Rufledt
Rufledt UberDork
1/30/17 2:06 p.m.

TL:DR: i want to make a better steam engine for literally no reason

I've been reading a lot of completely irrelevant stuff since i can't go back to work for another week, and i eventually made it to steam engines. The old kind, like, boilers, trains, stationary engines, etc... I think it's an extension of wanting to build stuff in metal, casting, machining, welding, and so on, but i can't be sure. When i came off the pain killers last week i was reading all about steam powered stuff so i don't remember how i got here, but i decided to stay awhile.

These days there are lots of people making small model steam engines, such as these, and some people using functional ones as tools in a shop. None of these are new-ish designs. Obviously internal combustion engines and elecrtic motors took over and have won the war for a whole host of reasons. I can't begin to fathom building an internal combustion engine from scratch, but it seems like making steam engines is common place. Why not try to improve them?

There are tons of valve designs. the 'normal' one is the slide valve like on a train:

it's reversible (because trains want to go backwards too) and has some 'tunability' for how long the intake port stays open. That's the thing, steam is hot and pressurized, so you get best efficiency if you use the expansion of the steam. you do this by shutting the intake port earlier in the 'power stroke' and letting expanding steam do some of the work. Some ship engines and stationary engines also put the steam through multiple cylinders to take advantage of all the expansion:

Those are called compound engines. Steam enters small cylinder, expands a bit, goes to the next one, expands more, and so on. Sometimes 4 cylinders with steam only entering the first one. These are actually more efficient than steam turbines at low speed, but WAY less efficient than diesel engines. larger cylinders are the lower pressure ones, allowing for all cylinders to have similar power output.

The most efficient ones seem to be corliss type engines, with valves that open rapidly at the beginning of the power stroke, and then cut off immediately at a variable point. It's controlled by the governor, so it's always operating at the same speed, but when more power isn't needed, it increases efficiency by cutting off earlier. It's crazy to watch: valves and pushrods and dashpots, oh my! Pic of something similar if you don't want to watch video:

It's really interesting, it runs the throttle wide open but varies the valve opening for throttle control. Sound familiar? This was invented in 1849. Genius. It also uses cylindrical valves that rotate, poppet valves weren't quite figured out yet and don't really work so well for steam i guess. One issue there is they can't move very fast. it uses gravity or air suction to pull the intake valve shut after cut off. The simplicity of the slide valve above also seems to be better for high speed applications, like trains, which need to go balls out (a term i recently learned came from centrifugal ball governors commonly found on steam engines) for a long period of time.

There has to be a way to get the efficiency of one with the speed of the other, right?

The other problem i saw was airflow:

Not sure how well air can flow smoothly there. THe exhaust (pink space) for max efficiency should escape as easily as possible. The corliss mentioned above has separate exhaust valves so the exhaust can go straight out the bottom, which also helps not cool down the intake port. but even then, it all empties into a box with a single outlet:

The intake goes into like a plenum thing, too, which i'm not sure is necessary for steam engines. I think it has more to do with technology of the time and simplicity. One tube from the boiler into a larger space inside a strong casting to hold the pressure. Pressure is serious on these things, no ambient +2 bar of boost like a high power internal combustion engine, we're talking hundreds of pounds of pressure sometimes.

You don't need to worry about helmholz resonance or whatever that's called, shorter is always better for intake runners. Shorter is better for almost everything, shorter distance from boiler to engine, shorter distance between cylinders in compound engines, basically always try to keep the heat in the steam engine until the steam is exhausted to air or a condenser.

The low speed vs. high speed engines have different 'clearance' too. Clearance is the amount of space in the cylinder at TDC. High speed engines have lots of clearance and open the intake valve slightly before TDC, i think the idea is to fill the space with as much steam as possible by the time TDC happens, because stuff is moving fast and some of those locomotives have HUGE cylinders. like over a foot bore. One nice thing is the lack of detonation, all pressure operated, no fire (except in the boiler).

Have any of you ever though about combining some of these with some more modern type valvetrain stuff? I've seen small model engines with a cam and roller instead of the steam engine 'eccentric' and i think that's a good start. Some large steam engines used cams but at very low speeds, and if you can get variable cam duration with valves that can close rapidly at high speed i think that would be the key. If modern valve springs can close a valve, rocker, and pushrod at 7k rpm, then surely something could be designed for a steam engine to close rapidly at less than half that rpm. I came up with some ideas, but i'm currently learning CAD so it'll be a little while before i have the software skills to actually model it.

what are your ideas? modern materials? make poppet valves work? should i seek psychiatric help? Seriously, if i could do any work in the garage i would be doing it (though technically i would be back at work if i could do that) but i can't just sit and do nothing.

