What contributes to steam engine efficiency?

[macardoso]

So I am about 60-70% done with a PM Research #6 engine and after a long hiatus, I am back into the build. I was thinking through steam engines from an engineering perspective and trying to understand how they differ from gasoline engines. In particular I was curious, if one to design a steam engine from scratch, what decisions could be made to improve the mechanical work done per unit of steam (specific work). Consider the the engine's fuel efficiency.



Here is the PV Indicator Diagram for a real steam engine as it travels through the engine's cycle. My #6 is a double stroke engine, so the return stroke would have a similar diagram but it would progress 180 degrees out of phase from the forward stroke..

This diagram shows the pressure and volume of the working fluid. The area contained inside the curve is the mechanical work extracted from the hot gas. Starting at #1, the piston is at top dead center (TDC) and full of high pressure steam. As the piston begins to stroke forward, we travel to #2. The steam valve does not close immediately so the piston remains at boiler pressure until we pass #2 (valve closing). From there, the gas trapped in the piston decreases in pressure while increasing in volume, driving the piston forward. At #3, the exhaust valve opens and pressure drops from #3 to #4. At #4, the exhaust valve closes and the cylinder retracts under the inertia of the flywheel to #5 where the steam value opens and fills the piston with steam. Once the piston hits top dead center #1, the cycle repeats.

The real steam indicator diagram is drawn as a blue line inside the ideal curve (red). This accounts for the non-instant opening and closing of the valves, valves opening before the end of stroke, and gas being trapped in the returning piston once the exhaust valve is closed.

If we want to extract the maximum work from the engine, we must design it such that the blue line lies as close to the red line as possible. Here are my personal thoughts on what would do this the best, although I could be wrong on some - would love to discuss.

#1) Steam must enter a hot cylinder. Any heat loss to the cylinder walls will drop the curve at point A from ideal as the pressure in the cylinder will decrease before the power stroke can begin.
#2) The steam valve should provide a high flow of steam into the piston (large orifices in the valve). This will allow the cylinder to pressurize as rapidly as possible. This would bring the segment E->A closer to the segment 5->1.
#3) The steam valve should open and close as rapidly as possible and be open for the least amount of time possible. This allows the maximum amount of stroke of the piston to extract work from the gas.
#4) The steam porting volume contained on the cylinder side of the valve should be minimized. This will minimize the volume of steam consumed per stroke. The exhaust pressure of the steam should be as close to atmospheric as possible. Any pressure left in the cylinder at exhaust is waste. This may not be optimal for power output of the engine.
#5) The exhaust valve should open as close to the end of stroke as possible, allowing the maximum time for the gas to expand and do work.
#6) The exhaust valve should remain open for as much of the return stroke as possible. This minimizes the compression of gas as the cylinder approaches top dead center and moves point E closer to ideal.
#7) If there is a minimum cylinder + porting volume at TDC achievable due to mechanical requirements and porting dimensions, then there is a maximum steam supply pressure beyond which any additional pressure adds no work and will be exhausted above atmospheric pressure.
#8) Steam pressure below this maximum pressure may extract more work from the gas (efficiency) at the expense of power output. There is a critical pressure which provides high efficiency and power. Higher pressure = more power, lower pressure = more efficiency (I think).
#9) An optimized steam valve orifice (not just a circular port) could improve the valve response and the efficiency.
#10) An optimized steam valve actuator design could allow for independently adjustable durations of the supply and exhaust segments, and well as rapid transitions between opening and closing.

 

[tq60]

They used multi-expansion engine where regarding steam flow the cylinders are in series and get larger as they go.

The expansion volume and pressure being different as each cylinder extracts usable power the still at pressure steam passes to next cyl.

Sent from my SAMSUNG-SM-G930A using Tapatalk

 

[benmychree]

A steam jacketed cylinder and superheated steam can add efficiency, multi staging only is advantageous with higher pressures; steam is generated at higher pressures more efficiently than at lower pressures ( less BTUs per lb of steam)

 

[Larry$]

Since we are talking piston engines and not more efficient turbines, how about doing a port and polish with opening and closing being activated by electronic control to optimize the engine and any speed. While it may be possible to operate the valves by electrical coils compressed air would likely be cheaper/easier. Mechanical valving would be very difficult to optimize for varying speed. Could efficiency be gained by trying to pull a vacuum on the outlet port by condensing. The last stage on a naval turbine is always under vacuum by cooling with sea water. Maybe a spray of water at the port would add something sort of like atmospheric engines. The closer to the port the better.

 

[benmychree]

What make a piston steam engine inefficient is condensation within the cylinder and consequent re evaporation as the stroke continues is where the poor efficiency is based, no valve operation tweaks are going to make much difference, those done in the past only worked with large slow speed engines, and relatively efficient as they were, they were way behind turbines. Condensing engines were more efficient, but turbines are also condensing.

 

[Lo-Fi]

How deep down the rabbit hole do you want to run?

#1) Steam must enter a hot cylinder. Any heat loss to the cylinder walls will drop the curve at point A from ideal as the pressure in the cylinder will decrease before the power stroke can begin.

Keeping everything in the steam circuit lagged and warm is a great place to start.

#2) The steam valve should provide a high flow of steam into the piston (large orifices in the valve). This will allow the cylinder to pressurize as rapidly as possible. This would bring the segment E->A closer to the segment 5->1.

