Welcome! Log In Create A New Profile Recent Messages


"Jakuba's Compounding article in the Bulletin"

Posted by jakuba 
Re: "Jakuba's Compounding article in the Bulletin"
August 30, 2014 03:53PM
Ok, I I’m confused- probably another semantics issue or problem with assumptions.

I assume we are talking about simple and compound engines expanding a given mass of steam from the same inlet conditions to the same outlet conditions.

The size of the cylinder of the simple engine or LP cylinder in the case of the compound engine should be the same in order to get the same density of steam at the end of expansion.

I suppose you all are getting tired of me bring up one hundred twenty year old steam locomotive designs, but this is one of the few cases I know of where direct comparisons were made between machines that were alike except for being simple or compound.

The main advantage of compound locomotives was in the saturated steam era due a reduction in cylinder condensation due to the reduction of temperature difference between inlet and outlet of the cylinder. Condensed water re evaporated upon release of the high pressure cylinder had a chance to do work in the low pressure cylinder. This advantage went away with superheating. Superheating also produced less dense steam at a given pressure so the boiler could fill the same size cylinder more often per minute.

The next advantage of compound locomotives was the increase in admission time (longer cut off) for each cylinder. This eased valve timing issues and provided more constant loads on all components. Steadier tractive effort caused less slipping of the wheels on the rails.

The next advantage, at least in theory, was a reduction maintenance costs if loads could be distributed over a larger number of axles, crank pins, rods, cylinders, pistons etc. This usually did not work out, due to the difficulty of dividing the work equally among the cylinders at all loads and speeds. Equal division of work is not difficult for marine engines and stationary engines that operate at constant load and speed.

Compound locomotives often ended up with increased maintenance charges due to low pressure components sized to carry half load ending up with the lion’s share of the load due to poor distribution of work.

Most French and some German four cylinder designs got pretty good division of work by providing separate valve gear for HP and LP cylinders but this caused required higher first cost and more components to maintain. Four cylinder engines that required cylinders between the frames and cranked axles had increased torsional fatigue axle failures.

Compound locomotives generally did have smaller boiler repair charges due to the steadier exhaust forcing a steadier draft. Boiler repairs were the largest maintenance charge for locomotives. The improvement was generally held to be not worth the effort. This is not a factor if boiler draft is not induced by ejecting exhaust steam up the chimney.

In the end, almost everyone gave up and went back to simple engines. As far as I am aware, only the N&W and C&O stuck with compounding to the end with their Mallets. Mallets were best case applications for compound locomotives and the N&W used them in best case service.

All of this must be taken with the context of boiler pressures of the day. Locomotive boilers with water tube fireboxes were often tried and found to be too fragile. Boiler maintenance charges increased geometrically with increase in boiler pressure (temperature). By the end of the saturated era 185-200 psi was regarded as maximum practicable. By the time superheating was sorted out all railroads used as much superheat as the lubricant would stand. Pressures went as high as 300 psi (with very few exceptions higher) by the end of the steam era with attendant increases in boiler maintenance charges. Railroads with less than desirable water (most of them) kept the pressure as low as practicable.

All of this changes if anyone develops a totally condensing steam car or locomotive. With modern boiler quality water, high pressure boilers are possible. (Locomotives will still have shock loads from coupling up and with trains stretching out and coming together, so water tube fireboxes are still an issue.) If condensers and high pressure boilers are a reality, then compounding is back on the table but the problems our great grandfathers struggled regarding equal division of work between the stages have yet to be solved.

Re: "Jakuba's Compounding article in the Bulletin"
August 30, 2014 06:30PM
Hi Kerry,

You are right that the size of the simple or LP cylinder should be similar in order to get the same density of steam at the end of expansion,
The size of the cylinder of the simple engine or LP cylinder in the case of the compound engine should be the same in order to get the same density of steam at the end of expansion. BUT this assumes you only have one simple piston but two pistons in the compound engine (or more if going to multi stage compounding). If we assume we want just as many power pulses per revolution and are willing to have just as many cylinders so as to have a similar parts count, then the 2 cylinder simple engine will have cylinders displacing half of the compound LP cylinder. There were vanishingly few single cylinder steam cars.

