Engine conversions

We all know of the Mercury outboard and VW Beetle engine conversions, but does anyone have any ideas for modern engines that might have features which would make them especially interesting for conversion to steam?

One that springs to mind is the Yamaha YZF R1 engine. This is a 1,000 cc (61 cid) engine rated between 150 and 200 HP (obviously at some pretty high RPM). What makes this interesting compared to any other inline 4 is that it has a crossplane crankshaft. A normal inline 4 crank has pins located at the 0 - 180 - 180 - 0 degree positions whereas the Yamaha is at 0 - 90 - 270 -180, just like the typical V-8 engine.

This might be desirable for a steam engine because they produce a power event every 360 degrees rather than every 720 degrees as in a 4-stroke engine. Converting the typical 4 cylinder engine leaves you with two cylinders reaching TDC simultaneously ... which puts a lot of extra stress on the crank at TDC and denies you the smoothness you would receive if the admission events were evenly spaced out. By contrast, the crossplane 4 should give you that smoothness and reduction in peak torque across the crank.

The advantage of the single plane 4 crank is that it has no primary unbalance, although all the secondary unbalance forces coincide for a vertical shake unless the engine is fitted with dual counter-rotating balance shafts turning at twice engine speed, the cross-plane has a strong primary rocking couple but no secondary vibration. Yamaha recognized this and fitted a single counterrotating balance shaft spinning at engine speed.

Anyone else have some nominees?

My vote is for an opposing piston design like the one used on the German U-boat.

Hi Rick,

Is there any engine having that configuration available for people to convert?

Another option for conversion might be an OMC V-4 powerhead. The V4 is nice because you get evenly distributed power strokes, a very compact package, and reasonable balance. You could either run as a uniflow using the 2 stroke ports, or resleeve like Art did with the PSL engine.

90 degree V-4 2-stroke

OR, there's a 4 stroke conversion that wouldn't need sleeving. It's a 60 degree V-4 rather than a 90 degree and I suspect the power delivery and balance will be a little less smooth -- but undoubtedly acceptable or they wouldn't be manufacturing them.

60 degree V-4 4-stroke

Regards,

Ken

Ken,
To answer your question, yes. One just needs to create an adaptor plate to receive a 2-cycle, cylinder head top and bottom. In other words, bolt two, 2-cycle engines head-to-head. Adapt a type of gearing to connect the 2 crankshafts. A chain or cogged belt could do the job.

The 2-sided adaptor head will introduce the steam and it would be a uniflow in 2 different directions.

This concept provides for decent expansion that seems to help with a steam engine for torque.

I haven't taken the time to develop this adaptor plate. However, the concept is sound. Every revolution provides a power stroke with the expansion of ~ 1 to 14-16. Karl Peterson presented a piston concept from Cyclone at a SACA Meet in the past. This compression release piston could be used effectively with this simple arrangement.

Attached is the Achades opposing piston diesel engine. Envision replacing the fuel injector with a steam injector.

Achates Diesel Engine

This engine has demonstrated superior efficiency and would also do the same with steam. I'm not aware of any existing steam engines of this design.

While visiting one of the maritime museums in Virgina, I saw the engine in the Monitor steam battleship, civil war. Designed by John Ericsson to minimize its height within a ship with low profile.

Monitor Steam Engine

This engine inspires the opposing piston idea, for me at least.

Rick

Hi Rick,

By a compression release piston, do you mean something like the attached patent drawing (William J Divine)?

The piston has a piece of flat spring steel on top (24) which is lifted by a pin (23) actuated by a rocker on top of the conn. rod (30) which allows steam to flow through the holes (21) in the top of the piston.

I had looked at something similar many years ago. (THRU PISTON). I used a poppet rather than spring steel because high superheat would anneal the spring steel. Not seen in this view are passages in the plane of observation that carry steam around the sides of the piston down to the uniflow ports.

Regards,

Ken

Yes Ken,
Those two compression release concepts look like they would work. I haven't heard of any production application except for Cyclone. And that shouldn't count I think. My guess is that the draw back would the total piston mass would be high and cause problems at higher rpm's.
Rick

Hi Rick,

My biggest objection is that you are still getting a bashing effect that I would prefer to avoid at all costs -- the motion at the wrist pin isn't large enough to get a useful eccentric or cam motion. Devine used the second pin to operate a bash valve in the head and I never really took the time to work out the motions. Both the piston and the conn rod rocker are approaching TDC and their velocity should be dropping rapidly -- I tend to wonder if the arrangement can be worked out so that the pin velocity is very low as it contacts the bash valve and then accelerates as you pass center. If that worked, it would be a novel way to minimize impact forces.

