Design of Reciprocating Piston Expanders

Guys,

I think it takes talent and enough money. But not too much money. Too much money seems to build organizations and drowned talent. It doesn't take that much money to get things built and tested. I only spent about a grand on my condenser design and it worked better than anticipated.

Paper engines quite often don't work as hoped but if it doesn't run well on paper it sure won't run well in steel. This phorum is where we might get some of the paper right. Lately a lot of good information has been brought to the phorum, and thanks for it.

Best,

Bill G.

Regarding poppets and variable steam pressure, I was always surprised by the use of automotive type poppets, since on big steam engines they used balanced poppets, similar to this picture:





Edited 2 time(s). Last edit at 01/15/2015 11:57AM by dullfig.

It's partly the complication of the valve and partly the increase in clearance volume.

Check out this link to the Duke, a British loco that uses one.

[www.dukeofgloucester.co.uk]

Caleb Ramsby

Bill,
Indeed, one hell of a lot of good information has been on our web site and in our Journal and Newsletter.
The great pity is that no one has gone through all of it and organized it with a reliable index so we can look up something without having to plow through all the back issues and website stuff.
Guess everyone is too busy on their computers designing dream steamers that never get built and arguing esoteric theories that never get reduced to practice.
Jim

Hey Jim,

Here is a quote from you, which was posted July, 05, 2003.

"Caleb,
You still need to go to the nearest college library and work your way through the old books on steam engine design, that is if they have not thrown them out for lack of interest.

You can't find the design information on Besler's Kaiser engine, I don't think the blueprints even exist any more. It should be well described in old SACA newsletters; but since we don't have an index, you would have to search all the early issues."

That was not much more then a year after I first got into steam stuff. I took your advice and then some!

I just wanted to point out that you have been bitching about the lack of an index for a long time and have done nothing about it.

You have also been bitching for years and years about people yammering away about theory and design and not building stuff, yet to the best of my knowledge you have never made a car with a powerplant of your own design either.

I must admit that it is extremely painful for me to read my posts from back then, the combination of ignorance and arrogance which I displayed in my early 20's is not something that I am proud of.

Caleb Ramsby

Caleb,
Want to bet?? When you are busy totally redesigning the E Doble engine and building eleven of them, who has time to compile an index? Prattling abut theory doesn't get a car built either, does it?
Care to drive by some day in your new steam car.
Jim

Hey Jim,

"Care to drive by some day in your new steam car."

Sure, in theory I would. . . that's a joke folks.

". . .redesigning the E Doble engine and building eleven of them, . . ."

This is exactly what I am talking about. You have stated on numerous occasions that those big Doble compounds were too heavy and expensive for what they did and that near the end of his life Doble said that a simple counterflow engine would give a better economy overall then a compound. So just why have you spent your expertise on those big chunks of metal instead of using it to make YOUR dream car. I believe that one iteration was a long Jaguar with a three cylinder engine under the hood, the cylinders splayed left, right and up, with a big Lamont boiler and an exhaust turbine draft booster.

My personal interest has shifted from an automobile to a motorcycle, more of a challenge from an engineering standpoint, but many fewer parts to deal with.

Would I be better off if I had spent the last couple of years doing what Cyclone has. . . or will mentioning them in a negative manner get this post deleted?

Caleb Ramsby

I thought the reason for componding had more to do with evening out the power pulses when using high pressure steam, not so much with efficiency. I suppose it also allows really long expansions, otherwise the steam is hardly pushing towards the end of travel

Dan



Edited 1 time(s). Last edit at 01/15/2015 08:28PM by dullfig.

Dan,

No, compounds are less smooth then single expansion engines on an overall basis. The only way to make them smooth is to adjust the cutoff of each stage for that given inlet pressure, temperature and rpm. Which in a car is always changing.

Here is an excerpt from a link bellow.

" In the Fundamental Design Calculations for the 5AT (see FDC 1.3), Wardale gives figures of minimum indicated Specific Steam Consumption for the Duke of Gloucester as 12.2 lb/hp-h and for the SNCF 141P Class 4-cyl. compound 2-8-2 as 11.2 lb/hp-hr, as compared to 11.24 lb/hp-h (= 5.1 kg/hp-hr or 1.9 kg/MJ) for the 5AT."

