Modified expansion engine?

[Rudy from Norway]

I need some advice from you guys.
I’m planning on my next project. It will take a year so I better do it right. I would like a steam engine. Twin or triple with lots of details and the style very often used in medium size steam boats. That would be for instance the Stuart Twin or the Stuart Triple Expansion. I like them both. Preferably the size of the triple since many details are easier to build if the engine is a bit larger. If I do the Twin I would add some of the features from the Triple anyway.
However, there is a catch (If I don’t go for the Stuart Twin, non expansion though). I want it to run flawlessly on both steam and compressed air. From what I have figured out, an expansion engine will fight it selves and need rather high pressure to run on air. And if the sound I have heard of these on YouTube is right, it sounds like they are not totally happy.
So.. I figured. What if I made the Triple expansion engine with equally sized bores and run the input and output ports in parallel? Not as an expansion engine but as a normal one? I would have to sleeve the larger cylinders, but who would know? The Triple has a vacuum pump I believe is an “inverted boos”? It sucks the exhaust out of the last cylinder? This will probably not be possible to do If I don't run it as an expansion engine. I expect much more exhaust in addition to that the steam will continue to expand after being spent and therefore the vacuum pump would probably not catch up? Am I right? I could of cause make the vacuum pump just as a display feature and leave it empty inside. Is it anything else I haven’t thought of here?

Rudy

 

[Charles Lamont]

Rudy,

The Stuart triple has an 'air pump'. This pump is intended to work with a condenser. The exhaust vacuum is created within the closed condenser vessel by condensing the steam, not by the pump. The function of the pump is to then remove the liquid condensate from the condenser. It is called an air pump because it is also required to remove the small amount of air that inevitably leaks in, for example throught the low pressure piston-rod gland.

The pump is not intended to evacuate the low-pressure cylinder directly - that just would not work thermodynamically, because the pump would need more power to drive it than would be gained in the cylinder. It does work with condensate though, because the condensate volume is so much smaller, it needs much less power to drive the pump.

 

[Jasonb]

If I were doing it I would do away with the cylinder casting and make a new one from solid so that the HP can be increased in size, you could then have all three cylinders at 1.25" and then plumb it up to run as a triple high which will be easier as you will be able to place the inlet bosses where you want them rather than have messy pipework trying to use the cast ones.

 

[Peter Twissell]

I'm not sure why the expansion would be any different between steam and air.
If anything, the hot steam will cool in the engine and expand less than compressed air.
Sure, steam can be condensed and provide a vacuum at the exhaust, but fitting a condenser also muffles the exhaust sound, so it's not clear what you are trying to achieve.

 

[Cogsy]

I'm with Peter - I've never read anything that would provide a reason for why compressed air would expand less than steam, except the change of phase from liquid to gas of course. They are both gasses under pressure and should behave similarly.

 

[Rudy from Norway]

Hm.. that gives meaning Cogsy. I have been thinking that steam continues to expand after leaving the cylinder. However, compressed air leaving the cylinder will actually also have pressure. When the exhaust port is opening, the same pressure that drove the piston is actually let right out the exhaust.
However, if cold air and hot steam did have different properties would not surprise me. So many say the expansion engine runs much smother on steam. If that has to do with timing, I don't know.

Rudy

 

[Charles Lamont]

Heated air would behave more like steam.

However, there is not much use in people speculating what might happen. Meaningful conclusions require some study of things like steam tables and ideal gas laws, an understanding of concepts like enthalpy and adiabatic expansion, and then doing some calculations. This is all (mostly first year) engineering degree thermodynamics.

To anyone who thinks this stuff would be easy if only they knew how, I would say that it might be easy for some people, but I gained a modest mechanical engineering degree (a 2.2 or 'desmond') at one of the world top-flight technical universities, and I don't find it particularly easy.

