in search of

Peter,
The Skinner dual cam gear with variable phase between them is awfully nice. Combine it with the 1912 Peugot desmo like cam follower if you must. Actually I don't think there is any problem with a 2,000 rpm engine.
Hydraulic looks good on paper, until you start adding up all the horrid complication and single points of failure that type could have.
Electric valves are silly unless you want a 200 amp generator to run them.
There is so much to recommend the double seat poppet that I really wonder why any effort is expended to try to fine something else. They seal, are light weight, take little power to operate, minimal clearance volume and don't need lubrication except for the stem, and that can be in a long graphite guide. What else does one want? Even Abner had to admit they were better than his beloved piston valves with high temperature steam.
Jim

Hi peter
Like the bladder idea. Make up water can destroy the engine and boiler,clorides, oxygen, rust.ect. The total system should be of non corosive alloys and water lube. Ihave experence in this area and know it works. Oil contamites, iron rusts.

At the time Professor Stump listed those 7 losses it was not possable to mathmaticly analyze a compression cycle.

Professor Stump was also promoting his Uniflow design. The fact is that a lot more loss can be attributed to the clearance space. The clearance space loss increasses at the higher expansion ratios of shorter cutoff.

At that time an engine efficiency of 80% was quite good. An engine with 10% clearance operatoring at short cutoff with little or no compression could easly have a 40% loss of efficiency. When you calculate efficiency loss due to clearance, it can account for a big part of the total loss.

I suspect that the Uniflow engine gains most of it's efficiency from high compression. I beleave that a high compression counterflow engine would be just as efficienct as a Uniflow engine. This proves out using a modified Rankine cycle that takes into account clearance.

With a counter flow design compression can be varied by valve timing. At low spead and long cutof were compression causes a uniflow engine to run jurky a counter flow engine would not run with compression.

The idea that less cylanders are better because a larger cylander has less surface to volume needs some investigation.

A smaller cylander that is square(boar equal to stroke) can have less surface area than a larger long stroke engine. Why isn't a square engine
better?

Take two engines of the same displacement. One having two DA cylanders and one having 6 DA cylanders. The 2 cylander engine needs more then 90 degrees of cutoff to obtain overlapping cutoff. While the 6 cylander only needs a little over 30 degrees of cutoff. That is a 10.8 to 1 expansion ratio for the 6 cylander engine compared to less then a 2 to one ratio for the 2 cylander engine. The higher expansion ratio (at overlaping citoff) of the six cylander engine can get close to 3 times the efficiency of the 2 cylander engine. Friction and heat losses will be higher for the higher cylander engine. But I suspect the losses could never be close to 60% which would nullify the 300% gain.

We would wont a larger 6 cylander engine so as to produce the same or more power at short cutoff. The 90 degree cutoff produces around 2.8 times the torque as the 30 degree cutoff.

There is a lot of misinfomation running around. A lot of missinterpitation of results. I can't find any scientific quantative testing of these so called losses. As an example: A fellow SACA member arguied that compression did not improve efficiency. He had tested it and found that with compression his engine produced less power. That was the wrong conclusion. It did produce less power. But power along is not the messure of an engines efficiency. The engine actually produced more power per pound of steam used. It ran more efficienct but at lower power. Another thing that happens is that someone says they did some one thing and the result was ... But when you get into it a bit deaper you find that it was not just one thing that was done. Several other things were done that could have had some effect on the result. The idea that the Uniflow reduced surface losses does not hold. The Uniflow engine also added high compression which could account for the reduced losses also. Here is aniother one: Running two different engines at the same power output and the larger engine using less steam proves that it has lower loss due to it's lower surface to volume ratio. That is a wrong conclusion. If the larger engine is running at the same output level. It is using shorter cutoff or lower pressure steam... To mesure the differance between the two engines one would have to run then identically. Having the same inlet steam properties and valve timmings. The larger engine would be producing more powere. One would have to compare the power per pound of steam used of the two engines. The surface are conclusions shows some merit. A six cylander engine has 1.4 times the surface as a two cylander engine. But without quantative testing one can not make any asumptions.

