Harry Schoell
in search of
November 11, 2002 06:11PM
Guidelines:
Higher temperatures = Higher efficiency
Higher pressure = Smaller machine
Smaller the tube = Higher pressure
Smaller the machine, the higher the RPM.
Smaller the machine the less radiant heat loss.
Super heat will determine the expansion ration.
Steam is an insulator.
Metal is a conductor.
Water is a semi conductor.
Everything after the highest heat is a condenser.
Don’t waste Heat!!!

Professor Stump stated in 1922 the seven losses in the steam engine:
1) Losses due to cylinder condensation (surface loss)
2) Losses due to the volume of the clearance space (Clearance Volume Loss)
3) Loss due to throttling or wire drawing
4) Fiction Loss
5) Loss due to leakage
6) Loss due to heat radiation and convection
7) Loss due to incomplete expansion

If you do the same thing over you can expect the same result.

With these things in mind, and with the knowledge shared with the SACA members, who I greatly admire. I am designing and constructing three proto type engines with complete systems, based on these principals. I would appreciate comments on the application of these principals in engine design.
Peter Brow
Re: in search of
November 12, 2002 03:12AM
Hi Harry,

Look out for tradeoffs. Reducing one loss can sometimes lead to greater losses elsewhere. A few things I have been keeping in mind with recent design work:

Smaller engines have a higher surface to volume ratio, meaning more radiant & convective heat loss both externally and internally (the latter being what Prof. Stumpf calls "surface losses"winking smiley. Increasing the number of cylinders, with the same displacement, also increases these losses.

Clearance volume losses can be compensated for somewhat with exhaust compression (filling clearance volume with steam near the inlet pressure, so that live steam is not wasted filling/pressurizing the clearance volume). The compression work can be recovered on next power stroke, with the compressed steam acting as a spring. This "spring action" also helps cushion the moving parts at higher speeds. With the right exhaust compression, even an engine with relatively large clearance volume may be more efficient than one with less clearance volume.

Higher moving part speeds increase friction losses, esp at low load. As David Nergaard noted in another thread, the kinetic energy of moving parts (and the friction losses at bearing surfaces constraining them) increases at the square of velocity. EG, a given mass moving 200 feet per minute = 4x the kinetic energy/friction of 100 fpm; @ 400 fpm = 16x the friction losses of 100 fpm, etc..

Valve gear designed to reduce or eliminate leakage, throttling, and wire drawing losses often introduces enough extra friction losses to negate those savings. At 20% engine efficiency, 1 btu saved in friction equals 5 btus saved in leakage or heat loss. Some engines are designed to reduce unexpanded steam or leakage losses, but the valve gear or other design features used to do so add 2 btu worth of extra friction losses for every 5 btu saved -- meaning that the "improved" engine is actually less efficient overall than leakier, lower-expansion designs. With 10% engine efficiency, 1 btu worth of friction savings equals 10 btu worth of heat or steam savings. Often it is better to accept some steam leaks, or lose some expansion, than to add friction losses in an effort to avoid them.

Losses in one area can compound or even multiply losses (and/or costs/weights/volumes) elsewhere in the system, too. Conversely, savings can be compounded or multiplied.

If designing with possible future marketing in mind, there are also complex & difficult consumer choice and economic factors to consider. EG, a design feature which adds $500 to the cost of the system, may only save $400 in reduced fuel usage over its service life. Or, it might save $600 over its service life, yet consumers might prefer the less-efficient system/vehicle which is $500 cheaper, and pay for the extra fuel over a longer period (and if the vehicle is bought with borrowed money, this might cost less overall than paying extra loan interest on the more expensive vehicle).

"The nut behind the steering wheel" can be not only the most dangerous part of the vehicle, but also the most perplexing part to design around.

Automotive steam engines have to operate across an extremely wide range of rapidly- and constantly-varying loads and speeds, and this complicates design decisions and efficiency calculations tremendously. On the bright side, however, it is a spectacularly engrossing design challenge!

Peter
Peter Brow
Re: in search of
November 12, 2002 03:27AM
Clarification:

"At 20% engine efficiency, 1 btu saved in friction equals 5 btus saved in leakage or heat loss."

This should read "20% CYLINDER efficiency", IE, the efficiency calculated from the theoretical indicator card. "Engine efficiency", properly speaking, factors-in all losses (leakage, surface losses, friction, etc).