Crucial. Work by Andre Chapelon and L.D.Porta among others showed that making the steam circuit as large and free flowing as possible is vital. Not just the valves, but passages themselves should be smooth, large and free of tight bends or obstruction.

#3) The steam valve should open and close as rapidly as possible and be open for the least amount of time possible. This allows the maximum amount of stroke of the piston to extract work from the gas.

What's missing on the model is a reverser. In locomotives, the reserver is also used to vary the travel of the valves, and therefore how much of the stroke steam is admitted for. You're right that expansive working is far more effiecient than simply letting as much steam as you can blast in. More on this below...

#4) The steam porting volume contained on the cylinder side of the valve should be minimized. This will minimize the volume of steam consumed per stroke. The exhaust pressure of the steam should be as close to atmospheric as possible. Any pressure left in the cylinder at exhaust is waste. This may not be optimal for power output of the engine.

Absolutely. Any exhaust back-pressure should be minimsed as far as possible. For a fascinating read, grab a copy of The Fire Burns Much Better

#5) The exhaust valve should open as close to the end of stroke as possible, allowing the maximum time for the gas to expand and do work.

This is a complicated matter at higher speeds, but broadly yes.

#6) The exhaust valve should remain open for as much of the return stroke as possible. This minimizes the compression of gas as the cylinder approaches top dead center and moves point E closer to ideal.

Yes, but with respect to the same valve handling admission as is the case with classic slide and piston valve systems driven from a simple eccentric, this isn't usually the case. The advantages of expansive working on the admission side far outweigh the increased back-pressure on the exhaust. I can see that you're already thinking on the lines of separate inlet and exhaust timing, though... Again, more below.

#7) If there is a minimum cylinder + porting volume at TDC achievable due to mechanical requirements and porting dimensions, then there is a maximum steam supply pressure beyond which any additional pressure adds no work and will be exhausted above atmospheric pressure.

#8) Steam pressure below this maximum pressure may extract more work from the gas (efficiency) at the expense of power output. There is a critical pressure which provides high efficiency and power. Higher pressure = more power, lower pressure = more efficiency (I think).

Kind of. What's been shown (I believe by test at Rugby for British Rail if memory serves), is that running the regulator full open and varying the steam admitted by the valves by varying the cutoff (reverser position) is most efficient. Standing on the footplate of a large locomotive with the regulator full open and winding the reverser forward is quite an experience, I can tell you!

#9) An optimized steam valve orifice (not just a circular port) could improve the valve response and the efficienc

Commonly, full size locos use annular ports with arrangements made for the rings to ride over them. As mentioned above, ports should be literally massive. Models have tiny pin-holes by comparison to what's ideal.

#10) An optimized steam valve actuator design could allow for independently adjustable durations of the supply and exhaust segments, and well as rapid transitions between opening and closing.

Yes, though sadly most systems did not achieve this. British Caprotti as used on British loco Duke of Gloucester is a great example of what you're describing. It's essentially variable cam-driven poppet valves like a modern car engine with inlet variable separate from the exhaust as well as rapid valve movements. Well ahead of its time! The story of the beast is quite fascinating and happily it's survived into preservation.

A few other things to note:

Steam leaks, even small ones, have a large effect on efficiency. There's a paper by Porta around somewhere.

Compounding, as I think mentioned above, is a great way to extract the most energy out of the available steam. Fench locomotives ran systems with dual reversers so cutoff could be adjusted indepentently.

Piston ring sealing and friction can have a bearing too. Again, L.D.Porta wrote a paper I can't find at the moment.

Probably lots I've forgotten! It's 4:30am here. I woke up and made the mistake of idly browsing the forum.... Been awake thinking about it since reading your post!

To close off, here's a lovely video of a model Duke running with the valve covers exposed:

 

[macardoso]

God that video is so cool. First off, I've been thinking about this for days, Come to a few personal conclusions which seem to be supported by comments here. First off is that the piston engine is going to be inefficient compared to a well designed turbine, and second, eccentric driven valve timing is easy to build but poor in performance.

I'm not familiar with the reverser you commented about and will research.

My comments about lowest pressure steam possible were based on the thought that any temperature above ambient in the exhausted gas is wasted energy. Now obviously you won't reach ambient in the expansion of the hot steam but getting as close as possible would be better. However I've certainly heard that super pressurized steam is the way of the future so I need to read more and understand that better.

 

[Lo-Fi]

Interestingly, there was a mainline turbine locomotive built and run extensively: The LMS Turbomotive. It performed well; slightly better than it's piston driven cousins in efficiency and performance. I have a feeling if it had been developed further, it would have become the dominant technology.

From what I know of steam turbines used in power stations (which admittedly isn't very much), the real efficiency comes from using many stages of compounding and holding the exhaust under atmospheric pressure to extract maximum work from the remaining energy in the working medium once it gets to the last turbine stage.
There's simply not the space to do serious compounding in a mobile application with the requirements for the low pressure turbines to be as large as they need to be to do any useful work at low energy, so this is probably the limiting factor.

That all being said, there's serious fun to be had making model steam turbines, clever valve gear for reciprocating setups... You could probably spend a happy lifetime tinkering with different configurations. Having seen your other work, I'm looking forward to seeing what you come up with if you choose to explore steam tech!

 

[savarin]

What about the una-flow engines?
I dont know anything except they operate similar to a 2 stroke engine.

 

[Lo-Fi]

The uniflow arrangement is something new to me - probably because my interest is primarily in locomotives, which it's not particularly suitable for. Just looked up, it's quite intriguing!