Valve Gear Design for Steam Engines - My approach

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I have read that on some of the first steam engines, a boy pulled on a ropes with actuated the valves, and thus the first valve-gear was a boy pulling on ropes. That would be a very boring job for sure.

What the D-valve does is combine all four of the Corliss valves into a single valve, and thus you get the action of four valves in a very simple and compact single-valve configuration. The D-valve is a rather ingenious device for sure.

The D-valve is not as efficient as a Corliss valve system, but the D-valve can be used for high speed engines (perhaps 300 rpm was considered "high speed"). Corliss engines generally operated at perhaps 70 rpm.

The D-valve saw wide use in many general purpose steam engine, and it was relatively simple to manufacture, and relatively simple to maintain.

D-valves over time will autmatically take up any even wear between the port face and the mating valve face, unlike piston valves.
It should be noted that a D-valve is not rigidly connected to the valve rod, to allow the D-valve to move in/out slightly, and move closer to the port face as wear occurs.
.
 
Last edited:
A typical D-valve admits steam to the cylinder via its outside edges, and the steam flows down the passages to either end of the cylinder.

The exhaust flows in the center (often domed shape) part of the D-valve.

The port openings in the valve seat (the piece of the cylinder that the valve slides over) typically have an exhaust opening in the center that is twice the width of the steam port on either side.

As the piston approaches the end of its stroke, the D-valve has to relieve the pressure inside the cylinder, and this is what causes the characteristic "puff" sound of a steam engine exhaust.

The valve has to open to exhaust early enough to allow full release of any remaining cylinder pressure, and you don't wait until bottom-dead-center to open the exhaust valve.

.
 
A somewhat odd occurrence is the "compression"point, and this is where the D-valve has its interior edge closed to exhaust, and before the outside edge opens to steam.
As the piston returns to top-dead-center, and after the compression point occurs, the cylinder is sealed, and so the piston compresses any remaining steam back to boiler pressure just before admission begins.

Many engineers spent many years perfecting the D-valve design, and there are many books written about valve design.

Steam engines were one of the few portable bulk power devices that existed for many years.

Prior to steam engines, much of industry was driven by water wheels, which meant that factories had to be positioned near fast flowing and constant streams or rivers.

People from the general public often ask me at steam shows "That is a fantastic and interesting device (referring to a steam engine), but what does it do?". The general public has no clue about steam engines, other than perhaps steam locomotives.
Steam engines accomplished what electric motors accomplish today.
Most people don't even relize that there was a time before electrification.

Ships like the Titanic were powered by gigantic reciprocating steam engines (the Titanic also had one turbine), and these ships had electrical power.
The electrical power was made using steam engine/generator sets, which is a steam engine connected to a generator.

.
 
Roughly speaking, the D-valve moves opposite of the piston.

The valve is moved by the eccentric, which is just a crank arm (like on a bicycle) what is made into a full circle, and instead of a pedal, a strap is placed around the eccentric.

.
 
This is a mechanical description of how a D-valve works.

The understanding of the thermodynamics of a steam engine is an entirely different subject, and a very involved/complex subject.

I will leave the discussions of thermo to those who have a lot more brain capacity than my tiny brain has.

.
 
The oscillating steam engine combines the valve into the cylinder, and the cylinder rotates on the port face.

Here is a typical Cretors oscillator.

For very simple engines, the oscillator is a good solution.
It should be noted that some of the old oscillating steam engines were huge (pehaps 10 foot bore), and drove sidewheel steam ships, so they were not used only to power small loads.

The Cretors oscillator drove the popcorn popper, as well as powering the rotating drum for the peanut roaster.

Most folks start out by building a simple oscillating steam engine, and it is a good exercise in both steam engine mechanics, and learning how to machine engine parts.
.
rCretors-Horiz-02.jpg
 
Hi GT.
You are correct with PV=mRT.... except for steam (or other fluid) when the temperature and pressure are at a change of state. Then the latent heat fouls the equation so it just doesn't work - hence the empirically derived steam tables. - Nowadays Ethalpy tables, that have been derived by calculation. Bread and butter to refrigeration engineers, as they make their engines work by vaporising liquids then condensing them completely to move heat from A to B to C. In theory, that is exactly what steam engineers are doing, but as we only condense the steam partly in our devices, it is not so easy to calculate. Just easy at the boiler end!
For Power stations, using dry steam that is still dry when it exits the engine (Turbines can be destroyed by wet steam!) then it is easy to calculate. Or in stages of multi-expansion engines using superheated steam... but to do the calculations for condensing engines such as most models, in micro-steps as in computer modelling, it all gets very messy, I think. Or at least in my head!
But your summary of valves, timing, etc. is excellent anyway!
K2
 
Here are a couple of animations I made for a ship oscillator.