Peter
Peter Heid
Re: in search of
November 12, 2002 12:01PM
Harry,

A great deal of information is contained in the forums previous discussions and reading the old posts, if you have not done so yet, will give you a great deal of help. Steam has been around for many years and most design ideas have been tried and previous experience is of the highest value in avoiding repeat mistakes. I wish you well in your project.
David K. Nergaard
Re: in search of
November 12, 2002 08:23PM
Almost any text on steam engines will list the major losses in an engine. Always the biggest is also the inevitable, the energy in the exhaust steam. The second largest is the "initial condensation loss", which Prof Stumpf referred to as the surface loss. It is this loss he addressed by designing the uniflow engine. In small engines using saturated steam, this loss uses at least a third of the steam and can be as much as three fourths. Please note that "small" in this case refers to mill engines, i. e. cylinders less than a FOOT in diameter!
Automotive engines are always so small that surface effects dominate their performance.
Harry Schoell
Re: in search of
November 13, 2002 03:30PM
Thank you your comments are appreciated. Yes, there are a lot of guidelines, didn’t mean to preach to the choir. Just wanted to bring up point from professionals about thinking out of the box. There has been a lot done, however slowly. The need is at its greatest now. Fuel is not a real concern, but the poison it produces in an internal combustion engine is.

External combustion is known to be clean but could be cleaner.

Efficiency is the concern because of the cost, cost of fuel and cost to manufacture the machine. History has shown gains, James Watt, Stanley, Doble, Williams, Carter. However they have each opened doors to the next.

Mr. Nergaard, thank you for your feedback. The loss you refer to is in the total system, which is one of the areas we are exploring. I look forward to all future comments. The first part of the list is gleaned from SACA members and history.

The first and second engines, Mark I and Mark II (cyclone), are to be 8 Kw generators and the Mark III (box motor) is a simplified version for cost purposes. It will power a small cycle car. Only hold up is the machine shop, about ¾ finished on parts. Please keep this dialog going as we all learn. It is good to think out of the box.

Tom Ward
Re: in search of
November 16, 2002 06:10AM
The old professor's list of how heat is lost triggered my imagination to come up with possible solutions.
1 Surface loss: Put a steam jacket around the cylinders. This would keep the inside wall temperature at or near boiler temperature. With no temp differential there is no heat conduction. It also provides a dramatic improvement in volume/surface area ratio. The trade-off is the additional packings needed for linkages which increases either friction or leakage.
2 Clearance volume loss: Use an opposed piston uniflow design. The admission port needs to be only 1/4 as high as the cylinder bore and its piston's stroke need only be 1/2 the cylinder bore. I haven't come up with the trade-off yet. Any guesses???
3 Throttle and valve losses: The uniflow opposed piston effectively provides an admission valve with the same diameter as the cylinder bore. Instead of moving towards the piston it is pulled away directly into the steam jacket. By incorporating a Stephenson linkage between the crank and admission piston its stroke can be varied and used to control speed. Therefore no throttle. The admission piston would lag the power piston by 90 degrees. The trade-off will likely be a need for a power assist device on the linkage control.
4 Friction: Graphite, graphite, graphite. Because of its plate like crystalline structure the minute particles slide right past each other. Being pure carbon its melting point is over 5000 F. It is uneffected by water and has no effect on the steam. I would use a coil of graphite fibers in the place of piston rings. I would use roller or ball bearings in the crank and wrist pins.
The trade-off is a graphite distribution system needs to be developed.
5 Leakage: A balanced bellows approach on push-pull rods could work if a proper material could be found. If the servos can handle the temperature then a gas tight grommet would eliminate some leakage spots. The trade-off with friction will to some degree always be there.
6 Radiation and convection: Fiberglass insulation of the boiler/engine compartment may or may not be cost effective. The flow of air through the condensor to the firebox then to the flue cannot be eliminated. The pipes between the condensor and hotter parts will always conduct heat away from where you would like it to stay.
7 Incomplete expansion: Cannot be eliminated in a cost effective way at all times. Making the stroke to bore ratio larger can help if space allows. Allowing the minimum mass of steam needed for the load at that time will reduce average steam use. Compounding helps if you want to deal with the cost of extra parts.
When looking to improve fuel efficiency you need to weigh the cost of the fuel against building costs. Since I drive less than 10,000 miles per year the savings from cutting consumption in half is roughly $200. If an 18-wheeler which uses $200 of fuel every day cut its consumption in half it would save roughly $35,000 per year. That's more than most folks' take home pay. A car, why bother? An 18-wheeler, go for the max!
Peter Heid
Re: in search of
November 16, 2002 08:18AM
Tom,