[5at.co.uk]

Here is a quote from Jim Crank from 12 years ago regarding compound engines.

"Compounding is not as efficient in a car as is short cutoff."

That is from the Stanleysteamers forum.

Hey Jim,

I dug up one of your "dream" cars.

[stanleysteamers.com]

The above is more evidence that I had ZERO clue about what I was talking about back then. . . I was also idiotically dreaming of inventing a "radical new engine design", that was before I knew better.

2002 or so was when I got into steam, via research on my favorite engine of the time. The Wankel! Said research led to Andy Pattersons website which has the compound Wankel idea on it and also to your posts on the Stanleysteamers website regarding the steam powered Wankel engined car you were "working" on back then. Here it is now 12 years later and I have put in well over 20,000 hours of research on the subject. My greatest dream is that my research pays off so that when I do. . . if I ever do get the motorcycle prototyped it will not be a repeat(as so many of the light steam powerplants have been) of someone elses mistake.

My observation has been that you pinned your hopes for a hot rod on the people at Cyclone.

Which do you hold in higher esteem Jim, the results from Cyclone or those of Lear?

Caleb Ramsby

I've been reading David Wardale's critique of the caprotti poppet valve system, and apparently those "double beat" (twin seat) poppet valves have leakage problems. That wouldn't be good. On the other hand, apparently piston valves can be made pretty steam tight. Sometimes what seems "obvious" in steam is not necessarily so. I gotta read some more.

A piston valve could certainly be driven by a voice coil arrangement instead of a valve gear. Force wouldn't change with pressure or back pressure.

Dan



Edited 1 time(s). Last edit at 01/15/2015 10:30PM by dullfig.

Hi Guys

Although the posters on this forum are mostly dreamers, including myself, there are a great number of viewers who are doing real things.

I put forward Roger White who has built a steam powered Land Rover 4x4. Admittedly it is not a racing car, but neither was the original petrol version. The engine is based on Doble's triple expansion but Roger did all the detail design, drawings, castings and manufacture. The new solid fuel monotube generator is all his own work and the electronic control system. Does this qualify as a real project or another bodge job?

He also manufactured a compound version which was dynamometer tested then installed in a commercial electricity generating plant in Ireland.

Just to show the LandRover works here are two videos of it in action. It is still in development so we test drive on a small hill (for Scotland) which really tests the system. In these videos he is using propane but last year he tested a wood fired system which has proven superior and I hope to get a video in the spring.

[www.youtube.com]
[www.youtube.com]

Brian

Caleb
You would never convince me of that. I’ve been building and running double acting compounds for over fifty years and they all run nice and smooth. In fact on the one in my big boat it sat on rubber engine mounts and never shook of vibrated.
Rolly

Hey Rolly,

We are talking about two different types of smoothness. You are referring to a mechanical balance, I am referring to the balance of torque pulses from the engine. All of your compounds have been in boats where the prop is "slipping" in the water, that is the water is acting as a fluid "torque converter", in effect smoothing out the torque pulses. Your engines are also very low rpm and setup by an expert(you) to balance the torque load at the cruising rpm. At least that is what I recall you posting about your compounds.

The Whites had a very fast gearing to the axle, depending on the model around twice as many revs per mile as the Stanley. One reason for this was to keep the engine bulk down, the other was to "blend" the torque pulses by making the engine take tiny steps as opposed to the long legs of the Stanley. Both had their ups and downs.

The White had to sit there and spit out water for five minutes after starting from cold. The Stanley's frames would tear themselves apart if pushed too hard for too long.

Hey Jim,

Thank you so VERY much for sharing your experiences with Lear and Cyclone. You are the only one here with enough guts to do so!