Dr. Hess
Dr. Hess MegaDork
1/30/17 2:21 p.m.

My ideas: Don't berkeley around with valves, rods, reciprocating engines. Go right for the real thing: Steam turbine. A multi-stage steam turbine engine approaches unity in efficiency. Do that with poppet valves. The first one created would fit on your desk and did 10HP. Make one of those.

RossD
RossD UltimaDork
1/30/17 2:39 p.m.

I work a bit with steam. I want a toy steam engine too. Just remember steam can be very dangerous, but you probably already knew that.

chaparral
chaparral Dork
1/30/17 3:14 p.m.

I agree with Dr. Hess. The only place reciprocating engines stuck around was on the railroads, where dirty road conditions and dirtier shops made the turbines, gearboxes, and draft fans unreliable. A 1000 horsepower steam turbine is smaller than the turbine section of a 1000 horsepower gas turbine. The problem is the boiler...

Rufledt
Rufledt UberDork
1/30/17 3:24 p.m.

Obviously turbines are better, but i'm only thinking of reciprocating ones. i probably should've said that at some point. A turbine would be great but i don't see anybody whipping a steam turbine up in their garage. I see plenty of poeple making reciprocating ones, however. I'm sure there are simple ways to improve them, most of the usable (non-toy, non-model) ones are almost all cast iron, old style designs.

Dr. Hess
Dr. Hess MegaDork
1/30/17 3:30 p.m.

The first (10HP) steam turbine was made like 100 years ago with technology that would be about what you could come up with in your garage today. Probably a lot more primitive than what you can come up with in your garage. If you are talking casting stuff, it certainly is doable.

T.J.
T.J. UltimaDork
1/30/17 3:55 p.m.

AFAIK, steam turbines cannot be 100% efficient. That makes zero sense from a thermodynamic perspective. The turbine itself can be very efficient at a certain speed, but you have to look at the overall steam cycle. I think 50% would be a good efficiency for one.

Dr. Hess
Dr. Hess MegaDork
1/30/17 4:14 p.m.

A multi stage steam turbine runs in the mid to high 90% efficiency range. I'm not talking about the boiler, just the turbine.

mad_machine
mad_machine MegaDork
1/30/17 5:54 p.m.

rather than reciprocating steam engines, why not try it with a rotary?

KyAllroad
KyAllroad UberDork
1/30/17 7:10 p.m.
mad_machine wrote: rather than reciprocating steam engines, why not try it with a rotary?

A steam wankel? That's GRM!

Dr. Hess
Dr. Hess MegaDork
1/30/17 7:25 p.m.

Here's one of them "rotary" engines:

Rufledt
Rufledt UberDork
1/30/17 8:29 p.m.

I understand how awesome turbines are, i just don't want to build one of those, i'm still thinking reciprocal here. A steam wankel would be cool, though. I bet you could make up for it's lack of torque! I wonder how you could do the ports for early cut off and no overlap. Side ports? You could probably make each side of the peanut it's own high pressure zone, kinda like how reciprocating ones can have pressure on both sides of a piston.

Appleseed
Appleseed MegaDork
1/30/17 8:41 p.m.

A hobby guy in a garage can fabricate a steam piston engine. I doubt the precision and metallurgy needed for a turbine can be fabricated in the same shed. Lots of broken dreams are paved with lofty, unattainable goals. Start with a simple piston engine and work up.

RossD
RossD UltimaDork
1/30/17 8:44 p.m.

What about a 2 stroke engine? Or just the turbine portion of a turbocharger?

MadScientistMatt
MadScientistMatt PowerDork
1/31/17 8:04 a.m.
mad_machine wrote: rather than reciprocating steam engines, why not try it with a rotary?

I have a reprint of a book called "507 Mechanical Movements" from around 1900. It shows a lot of rotary type - and even some downright Wankel-esque - steam engine designs.

Another thing that might be fun to work with would be an Ericsson cycle engine. Closely related to a Stirling engine, it's kind of another attempt to translate a Carnot cycle into a real device.

fasted58
fasted58 UltimaDork
1/31/17 8:28 a.m.

What about a Pelton (spoon) type rotor? Designed for water but modified could see it handling steam. Wouldn't have the torque as a piston engine but oughta spin like a mofo.

and what RossD said, be careful playing w/ steam

You'll need to log in to post.

Our Preferred Partners
HfHYljCcwvcYMbtl9Z7yKU1Gof2X0BIfHetY9n4on19SoczHh36RqDQinFlvyH5C