Ken

Hi Ken,
I was just down visiting Billy Barnes and looking over his motorcycle LSR engine. We have a question that I'm sure you know the answer.

What is the name of the offset crank to the piston alignment? We noticed that when the crank on this horizontal engine is at 90 degrees, the piston is not in the center of the travel. What is this referred to as or it's name?

Some pictures for everyone to enjoy!

Thanks,
Rick

Hi Rick,

I'm not exactly sure to what you are referring. In some engines, the wrist pin is not set in the direct center of the piston, that is referred to as piston offset. In engines where the crankshaft is located slightly sideways to the plane of the cylinders, we call that a Desaxe engine.

If you are referring to the fact that the piston travel is not equal to the crankshaft angle (for instance, the crankshaft is at 90 degrees but the piston is at 60 percent of stroke), then that really doesn't have a name and is just assumed from the formula for calculating piston travel. (note, the formula is easily found on google, but it only works if the engine is neither Desaxe nor has an offset wrist pin). I have heard the term indicated rod length tossed around as a means of compensating for crankpin and rod angularity during the stroke, but that figure would change with every fraction of a degree and it really doesn't seem all that valuable.

With EXCEL, it is easy to figure out the crank angle vs shaft movement, or there are online calculators.

Regards,

Ken

Thanks Ken,
Here is the reason why we are asking. On my 10 HP Stanley engine and Billy's 20 HP Stanley along with most all seasoned Stanley's, the eccentric angle is set 135 degrees to the crank angle. Forward is set back 90* plus 45* to add up to the 135*. Same for reverse in the other direction. This allows for 50% cut-off w/out hook up. Billy had me use a paint stick to mark piston position when we saw the positions on the valve. Because of this angle difference between 90* crank and piston position in the cylinder, I calculate a .569 or round it up to 57% cut-off. Note that this is theoretical on his engine and given that the valve would be machined to allow steam into the cylinder at TDC.

Note that Stanley valves are machined at two (2) different lengths. The shorter length goes to the top and the longer to the bottom of the cylinder.

Billy's valve is highly modified compared to the standard valve that would go into the Stanley engine. His is machined to allow pre-admission and limited travel during hook-up. His actual cut-off's at Top: .419 and Bottom: .263. His pre-admissions are Top: .018 and Bottom: .016. These can all be expressed as percent piston travel down the cylinder. In the case of pre-admission, it is the travel before reaching TDC or BDC depending on which end you are at. The standard valve was installed on the other cylinder and it didn't even open at the hook-up setting.

Billy confirmed that all eccentrics, angle setting, are the same on Stanley's except for the later ones. He said they changed it and it is actually easier to time. We aren't sure what changed.

This was a fun and interesting trip to perform this evaluation of the engine. All the methods for evaluating this stuff was in Billy's repertoire.

Rick

Hi Rick,

The relationship between crank angle and piston travel is dependent upon the ratio of crankshaft throw to connecting rod length. So, we could see a difference in valve events if they kept all the valve gear dimensions proportionately the same, but changed the ratio of crank throw to rod length.

The formula is x = r * cos A + ( l² - r² * Sin²A) where:

x = wrist pin position
r = crank throw (1/2 crank stroke)
l = conn rod length
A = crank angle (from TDC)

Honestly, the simplest way to analyze Stephenson's Link valve gear is to use the Charlie Dockstader valve gear program:

Dockstader Simulators

This program not only gives you a nice animation of your valve gear after you have entered all the dimensions, but it also shows admission and cutoff diagrams for both ends of the piston. You can change any variable and test it until you get the results you desire, measured at any cutoff. Sometimes you get surprises, such as finding out that a design that looks wonderful at full cutoff won't even work at short cutoff. It is certainly faster than trying to work everything out on paper or using CAD.

Regards,

Ken

I appreciate the effort for the simulator. I find it better to just use excel and simulate it step by step or angle by angle. Just my preference.

The main thing we established is that most all Stanley Engines have eccentrics set up for 50% cut-off. However, the actual percent is greater due to the 90* angle of the crank is ~57% of the piston travel.

So many already sophisticated responses!
I don't know if someone did this, but I had an old BMW motorcycle with a single, standing cylinder (from 1955) and so many times I thought how easy it would be to modify it to run on steam. You basically just have to build your own cylinder head and construct modified timing gear wheels and remove the cooling ribs of course.
I also had a 50cm³ 2-stroke moped and thought it would be especially easy to create a crude, not-so-efficient but working steam engine with a bash valve.

But my take is, that 2-stroke cylinder liners with their ports are not a good way to build something half ways efficient and 4-stroke, portless liners are much better.
There ARE very few (2) nice youtube videos out there with car engines running on steam. But the builders used the normal cams as I would guess, making them very inefficient.