My first question mark about Cyclone showed up when they had an "indicator diagram" on their website for the Mk-5. I had to sign a whole bunch of "clause BS" to simply view it. Then it turned out it was just a theoretical indicator diagram! It wasn't even complex enough to include the heat theoretical heat transfer to and from the cylinder walls, nor the resistance by the inlet valve ect. just a simple "grade school" theoretical indicator diagram, which was treated as gold and an absolute truth. The early efficiency numbers appeared to be quoted from said "indicator diagram", utter nonsense.

I have studied that Gillenwater bike, it sure was a screamer for its time! I for one would love to see that footage.

I seem to recall that in your youth you raced motorcycles, so you would know a fast one when you encountered it!

Caleb Ramsby

Why the Cyclone "indicator diagram" hit me so hard is that I recalled reading in Doble's notes that when he was testing different valves in his E type engine one would look much better on paper, then they built it and tested it on the dyno, it ran better on the dyno, then they put it in the car and drove their set "test route" which was 150 - 200 miles, something like that and would discover that it had less performance and ended up using more fuel. They also had the opposite happen where a valve looked worse on paper, ran worse on the dyno and performed better on the road, then there were some that performed as expected. I don't tend to forget things like that.

Expectations do not equal results.

Caleb Ramsby

Celeb
Yes I would consider them low RPM. My big boat engine under power she would run about 450 RPM at 275 PSI. I used a two speed transmission to step up prop RPM by 1.6 times, this keeps the pitch to diameter under 1.3 this also act as a disconnect to the engine so I could run at the dock without pushing the boat all over. Disconnected I’ve run the engine well over 700 RPM with no shack. Compounds need superheat, I run 200F over saturated. Without superheat there a dog.
Rolly

Hey Rolly,

I remembered Jim mentioning the balance of the two cylinder Doble F on the road, I looked it up.

"The two cylinder compound Doble Series F engine also had this problem in the first engines. Bill Besler told me that the engine was so out of balance that it shook the windshield of the Buick 3/4" back and forth at 70 mph, and this was a sedan.
Besler and his crew added two pie shaped counterweights to the crankshaft to try to eliminate this and according to all information, this did help balance the later F engines.
The White also has this problem to a lesser degree, as they added a massive counterweight to their crankshafts in the 1909-10 engines. "

and

"I seriously doubt 2200 rpm, 1200-1500 yes.
Besler did add two counterweights to this one engine, as he said: "It shook the windshield back and forth 3/4" when you stepped on it at 70, (Long pause here) and that was a sedan.""

The second quote was in reply to Nutz

"Langdon's "F" Doble engine was the one reworked by Besler and had extra counterweighting applied as(according to JIM) the original "F" had some real balance problems, not mysterious at all. It also had extra rings on the piston valve to cut down on valve leakage. I doubt if it could obtain 2200RPM, its horsepower curve was already dropping off at 1500RPM/piston speed over 1200FPM."

One way to get balance with a compound is to use three cylinders, all the same bore and stroke, with the center one being the high pressure and the two flanking cylinders the low pressure ones, that dictates a 1 to 2 ratio of hp to lp cylinders. This technique was used in locos, as was having the two flanking cylinder being the hp ones and a massive center one being the lp one. Then there was the Mallett with two pairs of engines, one pair being hp the other lp.

I should note that with the locos the compound engines were often refitted(as were some of the Malletts) to simple operation. The uniflow locos were also taken out of service and replaced with counterflow engine.

One thing that has always stuck with me is that there were NO compound shunting locos. Shunting is where the engine sits around in the yard idle half the time and must push around the cars at low speed to put them on the right track for the regular locos to hook up to. The compound would get cold and waste a LOT of steam trying to get hot again during shunting service. This is very similar to city driving.

Balance is an exponential deal. I don't think that there is an engine more suited to marine service then a triple compound. I also don't think there is one more suited to an road vehicle then a simple counterflow. A case in point is that you didn't design a compound for the Stanley you built.