The basic problem is that to get high efficiency a high expansion ratio must be used. The high expansion while more efficienct produces less power. That is the general rule of steam engine efficiency. Higher efficiency lower power. To get higher efficiency for a given power output a larger displacement engine is required.

For an automotive application we are space limited. We must get the boiler and engine in the automobile. The most space efficient positive displacement engine is the Wankel. However for most of us the Wankel would be way to expansive to develop. My crankrod design is an attempt to reduce engine space for a DA piston engine. The crankrod engine combines the cross slide and crank of a DA engine. It reduces he space requirementt for a given displacement by about 30%. The boiler space can be reduced as well. There are some very space efficient mordern designs. The higher efficient engine would also reduce boiler output requirments. I plane to use several small diameter tube boilers ganged together. The heat transfer capabilities of the steam generator(boiler) increases proportional to ~1/d^2. If you used a tube of half the diameter tube you would almost quaddrupple boiler output. But at the same time flow resistance through the smaller diameter tube would increase way to much. By using several boilers the tube length of each boiler is reduced and pressure drop(due to flow resistance) across a boiler can be kept the same or reduced.

Engineering involves makeing lots of trade-off decisions. To make those decisions one must know the results of the trade-offs. Piston steam engines are an abandended science. Verry little new development has been done in the last 60 years. The Williams brothers invented a high compression engine. Their car performed with outstanding results, With higher efficiency then thought possable at the time. They were unable to prove the science behind their high compression cycle. The Williams engine is the most advanced development in piston steam engine design in the last 100 years.

Hi Jim,

The Skinner valves themselves sound practically ideal. But, they must be driven, and the whole system seems too tricky for me to design, more complex, and not as light or compact as the Stephenson/piston valve system. Fitting any poppet valve drive system to a double-acting engine which has to fit into a tight space is not easy. There are lots of ways to go. Put the cams on the crankshaft, wider engine. Separate camshaft(s), more friction lost in drive and more cost. Sliding cams or sliding phase adjusters increase the camshaft length & packaging problems, among other things. Anywhere you put poppet valves increases the external dimensions of the engine, relative to centrally located piston or D-slide valves. Poppet valves are 4 moving parts per cylinder (2 w/uniflow), vs one for a sliding valve. The valve drive system and non-moving structural parts get more & more complex too.

Then, I'm not sure whether any of this is worth the trouble for an engine that runs at low & variable loads/rpms almost all the time. It's like installing a superb car stereo system in a no-muffler dune buggy; extremely adverse conditions prevent it from ever performing to advantage, so why not use the cheap crummy system.

Still, I keep sketching this stuff up, again and again. Some poppet valve idea comes along that sounds great, so I do the flow velocities and valve sizes, whip out the graph paper, and try to fit it all together in a neat production-friendly package with at least halfway simple, buildable, durable, & affordable machinery. Lots of stuff looks really good (and sounds even better) until I actually try to _build_ it. Well, just went over it again with Skinner valves & gear, same results. I should have expected this; they're just slightly more intricate poppet valves after all, with slightly more complicated valve gear. I've gone over poppet valve designs many many times before with nothing but never-ending design problems & unacceptable tradeoffs (cost, size, weight, complexity, etc). Maybe the next time I go thru it all again, something will click.

Agree, hydraulic and electric valves are troublesome. Maybe somebody will work them out, but for now it doesn't look like it will be me.

Harry: What kind of engine water rate have you gotten with cylinder water lube? Yes, makeup water eats the powerplant, but this can take 20-30 years or more with the right sytem, IMO. By then the car, its same-model-year IC competitors, and maybe its owner too, have been scrapped. Oil contamination can be minimized and easily removed. Oil protects; stainless costs.

Peter

Hi Andy,

Remember that the 6-cylinder 10.8:1 expansion ratio engine in your example, under starting conditions, will have a considerable cylinder temperature drop between inlet and exhaust events. With no compression (starting), the tops of cylinders -- heads, walls, and piston crowns -- will be much colder than the incoming steam. With short admission, there will also be very little incoming steam, which will spend a lot of time (low RPM starting) in a cold-walled space with very high surface to volume ratio (top of cylinder just after TDC/admission). So I think that an awful lot of the inlet steam temp/pressure will be lost to Prof. Stumpf's "surface losses". My guess is that experiments would show better results starting with longer cutoff in the 6-cylinder engine, and perhaps better steam rates at superlow loads/rpms (starting conditions) in the 2-cylinder competing engine with much longer cutoffs.