I doubt in any of the videos a proper "steam" cam shaft has been built for the project and the second video lacks due to the simple fact that the boiler is way too rudimentary for the big car engine.

I think the other answers are already much more sophisticated than my simplistic examples, but I wanted to show them in case someone was interested. I think a project like this requires furthermost a proper valve mechanism, properly constructed for steam AND as much thought for the boiler as for the engine. There are a lot of youtube videos out there of steam engine conversions where they just throw a small compressor tank filled with water into a camp fire and the engine ends up not performing as well as it could - even with makeshift valve solutions.

Steam Captain: I enjoyed your other post where they needed a plunger to get you off the ship.

What I would call the classic double acting steam engine that requires a cross head, piston rods, crank rods and counter weighted cranks, well it's bin developed over 200 years now. The Serpollet Engine resembles an internal combustion engine and was reported as an excellent engine with labyrinth.seals on the poppet valve stems. However, the life of the car produced was short.

I too wonder how to implement the valve timing as driven by a cam. The possibilities are great for you can command many different scenarios in the cam design. Here is a couple of thoughts in a strategy for an efficient, smooth running and single acting engine like an IC conversion. Do you provide for compression to account for the mass of the piston and rod to slow down and change direction. Do you want to have pre-admission of steam because you are accounting for wire draw. Wire draw as a reminder, is the effect where you cannot get full flow of steam when the valve initially opens. Then there is the all talked about cut-off and how one takes advantage of expansion of the steam leading to engine efficiency.

Oh, so much to figure out. I have a high regard for this fellow, Jay Carter. He converted a VW engine to a bash valve configuration and shown excellent efficiencies. The rpm's of this engine as compared to a similar "Classic Steam" engine are quite different. The double action provides for much more torque. How can one increase the torque on an IC engine conversion.

Here is a shot in the dark. Take a 5 HP lawnmower engine and remove the head. Take a weedwacker engine and remove the head and cylinder. Make a wolf compound head that will go on the top of the 5 HP and hold the weedwacker cylinder and bashvalve head; on top. This will be a steeple compound. It will be like a mini-Skinner compound. Just some thoughts. Think about the weedwacker performing expansion on the rod side of the piston.

This is a conceptual sketch for the cylinder head I was designing for an LSR automobile. It took some head-scratching to figure out how to use a single camshaft to drive an inlet valve that lifts to open and an exhaust valve that drops to open. I can't say that I had ever seen an engine with the valves set at 45 degrees to the plane of the cylinder bank ... but that let the geometry work itself out.

A new concept has come to me that should reduce the moving mass to permit higher valve accelerations for shorter cutoff or higher rpm. The problem is that it is taking more calculations in order to correctly specify springs that will remove potential lag times while holding forces to a desirable level.

I'm also trying to work out the specifics needed to make everything far more modular so that machining errors will affect just one component rather than the entire cylinder head.

Ken

Yes, yes the plunger
I just remember there are double piston engines. These might be moddable to ise one piston as the inlet timer like a 2-stroke, and the other one as the outlet timer. That sounds interesting. Although no adjustment possible

Edit: On analyzing the piston movements with function graphs, I can tell, that split single engines are, without any valve, are either massively inefficient or have abysmally low volumentric energy density. Thus not a good choice. So better forget what I said



Edited 1 time(s). Last edit at 01/29/2023 11:36AM by Steam Captain.

Compression release by a valve in the piston head actuated by the rocking of the connecting rod is very attractive. The forces on the moving elements must be considered when the rpm rises as they are likely to cause early fatigue.

Karl Petersen

Wilhelm Schmidt had a very interesting uniflow aux exhaust valve that is operated by compression. See patent.

Regards,

Ken

Interesting article on piston coatings:

Piston Coatings

Hi Ken,
What's your thoughts on this conversion?

1969 SE-124 Chevelle Converted to Steam by Besler. One of the few profitable ventures into steam powered automobiles.

Tom Kimmel - Chevelle Conversion

Kind regards,
Rick

Hi Rick,

We have to view that car in context. Besler really didn't receive enough money to develop a powerplant for the car, so he had to cheat matters in order to bring it in under budget and still make some profit. I doubt he made much.

He cut a V-8 in half and affixed double acting cylinders to the top. This would make the engine very unbalanced, but it was a proof of concept and we can assume he would do far better if scratch building.

A V-8 has all crankpins set at 90 degree angles, which is all wrong if you want to use two of these to drive piston valves. Besler dealt with this by fabricating a really ingenious linkage that phase shifted the motion. I found it while helping Tom organize Crank's files, and was really impressed. I found some schematics and other materials related to the car, too.