 

[Rudy from Norway]

Thanks for explaining about the pumps Charles. I'm starting to get the picture. Intricate stuff this external combustion idea.
Rudy

 

[John Antliff]

I think the main difference between compressed air and steam is the amount of heat energy contained in the gas. The expansion ratio is directly related to that heat energy and steam contains far more energy (latent heat of evaporation) than compressed air thus the work done by steam is also far more. I have had some experience with testing pumps and other equipment with both mediums and what works on compressed air often does not work on steam - the reasons are many and I don't understand all of them. To increase the power produced you need to increase the heat/energy available (steam pressure directly relates to heat available) hence the use of superheaters on steam engines. To increase the power available on ICEs you need to increase the compression ratio. That's my twopence worth!

 

[Peter Twissell]

It is of course true that a hot gas contains more energy than a cold gas, but that energy can only be converted to motion by a change in temperature of the gas.
In a steam engine, that means generating vacuum in a condenser, increasing the delta pressure between incoming steam and exhaust.
Otherwise, gases broadly follow Boyles law.
In the example of an IC engine, higher temperature means a greater delta between the cold induction charge and the heat of combustion, creating a greater delta pressure.
In a steam engine, a certain amount of heat is required to boil water, turning it to a gas. That is the least efficient part of the process. Superheating increases the gas pressure with greater efficiency.
All that happens in the boiler. It is only the resulting pressure that can be turned into motion in the cylinder.

 

[Andy Munns]

There is a lot to consider and was often a lifetime study... Yes the air pump (or extraction pump if a closed feed system) is functional only for steam plants as it pumps air and condensate from the condenser. Air will leak in from any leaky gland in the entire plant that is exposed to exhaust vacuum. Air also arrives dissolved in boiler feedwater as most older steam recip feed systems were open to the atmosphere.
Compressed air (N, O, CO2) technically behaves as a mix of superheated gases and there is no condensation to liquid during power stroke. Steam, unless considerably superheated, will condense on contact with colder cylinder walls and this results in a serious loss of power. Compound and triple engines reduce this loss by decreasing the temperature range/loss in each cylinder - This mainly benefits marine engines (firstly when a steamship could steam across the Atlantic without sails and without the cargo space filled with coal fuel). Factory, generator and power plant engines adopted compound expansion to increase fuel efficiency. Compound and triple engines were also smoother running and easier to start and reverse, plus individual parts were smaller compared with single cylinder engines of same power.
Big 3 cylinder steam rolling mill engines (look up River Don engine!) plus winches and winding engines remained simple expansion as economy was sacrificed for instant starting in both directions.
Stuart triple and compounds are marine engines and reflect operation on salt water where feed water had to be carried.
Locomotives remained mostly simple expansion due to the loading gauge issues with big LP cylinders.

 

[Jasonb]

It has been know for the exhaust to ice up on a triple run on air as the LP ends up sucking the air through the engine. Valve timing that works on steam is not the best for air even on a single cylinder so really needs to be set differently for air only. I have seen several tripple expansion models run on air but they needed alterations such as leaving the HP piston or valve out or modifying the pipework so at least the HP and IP both saw full pressure air.

 

[Cogsy]

Heated air would behave more like steam.

However, there is not much use in people speculating what might happen. Meaningful conclusions require some study of things like steam tables and ideal gas laws, an understanding of concepts like enthalpy and adiabatic expansion, and then doing some calculations. This is all (mostly first year) engineering degree thermodynamics.

To anyone who thinks this stuff would be easy if only they knew how, I would say that it might be easy for some people, but I gained a modest mechanical engineering degree (a 2.2 or 'desmond') at one of the world top-flight technical universities, and I don't find it particularly easy.

I agree it's not simple, but they're both gases so let's assume they are close enough to ideal gases for the ideal gas law to apply. Then PV = nRT and since you mentioned adiabatic and we have an unchanging mass of gas in the cylinder, then nRT is constant for a given intake of gas (let's use nRT = x for simplicity). So we have PV = x which we can rearrange to P = x/V. Now if we double the volume we halve the pressure or double the pressure halve the volume. This is the same for either of the working fluids so I still can't see a major difference in the two.