I think that optimum bore/stroke ratio is a result of MEP, superheat, rpm, cutoff, compression, conrod/stroke ratio, and probably other factors. Standard ratios were developed in the 19th Century thru extensive practical field experience. For modern engines with very different operating conditions, perhaps numerous engines, otherwise identical but with different bore/stroke ratios, should be tested to find new optimum ratios.

Peter

Andy, the important thing about the uniflow is NOT the compression, it is ISOLATING the hot inlet valve and the cool exhaust port. Increasing compression in a counterflow engine merely causes more condensation in and near the exhaust port, a fact demonstrated by many old indicator diagrams which showed compression to a fairly low pressure, then no increase in pressure to the end of the stroke. The compressed steam was condensing in the exhaust port!

Peter,
Water rate on the SC engine is expeced to be in the 5to6 range. The Mark111,Box engine , should be in the 6to7 range as it is a simple engine. They both have a reheat stage. Water is pumped into the bottom end for lubrication. A hydorlic seal is behind the rings in the lybrith groves. There is an ample use of aluminum also,its cheep. Also found bellows 450f 15psi. Gortiflex www.gortite.com thank you.

Hi Harry,

Water lube in the crankcase, okay. That might be workable with the right bearing materials & Jim's pre-oiler (pre-waterer?). In the cylinders, uh oh. A wee film of water in cyl wall micro-irregularities will condense or cool an awful lot of steam. There's no arguing with 5-7 lbs/hp/hr ... if you can get it! Sounds doable at full load, but I'm not sure about part load conditions, and that is what cars run at most of the time. Some features that increase efficiency at full load can actually decrease it (relative to a leaky "classic" design) at part load, esp. at very low part loads (eg, stop & go city driving).

Bellows are great; reminds me of the "roll-sock" piston road seals in the Philips/Stirling engines, hermetically sealing hot, high-pressure hydrogen (the ultimate sealing nightmare, h2 migrates thru solid metal!). 15 psi sounds a bit low for practical steam engines though, and bellows have their own pumping effect & surface/volume ratio ("surface loss" problems. Aluminum alloys have good strength/weight ratio and machine nice, but look iffy to me, at best, at steam temperatures. Watch out for thermal expansion, differential expansion & maintaining clearances, esp in a system with extreme hot/cold cycles (automobile warmup/cooldown). Okay in cold zones of engine, though. Belleville washers are sweet for taking up differential expansion in some places, but tricky to engineer in when balancing mechanical & expansion loads.

Best of luck; hope you can get those water rates (or even twice that) under road conditions, that would be a big step ahead.

David & Andy: "Surface losses" in engine steam passages are indeed a major factor. Not only do steam passages increase internal surface/volume ratio, but heat transfer between internal surfaces & steam is both radiant and convective -- high steam velocities & turbulence = high "wind chill factor". Eliminating passages with overhead poppets & separate inlet/exhaust valves (and/or unaflow) is an option, though side valve systems can be much cheaper, simpler, more compact, lighter and less draggy, so there goes the multimultifactor design tradeoff chess game again.

Even with unaflow, steam flows radially across piston crowns at exhaust, and exhaust collectors and interport ribs make nice cylinder-cooling heat exchangers, more radiant & convective heat exchange to watch for, plus axial heat conduction down the cylinder walls (which with high expansion ratios are nicely tapered, to maximize our fun with machining, thermal cycling/leakage/ring springing/ringgroove wear/ring fatigue) from the hot heads to the cool exhaust port zone. Most successful unaflows were big ol' factory engines with mostly constant/full-load running, constant cylinder thermal gradient, much lower surface/volume ratios, & warmup/cooldown comprising much less of the operating time than in a steam car engine.

Refractory internal surfaces are something I am looking into, probably good for any type of valve. But what is the surface-loss effect of oil films on internal surfaces? Oil has a pretty good specific heat, but I don't know its coefficient of absorption.

