Designing a new engine for a big(ger) boat...

[fredrosse]

Good work Bart. Your graph shows the net result of the two power strokes per revolution, with their interaction. Would it be possible to post the character of each side of the cylinder (head end and crank end) before they are summed?

In order to baseline your analysis, better predicting the actual expander efficiency that might be expected the following three actual Uniflow Engine test data information might be of interest:

The single acting Uniflow Engine of the Domestic Heat-Power Module (1980s) is somewhat smaller (23 CID), and got 21.1 pounds of steam consumption per brake horsepower-hour. That engine was running with atmospheric exhaust, 140 PSIG saturated steam, and used about 20% clearance volume. It produced 4.64 BHP at 1040 RPM.

The White Cliffs Solar Project used a 3 cylinder single acting Uniflow Engine (1980s), I think with 53 Series GM cylinder sleeves. This engine produced the highest efficiency I have ever found (in real practice) at 1500 RPM, 600 PSI steam pressure, and bash valves! Admission was so short that the bash valves actually held up well after adjustments got by teething problems. Steam consumption and data for this plant is on the internet. Some technical publications of about 100 years ago give higher efficiency for test engines, but those results were never actually applied to real working engines.

Officially sanctioned Ohio tests of a Baker Traction Engine, 9-1/4 x 10, done about 100 years ago, is attached.

[Lopez Mike]

I think you would be making a strategic mistake in designing a one cylinder engine with complicated reversing arrangement thus tying yourself to the weird rudder. I'd keep my options open. Is a twin that much less efficient or is it the added complication? A twin would certainly have better balance.

Mike

[barts]

I think you would be making a strategic mistake in designing a one cylinder engine with complicated reversing arrangement thus tying yourself to the weird rudder. I'd keep my options open. Is a twin that much less efficient or is it the added complication? A twin would certainly have better balance.

Mike

What's leading me in the single + Kitchen rudder direction is:
* To be self-starting from any point, I need either a twin with maximum cut-off of over 50%, or a triple
* shifting cut-off on the fly reliably means either three-d cams or a variable phase mechanism ala Skinner
* engine is significantly longer or wider, more bearings, cross heads, etc. Since the engine is in the
parlor, so to speak, keeping things less spread out is an advantage. I'm (so far) not having castings
poured for this project, so getting two twin cylinders close together is more of a trick. I'd really prefer
to bore the cylinder weldments for the liners on my 15" lathe rather than on the Bridgeport.

The back-up plan ( ) for the Kitchen rudder is a electric jog for the engine, and a shifting cam lever
in the pilot house (only a few feet from the engine, but...). I figure everyone with a single now relies
on a manual jog for reversing, anyway. But I really think that Kitchen rudder will work as others have
found.

- Bart

[barts]

Your graph shows the net result of the two power strokes per revolution, with their interaction. Would it be possible to post the character of each side of the cylinder (head end and crank end) before they are summed?

Sure; I'll do each one separately and post the results - simple enough. That's how I started.

In order to baseline your analysis, better predicting the actual expander efficiency that might be expected the following three actual Uniflow Engine test data information might be of interest:

The single acting Uniflow Engine of the Domestic Heat-Power Module (1980s) is somewhat smaller (23 CID), and got 21.1 pounds of steam consumption per brake horsepower-hour. That engine was running with atmospheric exhaust, 140 PSIG saturated steam, and used about 20% clearance volume. It produced 4.64 BHP at 1040 RPM.

The White Cliffs Solar Project used a 3 cylinder single acting Uniflow Engine (1980s), I think with 53 Series GM cylinder sleeves. This engine produced the highest efficiency I have ever found (in real practice) at 1500 RPM, 600 PSI steam pressure, and bash valves! Admission was so short that the bash valves actually held up well after adjustments got by teething problems. Steam consumption and data for this plant is on the internet. Some technical publications of about 100 years ago give higher efficiency for test engines, but those results were never actually applied to real working engines.

Officially sanctioned Ohio tests of a Baker Traction Engine, 9-1/4 x 10, done about 100 years ago, is attached.

I'll throw those into my cycle analysis program and see how things compare (predicted vs actual); that's a good idea. I wrote two programs
for this analysis - one to examine the torque, and the other to break the steam cycle into admission, expansion, exhaust, and compression.
I'm a little unsure how to treat compression work after I reach the steam chest pressure since it gets pumped into the steam chest and we get that back, but so far I've just treated that as lost energy - it doesn't make a big difference, and a slight increase in clearance brings the peak compression pressure reached below that of the steam chest pressure.

I have several technical books (Marks, Kent, Stumpf) from circa 1920 that list big (200 + hp) engine results (measurements from working engines), many w/ superheat, of course. The steam figures w/o superheat seem to align pretty well. My design is pretty much the same (jacketed head, insulated cylinder) as seemed to work well then, so we'll see.

- Bart

[barts]

I've spent a few evenings going over my model of a uniflow engine. The larger clearances definitely need to be taken into account in order to get results that approximate recorded reality; the power needed to compress the left-over steam in the cylinder was also important.