Another set of limiting factors was that they were not allowed to extend the engine compartment and were required to retain the air conditioning as well as power steering and brakes. Suffice it to say that space was at a premium.

Part of the reason that the car was underfunded was that GM was undergoing Senate hearings led by Edmund Muskie, who had brought in the Williams brothers who testified that their steam car ran like a rocket and had better fuel economy than a Beetle. They demonstrated the car, but it never went on a dyno during the hearings, or things might have turned out differently. So, GM put their main effort into SE-101 which had an engine based on the Williams patent designs, but built better. They put a lot of effort into controls, burners, boiler, and so on. This was their prime test vehicle because it could prove or disprove the contention that the Williams Cycle achieved better than Rankine efficiency. (It didn't -- classical thermodynamics won another round).

The SAE has reports on SE-101 and the test comparison of SE-101 and SE-123. Suffice it to say that neither car was all that impressive by 1960s standards. SE-123 was not even able to complete some of the testing because there wasn't sufficient room under the hood to fit a boiler properly sized to the engine -- the car simply couldn't reach the government's driving cycle speeds. The condenser was also inadequate due to space configurations. The specifications called for full condensing, which neither car achieved, but SE-101 turned in a respectable performance for a steam car while the Besler machine simply didn't have the cooling capacity to even make a good attempt.

Honestly, the Besler car was not very good -- but that was a foregone conclusion when the project started -- both the funding and the specifications precluded anything else. I have seen a letter written by Besler at that time stating that he was going to do just enough to deliver a vehicle, but he saw no future in the project and refused to take a loss on it by engaging in expensive development.

Now, everything in the world is relative. If this was a SACA member's project car, it would be utterly outstanding. The car drove smoothly and was reliable. The controls were automatic and by all reports functioned flawlessly. It was clear that Besler was a pro and understood what he was doing. That said, the Doble Series E was certainly a better car and the Series F, which Besler worked on, was better yet -- but neither had all the inherent compromises forced onto the GM project car.


Regards,

Ken



Edited 1 time(s). Last edit at 03/17/2023 10:15AM by frustrated.

How about given that Besler made the half a V8 into a compound, isn't the 90 degree crank angle a good thing? It was a receiver type compound.

I think the engine is a good conversion. It was condemned due to the specification constraints. It was as though they wanted a steam engine to be like a internal combustion engine. What if they were to open up the requirements to make the steam engine to work within its attributes.

Perhaps a clean sheet design would have lead them down the road similar to the Doble E or F but with anticipated improvements. I don't think the Doble car layout would be acceptable to the stereo typical GM car designer. This might have cemented the idea that it wasn't worth it to dive into the new steam car development. Hence Besler thought...just get what we can out of this exercise.

What if the platform was more open. Would the V8 conversion be good in this situation? I do like that it is a compound and uses piston valves. I know that you prefer the poppet valve and single acting with high rpm to develop HP. With that aside, I would say it's a respectable conversion.

There was nothing wrong with putting the whole powerplant under the hood, rather than scattering it around the chassis like a Stanley. This was becoming the norm with the last steam cars in the teens and twenties. It simplified manufacturing, reduced costs, and met the expectations of customers who had become accustomed to gasoline engines.

The V-2 had a rotary primary unbalance that could be canceled with a counterweight -- although Besler could never have managed that since the bolt on double-acting cylinders and piston rod added too much mass for the counterweight to handle unless, maybe, it was heavily slugged by Mallory metal. But even then I'm suspicious that there isn't enough counterweight since the inboard V-8 weights are necessarily smaller than the outboard (it's complicated). Even when perfect balanced, it has a secondary rocking shake. Since secondary shakes reverse every 90 degrees, the piston valves will partially add to that shake.

The thing to keep in mind was that this was intended to be a Doble technology car -- that was part of the contract. I suspect that was because the Doble was a known quantity and it would establish a baseline.

This use of piston valves is simply problematic. We need to inject oil into the steam, which has the potential to char in the boiler and cause tube failure. So, we're adding hardware that makes our boiler problematic. We also get foug efficiency losses, the first simply being that steam leakage is a thing. Secondly, that oil injection limits our steam temperature, which also degrades efficiency. Next, piston valves add clearance volume which ... you guessed it. Finally, we have admission and exhaust steam using the same passages, which effectively bypasses heat around the engine. The 1960s weren't the 20's and expectations had moved on.

The other option is what Art Gardiner was building when he died. That engine is in Tom's shop and I spent a few days reviewing Art's notes and can talk about it somewhat (the notes cover years and various engine iterations). Anyhow, like Besler's engine, it is half of a V-8 although it is a Ford 302 or much more recent vintage than a GM product. Six of one, half-dozen of the other, small block engines have been around so long that convergent evolution is highly visible. Basically, it was an improved version of the PSL engine with a vee layout and additional cylinder. Art had reversing and variable cutoff, which complicate the engine. Besler was using the car's transmission, which simplified matters greatly. Honestly, his linkage was brilliant but trying to make it reversible and variable, while housing it in the crankcase, would have been highly problematic. Besler couldn't have used that cutoff V-8 with piston valves without a transmission.