I like your thoughts on this matter. I was wondering why use a steam jacket around the cylinder because steam is a conductor of heat. Maybe not a good conductor but far better than a vacuum. Can we fit a couple thermos bottle sleeves over each end of the cylinders.
Jim Crank
Re: in search of
November 16, 2002 11:23AM
Tom, Peter,
Let us for a moment step back from all this postulating and review what is factual and workable. Tom, using your post of 11-16.
1) A steam jacket is not only heavy; but useless when you consider today's insulation materials. You STILL have to insulate the outside of the cylinder, jacket or not. So why not just insulate it well in the first place?
Doble tried this with the Greyhound engine and it blew up and almost killed one of the mechanics. He then went on to even more insane ideas for the PAXTON car, proposing 2,000 psi @ 1800°F. He was serious too!!
Why put on a jacket when the lubrication of the high pressure cylinder is in serious peril at anything over 750-850°F. If there is a lubricant out there that will work at 1,000°, I sure don't know of one. Cylinders can indeed be well insulated, so you can put your hand on them when they are running, as with my Doble at 750°F.
2) The opposed piston engine gives superb breathing and minimum heat loss to the metal. With the cranks at 180° the balance is perfect. The engine can be laid flat with the power output taken from the cross shaft. Heat barrier coatings on the piston heads, to keep from throwing heat into the oil system. Putting the crankshafts in any other phase angle than 180° introduces gross balance problems and turbulant flow patterns.
You still have to d
You still have to deal with water in the oil, as with any single acting engine.
I must agree though, in a modern car chassis, it is very hard to package a double acting engine and a single acting is more adaptable, as long as you use a dry sump system with water removal.
Forget roller bearings, just simply too expensive, big and heavy inertial loads. Consider commercial sleeve bearings with pressure oiling and a tiny oil pump on the burner blower to initially pressurize the main engine bearings when starting up.
3) How do you propose an inlet valve the same diameter as the piston?? Are you thinking about a single sleeve valve like the Bristol aircraft engine?
Or using one piston for inlet and one for exhaust? Use both together for exhaust and design a good balanced poppet valve. Minimum clearnace for sure and superb breathing.
4) Before getting serious about graphite composite piston rings, investigate the use of them with oil less air compressors. Commercially available. Be very wary about the wear factor. Piston rings leak 7 lbs per hour per inch of circumference, that is why the last Doble and Besler engines used so many rings. This is a serious leak path and pay attention to it.
5) Valve stem leakage is virtually eliminated with a long stem with labrinth grooves and a graphite valve guide for minimum oil need. Look at the Serpollet valve and learn from it. Also the Skinner Unaflow design. They spent a lot of money making them steam tight with minimal friction. I have seen many of them in operation and they don't leak.
6) Why use fiberglass for insulation? KAOWOOL or similar material is superb. My Doble has 1-1/2" of such insulation on the firebox and you can almost put your hand on it after the car runs for hours. Almost, it is still hot!! Fiberglass is not anywhere as good as the commercially available insulation materials designed for our precise use.
7) Compounding is only good at one speed and with the valve timing set for that specific condition. A three or four cylinder simple engine with very short cutoff at high speed is actually superior in a car, over a compound or even a triple.
A balanced poppet with a sliding cam is very hard to beat, when you consider the cost, size and mechanical losses to drive it. The Camprotti valve gear is way to complex to even consider. The only alternative that makes sense in place of a sliding cam, is maybe a shifting eccentric driving rockers that push the valves open.
Keep thinking though, it is all good information to consider.
Jim
Jim Crank
Re: in search of
November 16, 2002 11:41AM
Tom, Peter,
I forgot something in the previous post.
Dave Nergaard put his finger precisely on the main problems with steam engines. Initial condensation and re-evaporation and energy being thrown out the exhaust. The exact problem with any engine using the same porting for intake and exhaust, a bad water rate.
The unaflow with a balanced poppet valve is still the king of the heap. If you want to consider a compound on top of a 3-4 cylinder simple engine, think about a radial inflow turbine with a variable nozzle ring, geared to the engine.
It will accept some moisture, has a good expansion rate and has been used before for this exact purpose, and it did show improvement. Adding a very good heat exchanger to the exhaust is also of great benefit.
Getting minimal loss from the porting and using all you can from the exhaust energy left over is the key factor. Again, the unaflow and poppet valves, coupled with an opposed piston engine is a very good design. The second stage being an exhaust turbine is also well worthy of consideration from a size, efficiency and packaging point of view. Been done and it works.
Get a copy of Prof. Stumpf's book from a college engineering library, "The Unaflow Engine." He explains it very well in that volume, pay attention to what is in there, it is all good data.
Jim
Harry Schoell
Re: in search of
November 16, 2002 03:21PM
Hello Tom, I m using carbon for piston liners and rings. the rings are fragle. Did you mean to screw the ring onto the piston? could you please eleborate? "I would use acoil of grafite fibers in place of piston rings"
Harry Schoell
Re: in search of
November 16, 2002 03:32PM
Hello Jim,