The best improvement in accuracy came from calculating the additional steam needed to raise the pressure in the clearance space from the pressure after compression to the inlet pressure. My steam table software doesn't provide a lot of different lookup routines, so I ended up coding iterative solutions to compute amount of fresh steam needed to raise compression space to inlet pressure while making sure total mass * specific volume of mix was the clearance volume. Since thermodynamic state is independent of path, it was much easier to model admission as two steps - steam is added until clearance volume is at full boiler pressure, then piston moves and we add steam until cut-off occurs. It then expands until we uncover the exhaust port.... and then we compute the quality of the remaining steam, and compress it back into its clearance space.

For the Heat-Power engine, the model predicts a steam consumption of 22.65 lbs/ihp-hr at 4.66 hp @31% cut-off, rather than the 21.1 measured.
The numbers are pretty sensitive to exhaust port height which I guessed as being pretty low (5/16").

For the White-Cliffs engine, the situation is reversed - the model predicts a better result than was obtained, which feels better to me. In this case
I'm sure it's due to difficulty in modeling bash valve and remaining steam; the model predicts 6.6 lbs/ihp-hr at 36.3 hp, 1.9% cut-off rather than the 8.95 lbs/ihp-hr actually measured. There's also likely to be significant heat losses with this engine since it runs so hot.

The Baker engine has far more data available; I found the original paper from Transactions of the ASME v 39 on Google. I did not change the model to reflect the aux. valve; the results (lower power and more water than predicted) likely reflected both this and the valve gear effects.
I also had to guesstimate clearance. Across the tests, the model was consistently optimistic - but sometimes much less so than other times.
I wonder if these valves are small for the RPM being attempted.

I'm not sure if i know more than I did earlier... but I had some fun doing it, and I'd like to think that the model is more accurate than before. I think the Sea Lion's s engine will likely produce around 10 hp net at 5% cut-off and 200 psi steam; this should push her to between 6 or 7 knots depending on weight. Right now I'm planning on 5, 10 and 15 % cut-off cam positions (11.3, 18.7 and 24.8 ihp, respectively) plus reverse at 15%.
Of course, I used a fixed RPM of 400 for calc. purposes but the 28"x28" prop shouldn't speed up too much even at maximum cut-off .

Time to get back to the mechanical design aspects.

- Bart

[artemis]

Unless I've missed it, you don't seem to have "looked at" David Ayers article on "A Uniflow Engine" in the Mar/Apr 2008 Steamboating Magazine. It's a single cylinder uniflow with poppet valves and well worth you perusing same. While you're at it there's a very good article on "A Loo Brush Weed Hatch" in the Sep/Oct 2010 Steamboating Magazine.

Maybe I can coerce you into writing a detailed (multiple installments) article about the building or your boat - something ala Peter Cuthbert and the Francis Ann. He went for a good many months without paying for an issue of Steamboating with the free-issue-per-article temptation. And with the refinements you'll pick up here, It'll make fine reference material for a lot of steamboaters around the world!

Give it a thought! (and I'm not a pushy publisher or elitist editor)

[barts]

Unless I've missed it, you don't seem to have "looked at" David Ayers article on "A Uniflow Engine" in the Mar/Apr 2008 Steamboating Magazine. It's a single cylinder uniflow with poppet valves and well worth you perusing same.

Cough wheeze.. I hadn't vacuumed the pile of Steamboating magazines for a while... but I found David's article. I now remember reading when it came out. I particularly like the starting lever - nothing like few hundred ft-lb kick to start a steam engine off of dead center. Thanks for the reminder!

The weed hatch is also a nice touch... but I'll need one a bit bigger, I think.

I will definitely document and document the process of designing and building both the boat and engine. I did write a couple of articles back in the day on Otter's firebox & burner .

_ Bart

[artemis]

[quoteThe weed hatch is also a nice touch... but I'll need one a bit bigger, I think.

I will definitely document and document the process of designing and building both the boat and engine. I did write a couple of articles back in the day on Otter's firebox & burner .

_ Bart

Get a bigger "loo brush".

I will wait in anticipation for the articles. Your earlier stuff (going back some years) was really great. I know they will be appreciated/enjoyed

[87gn@tahoe]

Ron,

It's funny how quickly we forget. I've just stumbled upon some articles in issues from 2009 that pertain to issues my father and I heve been tackling for the past few weeks.

[artemis]

Ron,

It's funny how quickly we forget. I've just stumbled upon some articles in issues from 2009 that pertain to issues my father and I heve been tackling for the past few weeks.

Now you know why I keep suggesting that people subscribe to the Steamboating Magazine. Not a lot about meets, and shows, and camping trips, and photos - almost ALL technical/how-to articles. Save 'em and review 'em from time to time. And, yes, a full, complete, up to date Steamboater's Handbook on CD with search and index is in the works. Completion date is still uncertain.