One trick I sketched up for a unaflow design was refractory gaskets between outer walls of cylinders and bolt-on exhaust collectors, to cut heat conduction from cylinder to exhaust collector. Construction problems, and probably tried many times before, for all I know.

Andy, the extra size/weight/cost of a higher-expansion steam engine are indeed very important factors. All kinds of tradeoffs to consider there. I think your counterflow total-control valve regime can give optimum performance/efficiency results, under all conditions, for any given engine.

Peter

Uniflow engines were superb in variable load situations, their steam consumption was "flat" from 1/4 to 1 1/4 rated load. Witness use in the RV Atlantis II and numerous ferry boats.
The only surface in the engine exposed to high velocity steam flow during both admission and exhaust is the top of the piston. Perhaps an insulating coating would be useful here.
Heat flow to the exhaust ports and belt is not an issue. During admission that part of the cylinder is covered by the full length of the piston.

Hi David,

On reflection, it seems that the heat conducted down the cylinder wall will just heat the expanding steam which contacts the wall as the piston travels down. Heat should go from cyl wall into the steam probably as fast as it goes down the cylinder walls. So you are right, no problem there. I was repeating something I read long ago, without thinking it through. Ted Pritchard got good fuel mileage results under road conditions with a small uniflow, can't argue with success, though I wonder about the 80%(?) cutoff he used to overcome compression for starting/low speed, and how the steam rate at very low loads/speeds compares with classic steam engines with shorter cutoff under such conditions.

In highway-capable cars, engines often run at 10% or less of full load, during stop & go city driving.

Insulating piston crowns seems like a good idea for a uniflow -- and probably for any kind of steam car engine. Same for inside of cylinder head, as this will have both radiant & convective heat transfer from inlet to exhaust temps due to turbulence in cylinder.

It would be nice to have a small fleet of identical cars with various engines -- gas, diesel, uniflow steam, counterflow steam, etc -- and see how they perform on the same course, driving nose-to-tail with the same acceleration, deceleration, cruising, etc..

Peter

The easy way to make engine comparisons on fuel consumption:
Diesel hpx.055=gallons fuel
Gas hpx.o8=gallons fuel hour
2strokegas hpx.115=gallons fuel hour
What does your steam engine use in fuel?

Hi Harry,

Easy, but not quite accurate. Those are full-load fuel consumption figures. At part load, esp 10% or less of rated hp, gas and diesel engines use a lot more fuel per hp/hr -- often several times as much as at full load. Their efficiency drops dramatically when running at part load. Then you have to average-in fuel losses from idling (zero mpg), mixture enrichment during acceleration (gas engines), fuel waste during coasting and braking, transmission losses, etc.. Many of these losses are far less, or even nonexistent, in a steam car.

Most of the time on the road, any car engine is running at a tiny fraction of its rated horsepower. For most production gas cars, it is actually impossible to run the engine at full rated power or maximum efficiency on the road, even at top speed, due to the gear ratios used and other factors.

Chassis dynamometer tests for gas cars with large engines and automatic transmissions (which is most gas cars) show 7-10% net efficiencies (fuel tank to hp at wheels) under simulated road conditions. Even a good antique steam car can get that efficiency. With a much smaller engine, manual tranny, etc a gas car can run almost twice as efficiently as this under the best conditions (eg, level highway cruise), but relatively few people buy such cars due to reduced performance & other disadvantages.

The above is why I think that running identical vehicles with different motors under a number of varying road conditions would be necessary for an accurate efficiency comparison of different vehicle drive systems.

Peter

Hi Peter,
Think you will find them approximaly accurate. Dino numbers show this.
A gas car at 60mph 30mpg =2gal hr
2_
.08=25hp
If you know the dino numbers rolling and wind resistance can be calcucaled. A 300hp gas engine will use at full throtle 300x.08=24gal hr. Of course race engines will use more as some is used for expansion not for combustion. And that is why internal combustion is dirty, no time to mix. This is highway miles and do not relate to stop and go. This is where weight comes into play. hp/lbs ratio.

