 1899
  • Hardy 1859

    REVOLVING DOORS (to let rotor pass)

  • Flint 1805
  • Willcox 1806
  • Eve 1825
  • The Berrenberg Engine
  • Ritter Engine
  • Fearnley 1903 (Model Engineer)
  • Norton 1866


  • Murdoch 1799
  • The Pillner-Hill Engine
  • Birdsall Holly
  • La France
  • Keats 1909

  • Trotter 1805
  • Beale 184?
  • Lamb 1842
  • Hyatt 184?
  • Borrie 1844
  • Elijah Galloway 1846
  • Simpson & Shipton 1,2 1848
  • Simpson & Shipton (second Engine: 1851
  • Napier 1851
  • Dundonald's, Types 1-4
  • The Franchot Engine
  • Smith?
  • The Fletcher Engine
  • The Bartrum & Powell Engine
  • The Forrester Engine
  • The Soulé Engine: 1893
  • The Hay & Depuy Engine: 1899
  • The Toennes Engine: 1899
  • The Cooley Engine: 1901
  • The McEwan Ross Engine: 1912
  • The Incredible Hult: A Successful Rotary? 1889
  • The Gerard Engine: 1892
  • The Arbell-Tihon Engine: 1893
  • Jones 1841
  • Yule


  • Hornblower 1798
  • Stocker
  • Parsons
  • Walter: 1842


  • Kemp 1901
  • The Filtz Engine: 1897


  • Behrens

    I am no steam engineer (as will no doubt become apparent) but let's see if we can work out why the problems were so intractable. The fact that today there is virtually no market for such a device will be ignored for the time being.
    The intriguing thing about this particular technological dead-end is that at first sight there seems to be nothing inherent in the concept that makes it so completely impractical. And yet, literally hundreds of engineers and inventors applied their intelligence to the problem, but none were able to make it work acceptably.

    Contemporary comment makes it clear that sealing was a major problem; foreshadowing the later sealing problems with the Wankel. Sealing at both the rotor peripheries and the ends was difficult; anything approaching steam tightness meant excessive friction, and either way the machine was inefficient.
    Inevitably the thought arises that given modern seal materials and technology, it ought to be possible to make it work. I am no expert on rotary IC engines either but I understand that the sealing problems of the Wankel IC engine have finally been solved, though the low efficiency due to the long thin combustion volume remains an inherent drawback. Surely keeping a rotary machine steam-tight should be much easier, given that the steam will be both cooler and at a lower pressure?

    In many cases the rotary inventors seem to have been relying purely on close tolerances and accurate machining to control steam leakage. This is fine in principle, and when appropriately applied (eg the non-contact labyrinth seals used in steam turbines) but much less effective in the topology of a rotary engine, especially when friction and thermal expansion are considered.
    It seems inevitable that the inner rotor will get hotter than the external casing, so a rotor that is a snug fit when the machine has just been started will get tighter and tighter as it warms up. A further complication is that unequal heating and thus expansion of each part will result in changes of shape as well as size that will make attempts at steam-tightness through tight clearances quite hopeless.
    Few of the drawings above give any clue as to how the ends of the rotors were to be sealed against the side of the casing; it is not just a matter of making clearances small. The thermal expansion of rotor and casing are unlikely to be the same in all directions, and in any case it is equally unlikely that the various components will reach the same temperature at the same time, especially under varying loads.

    Contemporary comment also frequently refers to excessive frictional losses due to packing being over-tightened in attempts to obtain acceptable sealing.

    Another problem was that many of the designs seem to make no provision for the expansive use of steam, which would have made them very inefficient compared with normal reciprocating engines, even if the sealing had been frictionless and perfect: expensive rather than expansive. Likewise, many designers seem to have had no notion of getting steam into and out of the working chamber easily- the steam passages were often narrow and convoluted.

    Now consider the reciprocating engine and its piston. ("piston" is derived from the French for "pestle"- beautifully descriptive!) There is movement in one direction only, and the whole situation is mechanically far simpler. Thermal expansion is dealt with by piston rings that press against the bore due to their own elasticity. (and valve sealing does not seem to be a problem: cf using a compression tester on an IC engine)

    The essence of the situation is that sealing a reciprocating piston is a one-dimensional problem. With a cylindrical piston the sealing situation is the same all the way round, as it were, and the sliding movement does not alter the geometry. In contrast, the rotary engine presents a three-dimensional sealing problem, usually continuously altering in its geometry. This is quite a different matter.

