A lot of small engine makers will only run them on compressed air because of the regulations regarding certification of live steam boilers here in the UK. I am helping make a twin cylinder double acting self starting model engine from an incomplete set of castings which will be lubricated with motor oil and only used on compressed air for a short time.

I find an engine run on air does not need as much lead of the eccentric as one run on steam. As there is no expansion of the air retiming helps a lot as does reduced lap.

Although I tend to run my steam engines on air they will still do work and you can hear the change in sound when they are loaded up. You wil also find that those running “tabke top” engines on air tend to use very low pressures less than 0.5 bar is typical and regulate dright down a swe like a nice slow running engine but if you up the pressure to what you might expect from a boiler 4-6bar and open up th eregulator then they will do a lot more work as what energy goes in must come out

Interesting. I would like to have a good think about this but life does not allow it.

Air is a perfect gas while steam which is, or will end up, wet is not. To state the obvious steam engines are designed to use steam. If air is used the expansion will be PV^n = constant. The power n could be anything between 1.4 and 1.0. 1.4 is adiabatic (where not heat is lost during expansion) to 1.0 (where expansion is cold, difficult to visualize). High speed engines will be close to adiabatic, most slow running engines won’t. While a steam engine’s valve will allow steam to be admitted for a short time I think the exhaust will open far too early to make full use of a gas’s expansion.

This is complex and DO NOT compare it with the Otto or Brayton cycle.

JA

Werner, please keep us informed of your results.

On 19 November 2025 at 16:30 JA Said:

Interesting. I would like to have a good think about this but life does not allow it.

…

This is complex and DO NOT compare it with the Otto or Brayton cycle.

JA

Werner, please keep us informed of your results.

First, I owe Werner an apology.  My first answer told him stuff he already knows because I forgot his previous posts on the subject show he knows how to measure engines!

I’m in the same boat as JA.  I’m interested in thermodynamics but don’t have the time to master it.  And to be honest, my tiny brain is an obstacle!

Heat engines convert heat into useful power.  Theory identifies 6 ways in which the working fluid can be expanded to do work:

• Adiabatic.  Hot fluid expands and temperature drops rapidly.  Work is delivered quickly, pulsed, not smooth,
• Iso-thermal.  Hot fluid expanded such that temperature does not change.  Does most work, but slowly.
• Iso-baric. Heat added to maintain constant pressure.  Temperature does not change.   Less efficient, less power, but constant, which is sometimes exploited.   iso-baric is mostly used to add heat to the working fluid for an adiabatic or iso-thermal expansion in the next stage.  Diesel engines.
• iso-choric.  No work, temperature changes.  Mostly used to add heat to the working fluid for an adiabatic or iso-thermal expansion in the next stage.

These expansions can be exploited by designers to maximise efficiency, or power, or torque, or smoothness.   And, in a practical engine, to minimise pollution, shocks, excessive temperature, resonances, maintenance and rough running.  Ignition is a major factor, as are many other practical considerations.  Steam is not an Ideal gas!

Practical cycles include:

• Carnot – sets the upper limit on heat engine efficiency.   Although the Victorians put massive effort into building steam and IC engines using the Carnot cycle, it’s impractical.  Valuable theory.
• Otto – spark ignition engines.  Variations include Miller and Atkinson.
• Diesel – compression ignition engines
• Brayton – jet engines and turbines
• Stirling and Ericson – gas engines
• Rankine – vapour engines Steam engines

Model steam engines should implement the Rankine Cycle.  They sort of do, in that their proportions are derived from full-size prototypes, and they produce Rankine-like indicator diagrams.  As far as I know, no-one has built a model engine explicitly designed and built to a perfect Rankine cycle.   May not be worth the effort because small engines are grossly inefficient due to scale imposed friction and heat losses.

Compressed air does not perform well in a Rankine Cycle engine – a steam engine’s proportions and valve events aren’t optimised to produce the most appropriate adiabatic / iso-thermal expansions.  Compressed air works well-enough for testing, but the engine won’t get the best out of it, and, compared with steam, there isn’t much energy in air.  Enough heat in it to turn the engine over, but power output is low.  How low depends on practical unknowns: possibly the proportions and valve events in a model are so off Rankine that compressed air works better!   The only way to find out is to measure it by putting it on a dynamometer and taking indicator diagrams.  Indicator diagrams graph pressure vs volume as they change during the stroke, and the shape of the loop shows how close, or not, the engine is to it’s design cycle.

Practical example.  In WW2, British Agents were parachuted into occupied Europe.  Their radios used thermionic valve that guzzled electricity.  A serious problem when operating covertly from a wood!  One way was to charge lead-acid accumulators with a hand-cranked or bicycle driven dynamo, both exhausting hard-work.  Petrol generators need hard to obtain fuel and were noisy.  Stuart-Turner were asked to provide a small steam engine and boiler, also inconveniently heavy, but ran on water and windfall wood.   For obvious reasons, compressed air is unsuitable!  Compressed air is great in the workshop though.   Air-tools could be steam-powered, but…

Dave

Werner

Many thanks. My problem is that I only try to think about these things when it is cold. When it is warm I am, or should be, in the workshop.

Dave

Just a comment: Strange as it may seem the Diesel and Brayton cycles are the same (ideally adiabatic compression, constant pressure combustion followed by adiabatic expansion). Some gas turbines, before 1935, have used the Otto cycle where constant volume combustion is used. This has, unsuccessfully, been revisited in the last twenty years.

I have a few thermodynamics and fluid dynamics text books from my youth and they are very interesting but do not make light reading.

JA

Julie

To be pedantic, no system can be adiabatic. To do so it would contravene the Second Law of Thermodynamics. (As an aerothermal engineer “enough said, laws are laws”).

JA

Sheer curiosity – what would an engine designed from the start to run on compressed air look like?

Not very well, with a compound you are probably aiming for a larger expansion ratio, so as explained above, air would lose even more pressure. Of course you could alter the cut off in each cylinder to compensate, but as the main reason for compounding is to avoid condensation, it seems a big pointless. All the extra gubbins whirling round absorbing power.

Looking at the interesting indicator diagrams, the first things I would observe are the pointy bits at the right, and the sloping first half of the exhaust line. These suggest that the exhaust is opening too late on both cylinders, and that the exhaust port or passage area is too small.

We had an air engraver for part marking certain aircraft parts which was a straight version of a dentists drill. It took 1/2mm tip diameter carbide spherical tips with five flutes cut in them and ran at 250000 rpm. The power came from air striking a wheel of about 5mm diameter tangentially which had about 20 tiny scallops cut out of it, in the same manner as a Pelton wheel. When it was in use, some people dissapeared out of earshot, it sounded like a mad dentist was torturing somebody. It used the smallest ball races that I have ever seen.

My comment about not many pistons referred to the Ingersoll Rand link, only the very large there seem to be piston based.

It is possible air tools had/have a nutating disc but there are quite a few with various blades or vanes where the air enters at a tangent to drive them. Such as this drill