At this point we could only speculate. Art's engine was probably significantly more fuel efficient due to short-ish cutoff, higher steam temperatures, and separate admission and exhaust. He eliminated the need for a transmission by using a more complex valve gear, but I tend to think the transmission was a better solution since you don't need to make a lot of compromises in the engine operation. I think that was the one part of the contract stipulations that actually worked in Besler's favor.

We really have NO idea what Besler might have done if he were challenged. He was getting a little older and less energetic -- his letters point that out. He didn't have a budget. His mandate was to build a modernized Doble powerplant. He had to fit it inside the existing engine bay and retain power steering, brakes, and A/C. Under those conditions, I think he took the most prudent route with his design. There was no way that basic architecture could have been developed into a viable replacement for 1960s gasoline engines -- but he wasn't hired to do that. Realistically, he didn't have the resources to undertake a job of that magnitude. He was supposed to build something modern based on Doble technology and I think he did it in an economical and straightforward manner. (Except for that valve linkage, that took some brain sweat and wasn't straightforward. I think it was a great solution to his contract and also think that it would make a nice build for a hobby steam car -- if owned by someone who really appreciated the demands of steam and didn't try to operate it in the same fashion as an ICE machine.

Regards,

Ken

Hi Ken,
We haven't discussed the 3 cylinder Diesel engine conversion that has been toyed around with for awhile now. This is the engine that might be considered for an LSR attempt and know that you have done some considerable valve/head design. I believe this is a GM design also.

For the same reasons you state above, wouldn't this be problematic?

This Diesel conversion would be single acting in each of the 3 cylinders. RPM would be relatively high to achieve HP and geared down to get the torque out of the engine to the wheels. Higher RPM means more steam usage. The Besler engine conversion has significantly more torque at lower RPM than the Diesel conversion. I think this had a lot to due with the undersized boiler under the hood and to making it work. The GM transmission had a lot to add to this boiler deficiency also. However, the longer stroke in the Diesel does help to develop torque.

Wondering if there was any thought to compound the Diesel conversion? This exercise would include a poppet valve system also to allow a receiver type compound. The Skinner design comes to mind.

Any ways...food for thought.

Rick

Hi Rick,

There's certainly nothing wrong with the 353 as a steam conversion. It is a fairly old engine that generates its torque at lower rpm, and it is of an industrial nature, so it is essentially unbreakable.

Fortunately, they installed a balance shaft.

Sorry, operating at higher rpm does not inherently increase steam usage. It all comes down to the expansion ratio. We can run smaller engines at higher rpm and generate the same power, for the same steam consumption, if the expansion ratios are about the same. In both cases, we are taking one cubic inch of steam and expanding it to the same final volume. There might be a little more friction at higher rpm, but there will be less steam blowby past the piston rings, so one hand giveth and the other taketh away.

There was never any need to contemplate compounding, the 353 has uniflow exhaust ports. Thermodynamically speaking, this is inherently superior to any counterflow arrangement since cool steam is not passing through the head. Compounding also generates an interstage pressure drop which is pure lost power -- look at any indicator card of a classical compound engine and you will see a vertical gap between the low end of the high pressure trace and the high end of the low pressure trace. Then again, there is also a potential for loss due to radiation away from the plumbing and turbulence in the flow. The uniflow engine just eliminates these problems. It also eliminates the need for a receiver or an interstage charging valve.

We did add an auxiliary exhaust port to the cylinder head to remove the small amount of steam left in the cylinder when the piston covers the uniflow ports. This delays the onset of compression, allowing us to minimize clearance volume for higher efficiency AND higher peak power. Note I said it "vents a small amount of steam" -- yes, that will cool the head slightly, but not anywhere near the amount that you would see from a counterflow engine -- the reduction in clearance volume more than makes up for it. Note that Scott-Newcomb, Endurance, AC Staley, and GE arrived at the same conclusions. That was where my big contribution came in, I figured out how to get both valves to operate off of a single camshaft -- therefore keeping the engine simple. Otherwise, Tim Nye and I both had almost identical plans for the steam admission valve. I had a few other minor inputs, such as using solid aluminum stanchions for the camshaft and running them with pressurized oil -- we do that with gasoline engines and avoid the need for extra bearings.