Do you have a problem with high temp.1200f /3200psi,other than lubrecation?
Jim Crank
Re: in search of
November 16, 2002 04:48PM
Hi Harry,
Yes, I have a very big problem with super high pressures and temperatures.
The cost of the steam generator is about twenty times what it costs if you stay lower. Price seamless Hastaloy-X tubing, provided you can even find it, or other similar gas turbine materials. The life of the coils is much lower too.
Second, the safety angle.
Third, the now astronomical cost of any component in the system. Tell me the source of check valves, solenoid valves, water pumps, or anything else designed to run at 3200 psi and anything that touches 1200°F is priced like it came from Tiffeny's emporium.
I can destroy rear axles with 1200-1500 psi and 800°F in a 4,000 pound car, I don't need any more do I?
It is attractive from a theoretical point to go super critical; but the cost is so far beyond what a real steam car would be if done right, $50-80,000.00, that I don't want to even consider it.
Besler once made a 3500 psi 1400° steam generator for the Navy. I watched the tests, even they choked on the cost back then.
I will wait until someone makes a good engine with graphite rings, runs it on a dyno for at least 200 hours and measures the wear. Then I would sure use it for rings. It is an answer providing you get good life from it.
Jim
Benson in Maine
Re: in search of'' trouble''
November 16, 2002 06:44PM
To find a steam a leak at these temp/press',,,,Tie a rag on the end of a suitable long stick/pole,,,follow along the steam line,, WHEN ya find the leak,,it will either cut off the pole or set fire to the rag,,,Most things that will burn will ignite around 750 ,,,DO NOT USE YOUR HAND ,, I will not elaborate further,,Ben
Tom Ward
Re: in search of
November 17, 2002 05:08AM
Its more like a bundle of fibers coiled many times around the piston like thread on a spool.
Tom Ward
Re: in search of
November 17, 2002 05:18AM
I imagined a jacket like the pressure vessel of a lamont system. The lower portion would be liquid whose latent heat would keep the entire container at nearly a constant temperature. The cylinder itself would be mounted entirely within the super heated area. This would possibly result in nearly constant temperature expansion.
Tom Ward
Re: in search of
November 17, 2002 06:13AM
Jim,
Your decades of experience is just what we need to refine our design concepts before wasting time and money on bad ideas that didn't work in the past. I appreciate what you share with the forum very much.
I conceive of the steam jacket as more than just insulation. It would be a reservior of stored energy near the inlet port which would reduce losses due to pipe resistance. The steam line from the boiler would flow at only an average speed avoiding surges due to acceleration. I made the assumption that line loss was proportional to the square of flow rate like it is in electricity.
I conceive of the inlet piston as mostly as a valve. Its stroke be roughly one fifth the length of the opposite piston's stroke.
The combination of water and graphite may be a matter of basic R&D. Being of a single element graphite can't decompose at any temperature it may encounter in the cycle. The main problem I foresee is it may form a coating inside the boiler and condensor reducing their performance.
If Kaowool is better than fiberglass for this situation then it ought to be used.
Short cutoff will be acomplished by varing the stroke of the inlet piston via a robust Stephenson linkage. At the point where the piston seals the cylinder it will be moving at its highest speed instead of being at its slowest as with a poppet.