Caleb Ramsby

Caleb,
That three cylinder compound you mentioned is exactly what Bill Broebeck used in his steam bus.
It was quite smooth and was the most successful one of the three busses constructed for that program.
The first engine used the crankshaft and connecting rods from a 3-71 Jimmy Diesel. It bent the rods into an S shape from the torque. The rebuild used the two rods side by side from a 6V-71 (three throw crankshaft) for each crosshead and was successful. There was a complete EPA I think report on the bus system.
Jim

Hey Jim,

I remember coming across that report back when I was "hot for" compounds. That and who was it Besler that made that V-4 single acting uniflow exhaust compound with the single inlet valve. Those were the two best compounds for road use I came across.

For those who are interested the compound loco with the two outside hp cylinders and single inside lp cylinder had the hp cylinders set at 90 deg to each other and the single lp cylinder at 135 deg to the hp cylinders, it split their difference. The layout was to give the engine a self starting nature without any starting valves.

Connecting rods are an interesting issue. When I was laying out a Joy valve for my motorcycle design I had to resort to the old school oval cross section rod to get enough metal around the pin hole without causing serious stress risers. The Joy valve also drastically increased the bending moment on the rod which knocked the Joy out of contention for my purposes.

The force of steam is terrific. In my little double acting engine I am designing it for 9,000 rpm and 2,000 psi. The force from the steam is still in excess of that of the inertia from the piston, piston rod, crosshead, Walschearts linkage and connecting rod. The H rod is what I am going with, although laying out both it and an I section rod for the engine, the H rod is 20% heavier, it gives a much better load distribution at the transition to the big and small ends, which is critical when under heavy load.

Using plain bushings keeps the big and small end bores of the rods smaller. At 1,500 psi combustion pressure the Hayabusa has a loading of 20,000 psi on the small end bearing and 8,000 psi on the big end bearing, I think that at its redline the inertia force is greatly in excess of the combustion force, probably not with the supercharged ones though. I used to think that rollers were necessary for steam but after a lot of research, I just don't think that is the case. Everything from the legendary high speed F-1 engines to the massive marine diesel engines use plain bearings with forced lubrication.

Laying out the oval type of connecting rod with the Walschearts looks promising too though. The stress risers are even more reduced at the big and small ends as well as shot peening being more effective on it. It also allows for easier drilling for an oil delivery hole to the small end, which would also supply oil to the crosshead slippers.

I should explain the 9,000 rpm and 2,000 psi, the 2,000 psi would be for the drag strip and the engine wouldn't see it, it would simply be to get the most out of the boilers reserve. In daily use it would be at 1,000 psi, which would still be enough to turn the rear tire into smoke. If the engine were going over 3,500 rpm then you would get a speeding on the interstate.

I am curious Jim, with the Doble E type engines did you have to redesign the connecting rods? If so, what cross section did you end up using?

Caleb Ramsby

Hi Caleb,
The Besler Kaiser engine was designed by Bill Besler and Stanley Whitlock. Once that oil contamination problem with water from blowby was solved, it was an excellent engine. 85 hp at about 3200 rpm on the engine dyno. Tom Kimmel has it and the rest of the Kaiser now.
When Gawor bought the car from Besler, all the drawings, patterns and core boxes were in the back seat,
now missing. Who stole them we do not know. When Gawor got it, the Kaiser was a sweet running steamer, drove it many times. Vandals shot up and almost totally destroyed the car when Gawor had it.
This was the engine Bill was thinking about selling along with the rest of the components as a kit, put out a flyer to all the SACA members. The replies got him so mad he ended the idea right then. I was to build mine at the plant so he could see the problems and fix them, second generation Corvair convertible.

When I redid the E engine I did change the connecting rod, although none of the original ones ever broke,
The taper where the I beam shank blended into the actual bearing area on the original rods was very abrupt, so I went to an H section rod with a much gentler taper. Really now just standard practice for race engines, Carrillo and NASCAR for example.

I see my reply yo you about Cyclone vs. Lear got removed. Probably best we leave SACA out of this.
Your observation about the worth of the PV diagram should tell you what you wanted to know.
Yes, raced a bike once, scared myself silly, stuck to cars after that.
Jim



Edited 3 time(s). Last edit at 01/17/2015 01:04PM by Jim Crank.