The second one illustrates a feathering paddlewheel, which is designed to keep the paddles perpendicular to the water flow.




 
Hi GT.
You are correct with PV=mRT.... except for steam (or other fluid) when the temperature and pressure are at a change of state. Then the latent heat fouls the equation so it just doesn't work - hence the empirically derived steam tables. - Nowadays Ethalpy tables, that have been derived by calculation. Bread and butter to refrigeration engineers, as they make their engines work by vaporising liquids then condensing them completely to move heat from A to B to C. In theory, that is exactly what steam engineers are doing, but as we only condense the steam partly in our devices, it is not so easy to calculate. Just easy at the boiler end!
For Power stations, using dry steam that is still dry when it exits the engine (Turbines can be destroyed by wet steam!) then it is easy to calculate. Or in stages of multi-expansion engines using superheated steam... but to do the calculations for condensing engines such as most models, in micro-steps as in computer modelling, it all gets very messy, I think. Or at least in my head!
But your summary of valves, timing, etc. is excellent anyway!
K2
Thanks.

That brings to mind what I read about "superheating" steam, which is a discussion in itself.

Superheating steam adds energy to the steam, and raises the temperature of the steam.

And I guess superheated steam still has water vapor in it, but that water does not condense out of it easily ?

I also found out that steam can be superheated to some very high temperatures; maybe 1,000 F or more (I am not up on superheated steam temperature ranges). Turbines can handle superheated steam temperatures that are quite high.
Reciprocating steam engines can only handle a certain amount of superheat temperature.

The Titanic engines extracted the power from the steam across multiple pressure drops in its compound reciprocating steam engines, and then further extracted energy from the reciprocating engine exhaust by powering a low-pressure turbine.

I feel lucky to have some understanding of the mechanics of how a steam engine operates.
I leave the thermo aspects to the "rocket scientists", since that is somewhat beyond my comprehension.

.
 
I have a notion on Single acting engines that exhaust to atmosphere?
Could it be that the steam - that has lost much of its heat and has partly condensed during the power stroke - is expelled when the exhaust valve opens, but this expansion causes a further drop of temperature - and condensing - so that the pressure in the cylinder (at some indeterminate point in the stroke) drops below that of the atmosphere.
Real indicator diagrams confirm that does not happen. In some circumstances, eg light load and short cut-off, the pressure can drop below atmospheric before the exhaust valve opens, but that of course means negative work. I am not dead sure of this, but ISTR when the exhaust is released and can expand reasonably freely (depending on the back pressure that may be needed to create draught) then it tends to become drier or superheated, rather than wetter, despite the temperature drop.
 
The most fundamental arrangement of valves on a steam engine, and the easiest to understand, is a 4-valve Corliss arrangement, shown below.

The function of a Corliss valve arrangement is as follows:

1. Open the top left steam valve, force the piston down the cylinder.
2. Close the top left steam valve, and open the lower left exhaust valve.
3. Open the top right steam valve, force the piston back in the opposite direction.
4. Open the lower right exhaust valve.

The Corliss engines were very efficient, and much of that efficiency was due to the separation of steam and exhaust passages, and the very quick and often totally independent operation of the steam and exhaust valves.

View attachment 140326

Hmmm - - - wondering - - - - how would the Corliss engine efficiency compare with something like an engine designed like that of Dan Gelbart's?

He's using electrics and electronics to make the valve arrangement from a mechanical to and very very controllable function.
I've been thinking of trying something all his lines.

Is there a way of calculating any of this?

Comments/suggestions/(whatevers)?
 