I would use compounding as a power source for the auxillaries like Pritchard did with a Rootes turbine.
David K. Nergaard
Re: in search of
November 17, 2002 07:54AM
Graphite packing was used for a very high temperature steam engine in the 1930s. It was absolutely necessary to prevent any oil from reaching the graphite, so a complicated coaxial sleeve on the piston rod was used to keep crankcase oil from contacting the piston rod seals. Also, graphite rings are so brittle they will not survive going over ports in the cylinder walls, uniflow exhaust ports are OUT. The engine was a success, running until the late '40s, but not so good as to justify building another one.
Two problems with steam jackets. In small engines, the steam condensed in the jacket to keep it hot often exceeds what it prevents in the cylinder. Even in large triple expansion marine engines, jackets were found to be useless except for the lowest pressure cylinder, where the steam was usually wet. Jackets used as a reservoir for inlet steam make great separators, which means cylinder oil can not be fed with the steam but must be fed directly into the cylinders.
Jim Crank
Re: in search of
November 17, 2002 08:11AM
Tom,
The flow losses in the pipe are nothing compared to the losses going through a throttle valve and the inlet valve. The very best steam engines had wide open throttles and then did their "throttling" with the cutoff percentage. This could be adapted to a car.
The jacket, as you now describe it, would be exhausted right away and you would still be back on throttling and flow.
Inlet piston as a valve idea has been done a couple of times, seems to me that the early Whitehead torpedo engine had this, as did some other steam engines long ago. The Tesla generator in one form used this. There is a very bad leakage problem and the number of piston rings needed to really seal off the steam are very high. Talk to Jerry Peoples about this.
Kaowool is certainly used by lots of us.
A Roots blower as a compounding device will not give you a lower water rate. The Roots has no internal compression, hence no expansion capability.
What the Roots does have, is the ability to run on water, so it makes a dandy vacuum pump; but not a second stage expansion device.
The auxiliaries take too much power to be driven by a second stage device. The backpressure needed with a Roots to give this would ruin any engine's water rate to a high degree.
Also the time the Roots driven auxiliary is in operation does not mesh with the pumping demands. Any exhaust steam driven motor is just fine as a draft booster or fan-vacuum pump power source, then it matches those load needs perfectly.
Graphite needs more research; but when some composite does emerge, then the ability to throw away the cylinder oil will certainly advance the steam car's acceptability. Right now graphite has gross wear problems.
Jim
Peter Brow
Re: in search of
November 17, 2002 10:25AM
On poppet valves, the main friction loss I worry about is in the valve drive. Balanced poppets and proper exhaust compression can eliminate the loads from steam pressure differentials, and stem sealing can be both effective and low-friction, but with cam drive the energy put into accelerating the valves is mostly lost on valve deceleration (the valves, rockers, etc don't turn the camshaft). If the valves have to move very quickly, for very short cutoff, the energy loss is considerable. This is a big part of the engine's energy budget, even in 4-stroke engines where the valves move much more slowly than in short-cutoff steam engines.