Hi Harry,

I have located a few (hard to find) chassis dyno charts, which show BSFC (Brake Specific Fuel Consumption) figures under various simulated road conditions. When the BSFC numbers are converted to net thermal efficiency, it is apparent that in part-load road conditions, modern gas car engines are far less efficient than the 25-30% often quoted.

25 hp is a typical cruise output for modern cars. Since most modern car engines are rated at 75, 100, 150 hp or more, they are running at very low load/efficiency even at highway cruise!

Only a small percentage of modern driving occurs under highway-cruising conditions. ~70% of world population lives in increasingly-congested urban areas, and in such areas even highway traffic is increasingly "slow-and-go" or stop-and-go.

Peter

I don't know whether modern Otto engines are better in this regard, but from mid '20s right through late '60s the following rule applied: Otto cycle engine at 1/4 throttle used twice as much fuel per horsepower as at full throttle. Smaller throttle openings yeild lower efficiencies, the limit being efficiency equals zero at idle. Actually that is wrong, if you close the throttle at speed, the efficiency is negative: engine braking going down a hill for example.

Peter, Dave,
These approxemate numbers are close.We use them for fuel consumption on boats. Top speed, cruisespeed,idlespeed and they hold for fuel consumption/hp. Hourse power is hp is hp is hp. Chassey design is a nouther issue. Boiler.engine.condencer need to be concidered a single unit. We fill up at the pump. It will still come down to fuel to hp.

Hi Harry,

Okay for boats, perhaps, but you should see the charts for cars. The lbs/hp/hr varies quite a bit, in a steep curve, mostly at the low end where most driving occurs (which is deceptive at first glance). One of many factors is that transmission losses increase at lower speeds due to higher thrust loads/friction with higher gear ratios. Also, 90% of new gas cars have automatic trannies.

Hi David,

The BSFC charts for 1970s cars showed the same results as those from the 1930s. I have a published source for one, mid-1970s, need to look it up at library and maybe pull an unauthorized web republication (Fair Use Doctrine; educational purposes only). Kent has one chassis-BSFC chart, though it is hard to read and the vehicle is not identified.

I have only been able to look at 1990s charts ("eyes only" in meeting with mid-level industry source, funny how they guard this info, maybe a habit from DoD contracts), but the figures were unchanged from Kent & the published '70s charts. To bridge the document gap, I have noted that the 1977 Ford Fairmont got exactly the same EPA MPG figures as the 1999 Chevy Cavalier; these cars are very close in weight, engine displacement, transmission type, and frontal area. The Chevy even has a lower cd, so you'd think it would do better, at least in the highway cycle. The EPA test cycles changed a little between '77 and '99, but not much; I've been told less than .5 mpg in a typical car.

Another source told me that there have been internal engine efficiency improvements since the '70s, EG, fuel injection, computer control, and also vehicle improvements (air drag, tires, trannies), but they were relatively small, and eaten up by smog & safety modifications. One step forward, one step back. As Ted Pritchard says, pity the poor IC engineers.

I think that basic Otto fuel & cycle parameters have changed very little since the 1920s; they developed a formula that worked and stuck with it. This also means that they hit a brick wall with engine efficiency improvements. Hence the interest in hybrids, fuel cells, etc..

Maybe we should shift to a new thread; this one is getting extremely long.

Peter

Harry,
Great thread you have started, I wonder if you could clarify your 12-02-02 message as to what the .08 is?
Your example is typical of my old 1984 Dodge Caravan that would get 30MPG@60+mph=2 gallons per hour. However its partial throttle efficiency at this speed was very high compared to steam automobiles. That is 25HP X 2544/270,000 equals 23.4% thermal efficiency including driving auxilaries at about 1/4 throttle. This is going to be hard to match with steam engines using 10-20# steam per horsepower hour. Theoretically your 6-7#/hp engine with 80% boiler efficiency would not equal that(but close) if auxilary loads were taken into consideration. It would be a very interesting development to have a modern engine built and tested with the 6-7# steam rate, tough to maintain on the road.
Best of holidays, George

Hi George,
Transmission losses are not as great as sometimes persived. The varible belt is only about 3%. The largest gaines are with fuel injection,only because it doesnt have an excelerator pump. The basic engine no real gaines. The gas boys are running in panic cercels. the end of their phyics. We have opertunites here. The first is a clean burn. However efficency theory is easy to talk about.mostly hard work and lots of money. I am gratefull for the ideas and opinins. It may save Peter, myself and others who are inventing new systems some time and money.
Hi Peter,
Like your boiler,think you will have better than the 80% that George was refering to. You might concider a multi-gang pump, one on each tube to avoid burnout.