GT: Just for clarity: "superheated steam still has water vapor" Steam is pure water vapour: I.E. Water above boiling point (appropriate for the pressure) and purely in a gaseous form. It is this "gas" that is so useful to the engineer, because it can carry so much more energy than Air, or many other gases. I guess you are confusion the use of "Water vapour" (I.E. the water converted to gases or "vapour" state) with the white mist of condensed water that appears when the steam is at the boiling point, at which energy can flow into it and from it without a change of temperature, just a change of state. That energy of the change of state is the latent heat of "vaporisation". Most model engine cross the temperature and pressure curve at the boiling point (really, we should refer to this as condensing point in expansion of water vapour that is condensing). As they cross this point, some of the water vapour (steam) condenses to water droplets, which combine to "clog" the exhaust system for getting the remaining steam (water vapour) from the engine.
Just to try and make this more understandable: We talk of vapour in other contexts as well as for water. e.g. fuel vapour - by which we mean a gaseous state of the fuel. Things that are blown up can be vapourised = turned to gases - in the heat and pressure changes in an explosion, or high energy impact. (Such as NASA chucking a 500kg satellite into a 10km dia. rock. The metal etc. from the satellite, and some of the rock was "vapourised" in the heat of the impact and the jets of the mixed gases acted rocket-like (Newton's law of conservation of momentum) to push the lump of rock to change course). - Wow! - we have got into rocket science now! (I.E. Newton's second law).
But back to superheated steam... We make steam in a boiler - then pass it along tubes and heat the steam to much higher temperatures, in order to carry the energy within the heated vapour to drive the turbine, without ever crossing of the pressure-temperature line between water vapour and liquid water. - The water droplets would knock the blades off the Parson's design of turbine. SO, the turbine exhaust is dry steam - by design.
On this I would agree with Charles: the reciprocating steam engine that is using superheated steam acts similarly to my description of the turbine, especially in compound engines where the exhaust from one cylinder is at a relatively high temperature and pressure in order to power the next stage of compounding. But if you had a relatively early cut-off, then the steam that was introduced to the cylinder can expand (losing energy) to the point of condensing (extracting some of the latent heat from the steam) so the exhaust is "wet" when the valve opens. But with longer piston travel before cut-off, then the steam cannot lose enough energy for the temperature and pressure "condensing point" to be crossed, so you can exhaust dry steam. And when braking a loco, by putting the gear into reverse, the momentum (kinetic energy) from the train actually compresses the steam in the cylinder, thus heating if to a superheated state when the exhaust valve opens. (But it turns my brain inside out trying to visualise this!).
Now I'll take my worn-out brain and rest a moment to contemplate this... and how to model the pressure in the cylinder mathematically.
Oh, Charles, please can you translate ISTR for this poor humble idiot who doesn't know all the jargon?
Thanks! A very interesting thread and lots to ponder! (e.g. DO I really know what I think I know, or am I barking stupid?).
K2
 
Just to explain the Newton's second law - of conservation of momentum - re: the NASA DART mission.
Take 1000 million Tons (1000,000,000) Tons of rock at rest (or so it thinks in orbit of the asteroid - or whatever.).
Take a 0.5 Ton satellite travelling at 500million KPH: (500,000,000).
The combined satellite and lump of rock sees, before the impact,
a total momentum of 1000,000,000 x 0 + 0.5 x 500,000,000 = 25,000,000 TKPH of momentum.
But after the impact, it sees the rock and satellite debris going one way, and shrapnel and gas going the other way. So: the 25,000,000 TKPH momentum must be conserved and the 100kg of gas flies out at 1000,000,000kph plus the 1 ton of debris flies out at 10,000kph giving an equation of
25000000 = (999,999,999 x v) - (0.1 x 1000000000) - (1000 x 10000)
I.E. 25000000 = (999,999,999 x v) - 100000000 - 10000000.
=> 25,000,000 + 100,000,000 + 10,000,000 = 999,999,999 x v
=> 135,000,000 / 999,999,999 = v
So, I would expect the lump of rock to move away at about: v = 0.135km/hr. = the general "rocket science" calculations that NASA are pretty good at.
The incredible thing is that at such a slow speed moving away from its original path is something that NASA will measure!
Given a couple of thousand hours for it to move a detectable distance off course, it should be easy?
Hope you enjoyed the Maths? (Did you find my error?).
Sorry to detract from valve gear, but I was sure someone would ask....?
K2
 
And when braking a loco, by putting the gear into reverse,

Er, no. That is not how you brake a railway locomotive, or a train, in spite of what you may have seen in films. Holding the reverser slightly into reverse gear is done on a traction engine, however.