With piston valves & classic valve gear, the valve mass is larger, but energy put into accelerating the valve is mostly returned to the crankshaft on deceleration, minus friction losses. With good valve rod seals and valve rings, the seal friction is acceptable. I think that this system can be designed to lose less energy in extra steam leakage, steam passage surface losses, & lower expansion ratios than a practical shorter-cutoff poppet valve system would lose in extra mechanical friction -- with both systems at low loads.

I did some comparative kinetic energy calculations on poppet and piston valve systems a couple years ago, but don't have the figures handy. Even with the poppet valve system running relatively long cutoff (slow valve speed), it used more energy than a piston/Stephenson valve system for the same engine. Shorter cutoff dramatically increased the friction losses.

Perhaps a balanced poppet valve with some desmodromic drive method in which valve deceleration returns kinetic energy to the valve drive system? Some clearance is needed for good seating (.006" cold clearance in aircooled VW valves), but the final dead drop could be kept minimal so that little energy is lost to seating impact.

With poppet valves, the mechanical demands for slow/gentle valve acceleration & seating must be balanced against the need to open/close the valve quickly for acceptable cutoff & flow restriction. A practical tradeoff in a steam car engine may not give better results overall than classic valves & valve gear.

On "hot sleeves" for cylinders, perhaps instead of pressurized inlet steam sleeves we could fit unpressurized sleeves filled with circulating hot oil. Some synthetic oils can maintain the right temperature at ambient pressure. The liquid could be circulated through special coils in the boiler to keep it hot, and the engine could warm up faster. The extra pump and plumbing, and their heat losses, would be downsides. With heated cylinders, I wonder about heat transfer from cylinder walls to exhaust steam during exhaust and compression.

Peter
harry schoell
Re: in search of
November 18, 2002 09:22AM
Hi David,
By your comment,it apears that oil is the enemy of the carbon. Any coments on water lubrication.
Jim Crank
Re: in search of
November 18, 2002 09:52AM
Peter,
What definition do you put on "high speed"??
If the engine is turning 750 at 60 mph like my Doble, then the factors of high cam loading from inertia and such are very insignificant. If you are thinking of 5,000 rpm, then they sure do enter into the equation.
I, for one, really like the feeling of a steam car with a slow engine speed. The mechanical silence and the brute effort feeling such an engine delivers.
If you are talking about the type of engine that people used to favor, the bash valve converted outboard, then it makes a racket like any gas engine.
One of the great charms of a steam car is the relative silence.
Piston valves are certainly a great design, they work well up to 2,000 rpm and most certainly have been used there. The Stephenson gear is large and lots of parts that take a lot of room, a shifting eccentric gives the same valve motion and is most compact and easy to design into a new engine. With a high pressure engine, one has to use a good number of rings to slow down the leakage and they do require good lubrication or the ring wear is fierce.
Poppets take no lubrication.
Also, the poppet is adaptable to not only variable timing; but also variable lift, so that choice is open to the builder. The only thing to watch for is wire drawing with a tiny lift. No one said designing a good modern steam car engine was easy!!
Jim
Caleb Ramsby
Re: in search of
November 18, 2002 01:02PM
A few years ago I found this site [www.sixstroke.com] .
It is an interisting idea. It has some possibilities as a valve system. The top piston might need to have a controlled and variable actuation but the change of displacement for each cycle is interisting and presents various options for the standard piston engine design.
harry schoell
Re: in search of
November 18, 2002 04:07PM
Hi Jim,
Sence Biesler's time, we have a lot of new materials. Even Mcmaster Carr has ss valves, not to expensive for 5000lbs/400f. There are others. The limit on 304 ss is 1200f and 3200lbs is fluid so there is no need to go higher. The pump is primed with a gear pump to eleminate cavation. Tubes are small as they contain a fluid. No need to wonder where the gas is. Small bore and stroke,so it dosn't ring off the axel. However,if and when a modern steam car is built,it should be the baddest bear on the block. The only way to get attention.
Peter Brow
Re: in search of
November 19, 2002 01:57AM
Hi Jim,

Agree on the advantages of low speed engines. Performance and "feel" far outsell fuel economy, especially in limited-production exotic vehicles.

I found problems even at 1000 rpm max, with cam-driven 1" poppet valves (1/4" lift), even at 25% cutoff valve velocities, and getting worse with shorter cutoffs. I have to dig up the figures and report them here; currently lost in my notebooks somewhere. Under such conditions, the valve system loads & likely wear rates were quite acceptable (lower than in many durable gas engines), it was just the non-regenerative nature of the valve drive that hogged too much hp in friction, relative to the steam leakage savings, for my taste. Maybe the right kind of desmodromic drive would solve this problem, if it could be made simple enough.

Whenever the valve decelerates, it should ideally transfer its kinetic energy back to the crankshaft with minimal mechanical advantage/friction loss. Typical cam drives are a "non-return" mechanical system, similar to many types of automotive steering gear. During valve deceleration, they act like dashpots or dampers -- which is good for hi-revving IC engines. They can easily afford the friction losses.

In some of my comparisons, the piston/Stephenson valve system used more hp, but most of it went back to the crankshaft during valve deceleration, unlike sprung & cam-driven poppets.

This is why I like tandem poppet valves. The valves can run a lot slower & eat less hp at a given stroke/rpm., even at shorter cutoffs. Of course, these have their own downsides, possibly fatal ones. I haven't done a detailed analysis on tandem poppet systems yet.