Calib,
Giving a lot of thought to your steam motorcycle idea, as I once laid out the basics for one too.
Honda six cylinder Gold Wing, the sport model without the later Buck Rogers fairing. Snowmobile Wankel,
Enlarged radiator-condenser and saddle bags water tanks. Would be fun for sure.
Kim

Hello all:

been thinking about hot to put together an expander for a car. I know that car engines and steam engines have very different requirements, but quite frankly very few have the funding to build an engine from scratch. So here is my current thinking for converting an engine to steam:

take a 1.2L 3 cylinder diesel (kubota makes one for tractors and such) and turn it into a steeple compound. I was thinking that I could remove the current cylinder liners, and machine exhaust ports in them. The exhaust steam would then go through what used to be the water jacket. In other words, no water cooling, just use the water jacket to exhaust the steam.

the HP cylinders would be a sepparate unit machined from aluminum that would replace the engine head. So where the head used to be bolted on, there would now be a block with HP uniflow cylinders. Kinda sorta like Doble's Ultimax.

Has this been tried? any obvious reasons why it won't work? Would the diesel crank and rods be strong enough?

BTW, the engine is rated at 70HP. Since steam doubles the power pulses, does that mean that (given enough steam) it could put out 140HP?

Dan

Hey Dan,

For an IC to steam conversion I suggest reading up on Wetz's and Barrett's work in the old SACA bulletins.

In regards to horsepower, you are asking for a simple answer to a rather complex question.

See Jim's note above about the connecting rods of the diesel conversion failing.

Hey Jim,

Thanks for the additional info on everything.

H rods seem to be superior as to durability for anything except the highest rpm conditions, such as the 20,000 rpm V-10 F-1 era engines, then the lighter I rods trump them.

A Gold Wing sure would have plenty of room for the equipment. For my purposes I insist on a custom frame though, since I am experienced in the chain reaction that such a large engineering compromise produces. The modeling I did for the radiator/condenser shows that it really needs to be 6" thick at least and have a centrifugal pump recirculating the water and mixing it with the exhausted steam, then having the inlet to the feed pumps branched off of the delivery line from the centrifugal pump helping to jam the water through the pumps valves to limit vapor lock.

I don't want to discuss my bike design too much on here.

Caleb Ramsby

I'm back. My computer finely died Well the hard drive anyway. So I upgraded to a note 4.


Some great discussions here.

I would think that when comparing compounds to single expansion engines that the steam volume not be lgnored. You have longer cutoffs in the compounds. But blowby is not only effected by MEP area also comes into play. I would also think that blowby is related to pressure difference not MEP. MEP would apply to single acting. A double acting compound would have lower pressure difference. Small HP cylinders would have a smaler blow are and lower pressure differance.

The problem with a compound for.automotive use.is they classi, designe has a more limited efficiently operatoring power range.

Over expansion and compression both cause flow losses. The power range requirements are extreme compared to most statinary applications.

The problem that needs to be solved is speed range

Andy

Brian McMorran Wrote:

> 2) When the vehicle is beginning to move the
> mechanism will have to open the poppet valve
> against throttle pressure, there will be no
> compression to assist in opening the valve so the
> coil will have to be much stronger than an IC
> engine.
>

Ok, been thinking about this, and maybe one solution would be a hybrid concept. The engine would have a short cut-off cam operating the poppet valves, with a coil rigged to hold the valve open. So the opening of the valve would be mechanical, but at what point it close again would be determined by the coil. The advantage is in the manufacture of the cam, as it could be made by current auto cam equipment; and you would have electronic control over the admission timing.

Dan

You're describing a valve like the one in the Gelbart engine. See the picture below.

Lohring Miller

Interesting! Yes something like that, although I'm partial to opening the valves with a cam. I have my reservations about opening the valve by impacting it. That hammering cannot be good in the long run.

Dan

Hi Dan,

The idea of turning an IC engine into a steam engine by using the piston as a cross head or building it up as a steeple compound has been around for a long time; it’s certainly attractive but I think it would prove problematic. IC engine designers are very successful and they have refined their designs greatly over time; the engines are highly capable and robust but also not constructed wastefully, everything is pared down to what is needed and not a lot more.