Dutcher Industries taught us another problem with compound engines. When using very high pressure steam, they blowby badly. This is because the expansion ratio in the HP cylinder is low, and the mean effective pressure is quite high. I had asked Jay Carter, Sr. about his observations and he noted that blowby also went up as the expansion ratio dropped -- but he never went as low as a compound. By contrast, a uniflow can use a much faster cutoff and the mean effective pressure is much lower -- giving the steam far less opportunity to work its way around the piston rings.

Regards,

Ken

Hey Ken,
If OK I would like to further explore this Diesel conversion and understand some of its attributes to allow it to rotate at a higher rpm and still consume equivalent steam of a Besler compound. I'm sure you don't mind and it will get a little technical.

With piston valve engines, there is lead and lap. Lead breaks down to an advancement of the eccentric angle to allow the lap to work. They are related and the main focus is to have the lead and the steam lap to correlate at the point where the engine piston is at TDC. This is where the valve just opens. Now to correlate this to a poppet valve.

My understanding of a cam shaft and poppet valve arrangement is to perform the same admission at TDC. Also note that both piston and poppet valve can admit steam before TDC and reduce clearance volume. Also, an exhaust valve can close before the piston reaches Bottom Dead Center (BDC) to also reduce clearance volume. Engines that are configured with cross-heads use this to help with the reversing the extra mass used with the cross-head arrangement. Probably not necessary with a single acting engine without cross-head like the Diesel conversion.

How would you set up the percent of stroke before TDC and after TDC to allow the engine's steam usage to be equivalent to the Besler? What about the percent of stroke on the exhaust valve? In summary, how does this correlate to the piston valve arrangement. Note that I see by the Doble history that they started with a 4 X 6 double acting D slide valve engine. Then went to a Uni-flow and then to a Compound. I wonder why they left the Uni-flow Engine. Besler continued the compound in the SE123 Chevel conversion. The stroke on the Doble Engines started at 6" and with 5" along with in compound. What is the stroke on the Diesel conversion?

A couple of side effects that take away from the classic steam engine design when setting up the Diesel conversion. One needs to push start or use a starter motor with this arrangement. It is not self starting. The other consideration is to use a clutch because the engine needs to idle in order to maintain the piston valve set-up. The Doble's made it a stern point to always have the engine self starting.

This might make for a good discussion at the Spring Meet. What ever we come up with could be a neat presentation.

Again, food for thought...

Kind regards,
Rick

Rick
You may want to read a book by L Francis Herreshoff about his brother
The Title is Capt Net Herreshoff the wizard of Bristol.
The book is mostly about the sailing ships and steam ships that Net built.
But there are sections in the book on his great steam engines and designs. Triple expansion and valve settings.

As Jim Crank said Net was one of the best Steam engineers in the era.
Harvest Book Co P.A. 800-669-3397

Rolly

Hi Rick,

Honestly, for a poppet valve engine, you just don't worry about lap and lead -- you aren't trying to get a single piston valve to do two different things while being driven by a single harmonic motion. There's something to be said for using two different motions to do two different jobs. Note that really big and efficient piston valve engines often had separate admission and exhaust valves with separate valve gears.

Regarding exhaust, if your uniflow engine can run a good vacuum exhaust, such as on a steamboat, you simply make your clearance volume quite tiny and you don't worry about compression -- virtually all of the steam leaves the uniflow port and you have almost nothing to recompress. On a car, where you don't have access to a good coolant like the North Atlantic, you can either make your clearance volume larger to ensure that you aren't overcompressing the steam or you can add an aux. exhaust valve to delay compression onset. The second option yields higher efficiency and greater power.

As per Stumpf, your goal is to recompress your steam to the same pressure as the incoming steam which offsets much of the clearance volume losses. Stumpf is often misquoted on this point as people claim in eliminates these losses, but reading his book carefully, shows that the recovery is not total. My own spreadsheet shows the same phenomenon.

So, when designing our valve events, we need to pick whatever throttle pressure is most important to us and then choose to set the aux. exhaust valve closure at whatever angle will produce compression back to admission pressure. As per GM SE-101 -- and Franklin Railway Supply -- the valve is nominally closed before it hits the seat, so we need to advance the closure by something like 5 or 10 degrees -- depending upon valve dimensions, cam profile and rpm. Ideally, you work it out as close as you can and then run on a dyno while taking indicator readings to fine tune your result.

Honestly, a little overlap between admission and exhaust won't matter since your compression eliminates the pressure differential between the steam chest and the cylinder and we won't see a lot of steam movement until after TDC. An added benefit is that this reduction in differential pressure allows your poppet valve to move without imposing high axial loads.