I have read that Stephenson geometry gives a somewhat different motion relative to single shifting eccentrics, increasing lead/compression at shorter cutoff (which is desireable, and btw hard to duplicate with cam/poppet systems). Obviously the motion is identical at longest cutoff, but at shorter cutoff ... ?

So far, I have avoided shifting eccentrics due to negative durability reports on their use in locomotives. Stephenson gear was found to be more rugged in hard daily locomotive service, esp compared to wedge-shifted eccentrics. I too hate the extra eccentrics & links in Stephenson gear (and look out for mechanical interference & off-axis loads), but the alternative valve drives I've researched & tried to design so far have ended up with plenty of their own problems (fabrication, packaging, cost, etc.).

Admittedly I haven't analyzed every possible valve system in detail -- not sure if any lone researcher could find the time for that. There are a lot of options out there. Also, I ain't perzackly the brightest bulb in the chandelier when it comes to designing complex machinery, which also limits my design options.

Currently planning on piston/Stephenson, with 6 valve rings (& 180° staggered gaps in adjacent rings for maximum labyrinth-seal effect) on each end and the most bulletproof oil pump I can design (adjustable stroke & positive-action rather than spring-returned plunger), but I am keeping an eye on other valve systems. Maybe there is some valve system which is more easily-built, more compact, cheaper, etc., and with suitable valve timing characteristics for automotive steam engines.

I love the compact, no-eccentric Joy gear, but not their constant-lead motion. Beating the heck out of the bearings at high rpms, or jerky running at low rpms, what a choice. Valving is the toughest thing -- and the most crucial -- in designing steam car engines.

BTW, I am considering trying "TotalSeal" gapless rings for the power pistons (2 on each piston). I don't like the price, but slashing piston blow-by at low rpms is extremely attractive. Alas, they are not available in suitable sizes for my piston valves.

Peter
Jim Crank
Re: in search of
November 19, 2002 10:25AM
Peter,
The shifting eccentric can easily be designed with a curved slot, which will give you total control over the advanced lead you want at high speed. The Besler airplane and the Chevy conversion for G.M. had this.
The Joy gear can also have this advancing lead with short cutoff by use of a curved guide.
There are dozens of desmodromic valve gears going way back to 1912. Some are simple in concept, and some are machining nightmares. This is certainly a good way to run poppet valves, providing you can also include cutoff control. The Skinner dual cam with variable phase angle is also a most satisfactory way to do it too.
Also, if you are in a real bind, you can use two poppet valves in parallel and thus halve the valve gear running speed. That is if you can keep the clearance volume under control. I fought this one when designing the valve system to use with the Wankel engine, not an easy one to solve.
If the NASCAR guys can go to 9800 rpm and use flat tappets, per their rules, we certainly should be able to use 2,000 rpm in a steam engine and get good cam and roller follower life.
The trick is the opening ramp and the closing ramp angles, much easier with a steam engine than a gas engine where you want it wide open instantly. The steam engine doesn't need such violent cam action in actual practice.
You are absolutely right, the valve gear is more than just critical in designing for a modern steam car engine. It is the key to a good engine.
Then, just to further confuse the issue, don't ignore a hydraulic valve drive.
The old Dykes three piece gapless ring was a real dandy, used them in my Stanley. The Doble has step gap rings for the same reason, cut down on leakage. Gap rings are horrible and you need at least four of them.
Jim
Lowell McMillan
Re: in search of
November 19, 2002 03:50PM