ISTR : I seem to remember
 
Last edited:
Hmmm - - - wondering - - - - how would the Corliss engine efficiency compare with something like an engine designed like that of Dan Gelbart's?

He's using electrics and electronics to make the valve arrangement from a mechanical to and very very controllable function.
I've been thinking of trying something all his lines.

Is there a way of calculating any of this?

Comments/suggestions/(whatevers)?
I would guess that it relates to how an engine will be used, ie: with a constant and non-varying load, or a fluctuating load.

For a fixed load powered by a Corliss, you would probably be hard pressed to improve on the mechanical function of that valve gear, give how precise it cuts off admission, as well as exhaust.

With a fluctuating load, and a steam engine that perhaps operated at a higher speed such as 300 rpm, and automated control system may get better efficience, assuming a 4-valve system or something analogous was used.

There is a white paper out there of a steam engine with its valves controlled by a programmable controller, with linear actuators.
I will have to look for that.

Bottom line is the best of reciprocating steam engines are not very efficient, compared to today's diesels and steam turbines.

I think they did make a steam turbine locomotive.

They also made some compound (4-cylinder) steam locomotives, which I think were efficient, but prone to wheel slip for some reason.
I have an article on this too if I can find it.

.
 
On the history of valve gears:
In 1698, the English mechanical designer Thomas Savery invented a pumping appliance that used steam to draw water directly from a well by means of a vacuum created by condensing steam.

A new design by Thomas Newcomen in 1712: the valves of his Steam Condensing "Atmospheric" mine-water pumping engine had no rotative motion for valve operation, but the valves were operated by the Engine driver. His valves were one for admission of steam to raise the heavy engine piston and rod, and lower the pump rod, and a second valve that admitted water to the engine (later a condenser) to quench the steam and create the vacuum, so the atmospheric pressure on the outside of the piston could raise the pumping rod and pump the water up and out of the mine.
Move on to James Watt: His developments to improve the Newcomen design developed a new design to be introduced commercially in 1776: When he improved the Newcomen design with a condenser, the second valve became an exhaust valve from the cylinder.

Some later engines had dogs and levers operated by the piston rod - the timing produced by the position of the piston rod.
In 1781 introduced Watt a system using a sun and planet gear to turn the linear motion of the engines into rotary motion. He could not use the crank as someone else held the patent for that. Later this became out of patent and he used cranks.
Because he developed the sun and planet gear arrangement, he had a reliable way of converting the reciprocating motion of the piston into rotative motion of the crank. But the valves were very like the original "reciprocating" lever operated valves on some engines (right up until the late 19th century?).
https://en.wikipedia.org/wiki/Watt_steam_engine#/media/File:Watt_steam_pumping_engine.JPGIn his first sun and planet gear driven flywheel engine you can see the rods to operate the separate valves per beam position.
https://cdn.britannica.com/35/60535...tt-steam-engine-gear-drawing-Science-1788.jpgAlthough by 1832 Watt appeared to be using D-Valves on his beam engines.
https://en.wikipedia.org/wiki/Watt_steam_engine#/media/File:Maquina_vapor_Watt_ETSIIM.jpgHowever, Richard Trevithick, amongst others, keen to develop engines using high pressure steam, started using a 4-way valve, driven from the crankshaft. But possibly he was not the inventor of the steam carraige. It is not clear what valve arrangement was on his first steam engine powered carriage (1803), but appears to be rotating valves (possibly his 4-way valve?) operated by the engine driver - operating the long lever to the valve?
https://en.wikipedia.org/wiki/File:Trevithicks_Dampfwagen.jpgWilliam Murdoch had developed and demonstrated a model steam carriage, initially in 1784, and demonstrated it to Trevithick at his request in 1794.
Certainly, in 1803~4 Thevithick still had "engine driver" arranged valve gear: as can be seen on his Coalbrookdale loco here: - The looped handles that were pushed and pulled to operate the valves for steam and exhaust.
https://en.wikipedia.org/wiki/Richard_Trevithick#/media/File:Coalbrookdale_loco.jpg
The Blenkinsop engine: predecessor of George Stephenson's locos - shows rotating valves for steam inlet and exhaust: (at 2.53mins). You can see the valve drive rods at 4.44 mins of this video.