I'm wondering if the real difference may be expansion of air in the cylinder due to the heat of the steam that is introduced? Of course this is not adiabatic but in the real world I don't think there is a 'true' adiabatic process. It's been too long since I took thermodynamics for me to even take a stab at a meaningful enthalpy discussion.

 

[Andy Munns]

You can take indicator cards on larger engines, and those with slide valves will show imperfect diagrams. The process of balancing power distribution between HP, IP and LP is complex and counter-intuitive. Read Youngson (best) for the detail. If it is that difficult with full size triples (Youngson even details the requirements for length and angularity of the connecting rod), not even going to try on a tiny model. That said, you could plot theoretical indicator cards and calculate mean effective pressures in each stage, and then calculate powers, but why? They did this on big engines because the fuel saved was significant and the engine ran better and required less bearing work. Also the cylinder that ran the pumps and was set to give a slight power increase.
Also running on air that contains moisture may show icing - A refrigerating response to expansion of a gas through a nozzle, but perhaps take a leaf out of the tyre industry and experiment with moistureless Nitrogen gas?

 

[Rudy from Norway]

Thanks for all the input guys.
I did place an order for plans for the Stuart Twin Launch and the Triple Expansion.
I have some ideas on how to make the cylinder block and also some other mods I want to do, so a kit is probably not necessary.
Planing for a bit of time in the shop these days, so a hybrid Stuart will probably be the result.
Rudy

 

[Mechanicboy]

Icing on steam pipe between HP and IP/LP is visible due the air don't expanding as steam can do it, the IP/LP has larger volume than HP then it create vacuum by IP/LP. It cool down very effective.

 

[Mechanicboy]

The airpump who are pumping the condensate water out of condenser and create vacuum in the LP cylinder in exhaust side. The correct name is Edward's air pump.

Look at this how it works under test.



Picture how the Edwards air pump are build. Piston on way down while the valves are closed and create vacuum. When the piston are in bottom then the ports is opened then it suck water into the cylinder from "W" and create vacuum in condenser/LP cylinder in exhaust timing get vacuum then the LP piston is sucked up to TDC each time LP piston is pushing the used steam out of cylinder. Then the piston in Edwards air pump on way up and the valves are opened to get condensate water/air out of the air pump.

 

[TRIPLE]

Hi Rudy, have you seenthe blog of Johnsmachines.com? He build a triple that runs on air and on steam, very nice made.

 

[Andy Munns]

I teach full sized marine steam students that the vacuum is created by the condenser as steam contracts back to water. This vacuum is harmed by air that leaks in and occupies a huge volume. On small plants the single air pump must primarily remove this air, but it also is tasked with removing the condensate. Large plants often had an extraction pump designed for air plus a condensate pump designed for water.

 

[peterl95124]

what they don't tell you in high school (or even college) is that
PV = nRT is only half the story, the other half relates to "work"
E = Tin * Cp * ( VR ^ (1-gamma) - 1)
E = energy or work
Tin = input Temp
Cp, Gamma = coefficients depending on gas
VR = volume ratio, ie expansion ratio
the upshot is that the work done by expansion is directly proportional to the input temperature, and steam is generally way hotter than compressed air and thus does more work for the same expansion.

in a piston engine combustion increases the pressure, so its seems intuitive that it should work in that the power stroke has more force than the compression stroke took. but in a turbine engine combustion doesn't increase the pressure (if it did the air would flow back out the intake), so how does it work ???. in this case combustion increases the temperature, and that increase in temperature means that the turbine can extract more mechanical energy from the gas than the compressor took to compress it (at a lower temp). And the above equation (that they don't teach you in high school) lets you calculate those amounts of work.

I don't know how Cp and G for air compare to steam, and there's the added complication that for steam those values are fairly temperature dependent, hence "steam tables" for those wanting to do the math, perhaps someone reading this will have that info on hand.

HTH, Pete.