I was reading your ideas of a more effecient steam engine . I have a couple steam engines . One is designed to run up to 800 rpms , and the other that run's at 300 rpms' . It seam's that the slower steam engine is more effeicent . The only thing I can figuar is the slower speed's transfer's the energy better , maybe allowing the steam more time to transfer the energy . But Lllike the I studied the concepts and idea's , and my first thought was the valves could be greatly improved . But when n actual use it turned out the low rpm and low presure worked the best . Yes there is energy losses , but these energy losses are going to lubracation , so there in almost no wear at teh low speed's .
But if you do deceide to futher your investagationand developement , contact Mike Brown on the internet . He's has some developed technonogy along this line aready includding electric valves last I heard .
Once you've looked into this alittle more you will find that the piston can handle 144 pounds per squar inch , so valving timming become cridical at the higher presures . The increase heat though causes the engines to heat up more , and lubracation from the water to decrease lubraction . It's important to have condensendation in the cylinder for the lubracation if engine life is a concerning factor . These big hudge low rpm steam engine last forever , and are so effeceint . I've heard of one running 97 years' . They improved the valving but that was it in the 97 year's of service , why because the slow speeds are effeicent , and the lubraction of the water seals the piston and lubracates almost friction free . The faster you spin the engine the more energy it takes , beleive it or not . Just my though , but good luck just the same . And if you know were I could find a flyball governor about inch and half let me know thank s lowell
harry schoell
Re: in search of
November 20, 2002 02:09PM
Hi Peter,
Maybe i'm missing some thing. Why would you concider a Stanley engine as a modern car engine. The water rate of 20lbs per hp. With this in mind you could have a great toy,commercial sucess could be elusive. A 5" stroke at 150rpm is .4 sec. travel time. A lot of time to leak. Lot of make up water,lot of dirt,rust. I feel that it is necessary to have a total sealed system, water lube. A low speed has to small of a power band, to much input torque at the in put shaft,requiring a larger transmision.
David K. Nergaard
Re: in search of
November 20, 2002 03:04PM
Harry, I think you are missing something. No one claims the Stanley is efficient, single expansion with slide valves just isn't that good. However, it is as good as many "modern" attempts to build steam engines! There is no reason to run an engine fast other than to get more power out of a smaller package. As the boiler and other components dominate the space required for a steam system, there is no real reason to make a really small engine. The higher efficiency, smoother operation and easier to build valve gear of a slow turning engine are very attractive.
Peter Brow
Re: in search of
November 21, 2002 03:29AM
Jim: Thanks for the good info & ideas. Lots of food for thought there. I can see how a curved slot shifting eccentric can give variable lead, & might be "ruggedizable" if point or line contact can be avoided in the curved slot. With only 2 cutoffs, a simple arc slot between the 2 settings could give a good (curved) plane contact area. Didn't know that about Joy valve gear, thanks. I will have to study up on this type. Maybe I could build a small working model with everything adjustable in length, position, etc.. Computer can do it, but the model would be easier for me, and more fun to show off. Hydraulic valves, oh boy, don't get me started. I actually designed a system which I think could work, but both the plumbing and the maze of design options to explore are real snake nests. The KISS principle is keeping me away from those vipers for now.

Lowell: Engine efficiency depends partly on the ratio of piston work to friction losses. With low pressures and high moving part velocities, the efficiency is lower. I would guess that this was the case in your 300/800 rpm engine comparison, assuming the designs were otherwise similar. At higher pressures (MEP) though, the work/friction ratio & efficiency can be quite acceptable with higher component velocities. There are lots of other factors to consider too. Water lube is great for really big engines in some applications, but in small engines that water tends to condense and waste pressure/volume of incoming steam, upping the steam consumption per hp/hr.. Many small saturated-steam engines eat 40 lbs/hp/hr or more. Some of Mike Brown's engines give similar results with saturated steam, but I hear he has a hot new engine on the way, designed by one of our top steam guys. Superheat and properly lubricated dry steam give much lower steam consumption in a given engine. The only thing that should condense onto cylinder walls in an automotive steam engine is oil.

Harry: ditto David's comments, and I'm not talking about a Stanley engine, though it will be similar in some ways (and very different in others). See above on water lube. Water consumption, dirt, rust, etc are functions of condensing capacity and sealing of system. Agree, hermetically sealed is best. My current design gets about as close to hermetic sealing (no open vents) as I can figure out at this stage. But it does use check-valved vents for some conditions, mainly warmup and cooldown. Cooldown air-inlet vents to avoid vacuum-resistant water tank/condensing system, and warmup steam/air vents to avoid ruptures & pressure leaks in tank & condensing system. An inert-gas bladder on vents would be ideal, but there will probably be excess steam to vent under some conditions, gas loss/supply is problematic, and allowing air in on cooldown shouldn't be a problem as every surface the air touches will be oily. A few places will flood with oil on cooldown, to exclude air from those zones.

Vents are also nice in case I've totally blown it with some calculation, and end up with clouds of steam trailing the car. smiling smiley

A lot of steam leakage at valves & pistons is sometimes preferable to a little extra friction in seals, valve gear, high rpm, etc..

Peter
Sorry, only registered users may post in this forum.

Click here to login