George Stephenson's early engines - e.g. Killingworth no 2 - appear to have rotating valves, operated by some linkage not shown in this engraving.
http://www.trainhistory.net/images/trainhistory/killingworth-locomotive-george-stepherson.jpgOnce locomotive design took off (in a hurry?) the rotative motion from the cranks became the driver for the valve gear - using an eccentric - and reversing valve gears were developed. The Stephenson gear was probably the earliest commercial design of this? And I figure it was after 1833ish? But the earliest rail locomotives on the Stockton and Darlington railway were still required to stop before changing from forward to reverse direction. In 1831 or 32 - my direct ancestor - was recorded in the railway records as being the only engine driver who could reverse the engine's gear while still in motion. (Thus acting like a brake!). Against the usual practice of stopping the train, dismounting and changing the gear to reverse operation.

What I do not know, is where in the line of history did the development of the D valve, and similarly the piston valve, get introduced? I figured somewhere within the James Watt development?
And when did "cut-off" become a known advantage for economy? - Possibly as early as Newcomen?
I must read more history on this. e.g. the valve gear of Trevethick's "Catch me who can"
https://en.wikipedia.org/wiki/Richard_Trevithick#/media/File:Trevithick's_steam_circus.jpgand Stevensons Rocket all seem to have had rotating valves, later driven from eccentrics (on the Rocket)? But Newcomen had single valves hand operated. And Watt, through to Stevenson had linkages driving separate rotating valves.

Also curious to note that when Otto developed his first engine (1867), it used some gear mechanism driven off the piston rod to operate the valves - rotating type.
https://en.wikipedia.org/wiki/Otto_engine#/media/File:Enginy_Otto-langen_1867.jpg
So, with the later steam engines using Corliss gear and rotating valves it seems the design has gone full circle?
K2
 
Hi Charles,
Not sure about using the reverser to brake a loco... as I am not a loco driver! (One training lesson on the local club track nearly broke my back, so never again!).
But a couple of decades ago I went up Mount Snowdon on the rack and pinion railway... and was amazed when the driver used the forward gear with a small amount of steam to brake the train when reversing back down the railway. The loud "crack" of the exhaust was much louder than when going uphill... I think I chatted to him about it afterwards?
But My memory plays tricks, so you are probably right...
K2
 
Er, no. That is not how you brake a railway locomotive, or a train, in spite of what you may have seen in films. Holding the reverser slightly into reverse gear is done on a traction engine, however.

ISTR : I seem to remember
ISTR - - - also used for 'it stands to reason'

acronyms can be useful - - - they can also be a right royal pain!
 
I would guess that it relates to how an engine will be used, ie: with a constant and non-varying load, or a fluctuating load.

For a fixed load powered by a Corliss, you would probably be hard pressed to improve on the mechanical function of that valve gear, give how precise it cuts off admission, as well as exhaust.

With a fluctuating load, and a steam engine that perhaps operated at a higher speed such as 300 rpm, and automated control system may get better efficience, assuming a 4-valve system or something analogous was used.

There is a white paper out there of a steam engine with its valves controlled by a programmable controller, with linear actuators.
I will have to look for that.

Bottom line is the best of reciprocating steam engines are not very efficient, compared to today's diesels and steam turbines.

I think they did make a steam turbine locomotive.

They also made some compound (4-cylinder) steam locomotives, which I think were efficient, but prone to wheel slip for some reason.
I have an article on this too if I can find it.

.
Hmmmm - - - - as I'm looking for something where there will be a varying load - - - well that's why I've been chasing this particular rabbit - - - grin!

When you find said paper a link or copy, if possible, would be wonderful - - - TIA!

Hmmmmmm - - - - Uniflow Steam Engine - History - Skinner Unaflow includes the phrase "This engine operated in a steeple compound configuration and provided efficiencies approaching contemporary diesels."

My interest is because of an inherent capacity to use not only less refined fuels but also a wider range of such.

A compound steeple is what I was thinking of using - - - so it means mods to mr Gelbarts single cylinder uniflow design.

All part of the fun - - - imo.

Thanking you for your assistance.
 
Back
Top