    This does does rather beg the question of what might present a two-dimensional sealing problem; I suppose it would be something like a square reciprocating piston. I imagine that would present interesting sealing problems at the corners.

    Well, essentially there isn't one. In the relatively few places where a small steam-engine is required, a steam turbine, usually geared down to its load, will be more efficient, more reliable, and easier to maintain. One example is on board a turbine-propelled ship, where there is plently of steam available but electricity is in relatively short supply. Turbines were often used to drive the fans that pressurised the boiler-room, replacing the small reciprocating engines that first took on this role. On land, no-one today would dream of installing a steam boiler without electrical controls. And if mains electricity is available and already essential for operation, any minor requirements for power, such as driving pumps, might as well be met by an electric motor.

    However, let's suppose you come up witha rotary engine that really works, and is durable economical, etc, etc. It is a little difficult to workout a scenario where it would actually be useful.

    If small amounts of rotational power are required, an electric motor is almost always going to be the answer; in fact wherever mains electricity is available. Where it isn't, the usual answer is a fossil-fuelled internal combustion engine.

    So, we will assume we are off-grid, and all the fossil-fuel in the world has been used up. That has to include all the coal (of which there is an enormous amount) as otherwise we could simply synthesise hydrocarbons from it. The technology is nearly a hundred years old and is well-understood, though it does use up a lot of energy to make the conversion.

    So, we must assume no coal, and perhaps we are forced to use steam, presumably using wood as a fuel. Even then the world is probably not going to beat a path to your door, as for even quite small amounts of power steam turbines are simpler and more efficient than any form of piston engine.

    The very first electric "motors" were actually reciprocating machines, usually called
    electromagnetic engines to distinguish them from the usual rotary electric motors. However, it became clear early on that electromagnetic engines were always going to be inefficient and unsatisfactory, and that rotary was the way to go.

    THe first practical rotary electric motor is usually considered to be that introduced by Moritz Jacobi of Prussia, in May 1834. Four years later in September 1838, his second motor was powerful enough to drive a 28-foot boat carrying 14 people across the Neva river. (Jacobi did most of his work in Russia) It was not until 1839/40 that other developers were able to produce comparable motors.

    It occurred to me that this success story might have encouraged the belief that rotary steam engines would supplant reciprocating steam engines in the same way. In actuality the modes of operation and the technical issues of electric motors and steam engines are so different that such a proposition is less than convincing, except perhaps to a doggedly determined inventor.

    By 1910, the British Patent office alone held over 2000 patents for rotary type pumps and engines. Many promising applications, such as driving electrical generators on steam locomotives, were taken over by small steam turbines, which gave a usefully higher rpm.
    However, rotary air compressors and motors have been highly successful; presumably because leakage at the clearances is less of a drawback than for a steam-engine. Note, however, that compressors for very high pressures are invariably of the piston type.

    The Rotary steam engine was a seductive concept, but the practical difficulties are severe, and it languishes in almost total obscurity. No doubt the problems could be overcome given sufficient incentive; consider the slow and painful but ultimately more-or-less successful efforts to make a durable Wankel engine. Consider also that the piston engine remains near-universal...
    Today there are very, very few possible applications for the Rotary steam engine; in almost all cases an electric motor powered from a central power station will beat it on efficiency, convenience, and just about every other consideration you can think of.

    Robert Scott Burn "The Steam Engine" 5th edition. Publication date unknown but post-1844.
    This book is the only one I have ever seen that has a whole chapter on rotaries. The
    Simpson & Shipton engine is on the frontispiece, so the author presumably thought it was not just a waste of time and steam.

    "A Descriptive History of the Steam Engine" by Robert Stuart, p174. Second edition, 1824"

    "The Engineer's And Mechanic's Encyclopedia" Luke Hebert 1846

    Scientific American, Feb 25 1860

    "The Book of Modern Engines" Vol II. Caxton 1912.

    Dickinson "James Watt & The Steam Engine" OUP 1927

    "The Wankel Engine: Design, Development, Applications" Jan P Norbye 1971

    Back to Home PageBack to The Museum EntranceBack to The Rotary Steam Engine WingTop of this page