Converting an engine into a steeple compound, or even just adding a crosshead, means there will be much greater reciprocating weight in each cylinder. The crankshaft is not counterweighted or balanced for this kind of weight; unless the rpm is kept very low, this added mass will generate quite a bit of potentially destructive shaking force. Likewise, the crankpins, connecting rods and bearings were not meant to absorb the kinds of additional peak stresses this can potentially impose on the engine.

Other factors have to be taken into consideration. For instance, traditional steam engines typically produce power at lower rpm and higher torque than do ICE’s, this should be considered carefully. Anyone driving a car with a stick knows that you have to avoid “lugging” the engine, placing a very high load on the engine at low rpm. IC engine bearings reduce friction through hydrodynamic suspension and they can be easily wiped if operated at traditional steam engine starting torque and rpm.

I think the short form answer is that anyone converting IC to steam would be well advised to design the conversion in such a way as to mimic the donor IC engine as closely as possible.

Ken

Hi Ken:

Yes, for one I think you need to keep the engine idling, the bearings are not made for stop and go. You'd need roller bearings for that.

The allure is strong to connect the engine directly to the wheels, and run the car like a locomotive. But then you need am auxiliary expander to run accessories. So you might as well just idle the engine.

Dan



Edited 1 time(s). Last edit at 01/20/2015 08:07PM by dullfig.

Ken:

I kinda figured there would be extra stress, it's why I thought i'd convert a diesel instead. Might have to make stronger connecting rods (and who knows what else). Anyhow, I won't know for sure without trying.

Oh, and as I understand it, three cylinder engines are not balanced unless the crank has counterweights. I've seen pictures on Google where you can see three cylinder cranks with counterweights. One of the reasons I leaned towards a 3 cylinder donor.

Dan



Edited 1 time(s). Last edit at 01/20/2015 08:17PM by dullfig.

I'm a little late to this discussion (so what else is new?), but the original question was about why GE built those two single cylinder engines. The deal was that was one of the contracts put out by the EPA in the early 1970s clean air program to develop alternate car powerplants. GE was to build two test engines to evaluate things like oil-free operation.

In my humble opinion, however, the GE project was a waste. They designed and built the technologically best possible test engines, then had almost no money left for any tests.

One of the things they were supposed to investigate was carbon rings to eliminate the need for cylinder oil. The little testing they did came up with nothing.

There was a report attibuted to H. Frank and J. Rais, "High Speed Steam Engine Without Lubrication", The Engineer's Digest, 2, 516 (1945) that Skoda Works in Czechoslovakia had run a 440 hp 750 rpm engine on 600 psi/725F steam for thousands of hours using segmented carbon piston rings. If anyone can find the original article, let us know. Even Tom Kimmel doesn't have this one.

The Cyclone engine is reported to use piston rings of PEEK, a very expensive engineering plastic. How well it works? There's a February 12, 2008 post from Harry Schoell saying "The rings are working great.", but I haven't found any more information.

As memory serves, the GE people, when designing their engines, talked to American Bosch about hydraulic valve actuation. American Bosch said it would never work. Then at an EPA contractor's meeting, ThermoElectron displayed the hydraulic valve actuation system American Bosch had designed for their organic reciprocating expander...

Tim

PS: Ken, have you seen patent US 2688955 "Fluid pressure engine of the uniflow type" by Harry Ricardo? Cam-operated series admission valves with variable cutoff. 20 years before the Ricardo engine for the SES car.

All files from this thread

File Name	File Size		Posted by	Date
Surface area and clearance % to bore and stroke.xls	333 KB	open | download	Caleb Ramsby	01/13/2015	Read message
XSteam_Excel_v2.6_US.xls	397 KB	open | download	Caleb Ramsby	01/14/2015	Read message
X_Steam_Tables.bas	153.6 KB	open | download	Caleb Ramsby	01/14/2015	Read message
Uniflow Engine Section Details.png	218.8 KB	open | download	lohring	01/19/2015	Read message