Likewise, we set the cutoff at whatever produces the best efficiency and still guarantees the power we need. Besides cutoff, we can either make a bigger or smaller engine or change rpm somewhat. Steam consumption is going to be based on the cubic inches of admission volume per minute. Longer cutoff will produce more power, but also throw a lot of that away,as well. Cutoff that is too short will produce diminishing returns since friction MEP will start eating into the already low MEP we get from high expansion. Art Gardiner was looking at 15 to 20 percent, and given the fact that his little Chrysler conversion almost took the land speed record away from a multimillion-dollar British venture, he has some credibility.

I really could care less about Doble and self starting, whatever makes the car run is just fine. His allergy to electric starters would ring truer if he didn't use so many other electrical components in the steam plant. The Jay Carter powerplant used a starter motor and, as far as economy and power to weight ratio went, it utterly curb-stomped every Doble powerplant. If we look at the complexity of making an engine self-starting versus the hassle of buying a starter at Auto Zone and just bolting it in, then the starter looks like the winning proposition.

Let's face it, making an inline triple self-starting is already problematic, so it's easier to just not make unnecessary problems.

As far as a clutch goes, or even a gearbox for that matter, we only need to refer to Jim Crank. I exchanged hundreds of e-mails with Jim and one thing on which he was explicitly clear was that the White was a superior driving machine to the Doble, or the Stanley, and he had extensive hours in all three. He largely attributed this to three factors:

1. The White flowmotor control better regulated the steam generator than Doble, almost as well as Stanley.
2. The clutch allowed you to idle the engine and thereby keep the pumps on line when stopped.
3. The low speed on the gear box allowed you to run the engine faster, and thus spin the pumps faster, when operating in high-load/low speed situations.

Jim told me that he spent every parade in which the drove the Doble staring at the temperature gauge, hoping it wouldn't climb too high as the water level dropped in the steam generator. The same held true for city traffic or long uphill grades. You can see photos of Stanleys stopped on the side of mountain roads with one wheel jacked up so that the engine can be turned over to operate the pumps at higher speed and shorter engine cutoff in order to refill the boiler. In the White, operating conditions that might run the steam generator level down were addressed by simply shifting the gearbox to low and keeping your attention on the road while the flow motor controls kept things right. Popping the clutch and letting the engine idle was also advantageous when forced to stop for more than a moment -- the steam lines and engine remained hot and there was no need to vent condensate before restarting.

When it comes to poppet valves, I have a hybrid design that is part of the engine that I am designing, one element of which is meant to allow faster and crisper opening and cutoff. As I have found in the last couple of weeks, I essentially, once again, reinvented someone else's basic mechanism. In this case, it was actually built and run for a few years in a small stationary powerplant -- and is still apparently running in a steam launch. That reduces the risk a bit. It will be covered in the next Bulletin .. I wrote it up the article last Thursday night.

Regards,

Ken

Hi Ken,
Let us put self starting, clutches and even transmission aside for this discussion. Let us focus on the valve events that produce the steam rate or the effective use of steam. In other words, to make the best use of the steam from the boiler/generator.

As per Stumpf, the objective is to use compression to match the steam chest pressure. I would totally agree that Abner Doble did read and study Stumpf's work on the Uni-flow Steam Engine. The other objective is have a new lower percent range for cut-off than what was typically acceptable to the stereotypical steam engine designer. So the range of 20 to 30 percent travel would be a achievable event for a counter-flow arrangement. Where a uni-flow would entertain single digit percent. However, the uni-flow would also use longer cut-off in starting or accelerating from a slow speed. The intent is to use the full cut-off spectrum in operating a uni-flow.

As we all know that the Doble Detroit was a big failure. The uni-flow engine was an attempt to implement many of Stumpf's developments in engine efficiency and to utilize steam from the generator at it's upmost conservation while maintaining supreme performance. The engine was double action and with a 6 X 5 bore and stroke respectively. Interesting to know that Doble used a 2-part D slide valve in this engine. The idea was to allow the valve to lift in order to achieve the reduction in compressive work back into the steam chest. I think this to be an excellent idea.

Crank states that the lack of vacuum had a lot to do with the failure. Also, the operator wasn't used to the special needs of the uni-flow in operation all the levels of cut-off. I think that the ocean vessels with uni-flows is a good fit. I was totally impressed with the Badger and the Skinner engine. As Rolly mentioned, the vacuum is the ticket.

The other thing worth mentioning is the relationship between bore and stroke is reversed in the Doble Detroit's Uni-flow. This is a unheard of deviation from classic steam engine design. Although this is acceptable in a compound but not so good for a single expansion engine.