Harry,
Using one pump plunger for each circuit in a monotube's economizer section was something that YUBA tractor did with their steam system.
Sure eliminated the problem of equal flow in all sections of a multicircuit system. That is until one valve in the pump decides to quit.
JC

Hi Jim,

Yep, and I sure had fun designing a bypass control for 8 separate feedwater circuits. Ended up with an 8-port pivoting plate valve with hydraulic servo. Little differences in packing leaks, etc, all sorts of things can screw this up.

Peter

Peter,
Wasted effort. A nice manifold with no sharp bends will work just as well as all this complex nonsense.
Three parallel circuits with finned tubing on the economizer coils will give you everything you can want.
JC

Hi Jim,

Agreed. I'm off to greener pastures in the boiler department now anyway.

Peter

The talk about insulating the head and steam head preheaters like on a piston steamship engine.Had me wondering if anyone ever run a cast iron sleeved ceramic cylinder and head.I have had tool inserts glow at tempertures that metal can not handle. And still be cooler to the touch, at the end of a cut, than metal.
Assuming the price isn't too far out there.(diamond hones&cutters for the ceramics items) With the insulation mentioned earlier by someone else,you would have an engine exteriior on par with a water heater shell.With more heat coming from the plumbing than the engine exterior.

Hi Arnold,
All ceramics are not insulators. The high alumina ceramics is nonporous,needs diamond to cut. Others are porous,altho expansion is nill. I had high hopes for more of its use and had to look else where. Also very expensive. Using some for some for some bushings,small. The castables didnt work well for me. This has been my experince but dont let me discourage you. anouther try might be the right one.
have a good day

Hi Harry

Good to see you at Danville.

How much experance have you had with the peek baring meterial. other then your engine development. It sounds vary good but in operation problems with contamination might be the killer. As you demenstrated. It only takes a small bit of contaniminate to upset their operation.

Andy

A correction that was made in one or two other threads, which I want to repeat here:

Properly-designed cam-driven poppet valve systems do NOT use a larger percentage of an engine's developed horsepower than "classic" valve gear (eg, piston or D-slide valves plus Stephenson, Joy, or similar drive linkage). My statements to the contrary, in this and other discussion threads, were based on incorrect/insufficient information. New information provided in this forum (thanks guys) indicates that the % of developed horsepower needed to drive these types of valve systems is approximately equal in engines of the same hp/MEP.

Ceramics come in a very wide variety of types. The ones I referred to above are the structural/technical ceramics like titanium nitride, which are nonporous and have high strength and surface hardness. They are suitable for bearings, turbine rotors, pistons, cylinders, valves, etc.. They are also enormously expensive; one major maker quoted me a price of ~$6000 for 2 cylinder sleeves of simplified geometry & minimum acceptable tolerances, with very little price reduction in production quantities.

Peter

Hi Andy
The system that was shown was a total seal system. Water can be filtered through a finer micron filter than can oil. Other contamints ie. rust is not an issue.
Hello Peter,
As you well know as an inventor the cost is high, and 90% can end up in the dumpster or Tom Kimmel's museum. Showed cerimic pistons at Danvill, there is a welth of information from the members,more should attend,it was a lot of fun.

have a good day

Hi Harry

I did some 3D steam charts into the super critical region. 100 PSIA to 4000 PSIA. I think wether it acts as a gas or liquid would be dependent on it's temperature. It can expand. But only to the limit of it's heat content. Below the critical point there is a fault line on the P,T,v 3D surface that is the saturation line. Above or below the critical point fluid is compressable.

Andy

Forgot the link.

[www.steamautomobile.com]

Andy