Another concept worth exploring is the dynamics of the steam engine with the mass of oscillating parts with the rotating parts. The oscillating parts need to start and stop, changing directions. With engines set up as a cross head, the mass is significant and worthwhile to include exhaust lap on the valve. In the case of a poppet valve in a cross-head arrangement, just implement more compression to account for the force from the oscillating mass. I know this first hand from my Steam Scooter design and in implementing 1/8th inch exhaust lap on the D valve to help with not only steam conservation but with dynamic force balancing.

Again, what is the bore and stroke on the Diesel conversion? Also, it will be interesting to review your hybrid valve design.

Kind regards,
Rick

Hi Rick,

To quote an online source:

"The Detroit 3-53 is a two-cycle engine, which features a bore by a stroke of 3.875 by 4.5 inches, or 98 mm by 114 mm. The total displacement of the engine, as suggested by its name, is 3×53, equaling 159 cubic inches, or 2.61 liters. Comprising of three cylinders, the compressions ratio of this engine is 21 to 1. The engine offers a power output of 101 horsepower at 2,800 rpm, while it is known to provide a continuous gross power of 70 horsepower at 2,400 rpm. It produces a maximum torque of 205 ft-lbs at 1,800 rpm."

Note, however, that people have packed turbos onto 53 Series engines and run them far above their ratings, for a full season, without rebuild. That kind of over-engineering is probably why Detroit Diesels from WW2 are still in widescale use throughout the world. For a street machine (i.e. daily driver) I would want something considerably lighter -- but lightness really doesn't matter at Bonneville where acceleration isn't much of an issue. In fact, given the track's relatively low coeffficent of friction, it can be argued that extra weight is a net benefit.

You can find the repair manual here:

REPAIR MANUAL

And a brief poop sheet here:

POOP SHEET

I have to note that Jim Crank told me that his original intention was to set the Land Speed Record using a modified 353, but one of the pricier aluminum marine versions. He went with the Lear turbine when that became available at a ridiculously low price. Jim also told me that, if he'd had to do it over, he would have gone with the Detroit Diesel since the turbine rig was so problematic.

Regards,

Ken



Edited 1 time(s). Last edit at 03/23/2023 06:47AM by frustrated.

All files from this thread

File Name	File Size		Posted by	Date
William J Divine.png	149.4 KB	open | download	frustrated	12/07/2022	Read message
THRU PISTON.png	17.1 KB	open | download	frustrated	12/07/2022	Read message
20230114_111044 reduced.jpg	943.2 KB	open | download	Rick.H	01/15/2023	Read message
20230114_111221 reduced.jpg	934.4 KB	open | download	Rick.H	01/15/2023	Read message
20230114_114905 reduced.jpg	828.7 KB	open | download	Rick.H	01/15/2023	Read message
Stephenson Linkage Settings reduced.jpg	858.5 KB	open | download	Rick.H	01/15/2023	Read message
20230115_082127 reduced.jpg	649.6 KB	open | download	Rick.H	01/17/2023	Read message
Crank Throw Angle.jpg	483 KB	open | download	Rick.H	01/17/2023	Read message
Eccentric Angle.jpg	515.9 KB	open | download	Rick.H	01/17/2023	Read message
10hp Valve Simulation 180 Deg.png	51.5 KB	open | download	Rick.H	01/18/2023	Read message
head.png	208.2 KB	open | download	frustrated	01/19/2023	Read message
pat562039.pdf	213.1 KB	open | download	frustrated	02/20/2023	Read message
ENG0023.jpg	66.7 KB	open | download	Rolly	03/25/2023	Read message
DSC_0008.jpg	125 KB	open | download	Rolly	03/26/2023	Read message
267534.PDF	137 KB	open | download	frustrated	03/26/2023	Read message
3.5 x 7 x 5.5 compound.JPG	228.5 KB	open | download	Rolly	03/28/2023	Read message
20 HP box frame.jpg	191.1 KB	open | download	Rolly	03/30/2023	Read message
Hydro test at 2500.jpg	85.4 KB	open | download	Rolly	04/03/2023	Read message
20210503_184335 reduced.jpg	932.4 KB	open | download	Rick.H	04/06/2023	Read message
20230309_181848 reduced.jpg	595 KB	open | download	Rick.H	04/06/2023	Read message
20230309_182057 reduced.jpg	634.4 KB	open | download	Rick.H	04/06/2023	Read message
2023-04-06 16-08.jpg	480.5 KB	open | download	Rick.H	04/06/2023	Read message
2023-04-06 16-08_3.jpg	231.2 KB	open | download	Rick.H	04/06/2023	Read message
2023-04-06 16-08_4.jpg	418.1 KB	open | download	Rick.H	04/06/2023	Read message
IMG950547 (3).jpg	199.4 KB	open | download	Rick.H	04/07/2023	Read message
20221004_143535 reduced.jpg	624.7 KB	open | download	Rick.H	04/07/2023	Read message