Where locomotive comparisons break down is when you start looking at temperatures and pressures needed to be competitive in the current market. Back in the 70s and 80s Dutcher did the most intensive automotive compound engine development that I know of. They achieved great efficiency but never really got all that close to their target because of HP cylinder piston ring blow-by; even after installing 4 rings they were still losing something like 15% of the HP cylinder steam. This wasn't just due to the high pressure, but the duration of time the rings were exposed to high pressure. The HP cylinder obviously is relatively pressurized at all times, which permits a great deal of flow. This is much less an issue with high expansion simple expansion, the pressure is only high for a small period of time and the pressure drops off about the time any blow-by is starting to gain hold. Jay Carter was quite clear that he had done a lot of testing on this point and it jives with other reports. An efficient, low friction, high pressure piston seal really needs to be demonstrated before serious compounding at automotive pressures can be expected. If four rings aren't doing the job, then we have a real problem; especially given the friction they produce.

Another issue that worries me is carry over from the HP to LP (or MP) stage, not water but lubricating oil carryover. If the flow was straight from cylinder to cylinder this would be negligible, but if the flow is passing through a resuperheater we may have some serious issues with thermal decomposition and accumulation in the resuperheater.

And, of course, I have a whole slew of other issues. It isn't the theory of compounded reciprocating engines that worries me so much as the practical difficulties in implementing them.

Re: "Jakuba's Compounding article in the Bulletin"
August 31, 2014 05:39AM
I’m having a problem understanding where you’re coming from with some of your statements.

I assume we are talking about simple and compound engines expanding a given mass of steam from the same inlet conditions to the same outlet conditions.
The size of the cylinder of the simple engine or LP cylinder in the case of the compound engine should be the same in order to get the same density of steam at the end of expansion.

A given unit of steam volume at 600 psi and 200D superheat will expand about twenty-five times in volume down to 15psi saturated as it gives up it’s heat.

To get the same work from a piston as the high-pressure cylinder in the lower pressure or compounding cylinder it has to be larger in diameter and the diameter chosen to correspond with the incoming lower pressure of the exhaust from the HP cylinder.
Re: "Jakuba's Compounding article in the Bulletin"
August 31, 2014 10:00AM
Thank you, Rolly.

I was trying to get a better understanding of the exchange between Keith and Ken particularly the following statement:

“All I meant to point out is that for a given output engine, if it isn't compounded, the single cylinder must be larger to do the same work.”

I was considering the specific output of a hypothetical engine. Is it possible to produce an engine of a given size, is it will do the required work?

I was ignoring the size of the high pressure cylinder of the compound engine. I was also ignoring the difficulty of achieving high expansion ratio in a simple engine.

I was contemplating whether the low pressure cylinder of a compound engine must be the same size as the cylinder of a simple engine in order to discharge the same quantity of steam at the same conditions.

I was mis-reading Keith’s post. I was focusing on cylinder size, and I believe he was talking about the size of the engine.

I was also expressing doubt about equitable sharing of load between the stages of a compound engine except at constant load and speed.

Re: "Jakuba's Compounding article in the Bulletin"
August 31, 2014 12:15PM
The problem is that with high expansion there is little or no room for throttling down the engine. Rolly example:

"A given unit of steam volume at 600 psi and 200D superheat will expand about twenty-five times in volume down to 15psi saturated as it gives up it’s heat. "

illistrates the problem. Compounding or simple what happems when you throttle down the inlet steam to say 100 PSI? When does the pressure reach 15 PSI? In a tripple expansion the second and third stage would not have any pressure.

Acording to the most accurate steam properties Rolly's example would be well into a wet steam mixture. It would hit saturation at 131.75F having expanded 3.23 times in volume.

Re: "Jakuba's Compounding article in the Bulletin"
August 31, 2014 01:02PM

"I was also expressing doubt about equitable sharing of load between the stages of a compound engine except at constant load and speed."

You are right, the two cylinders will not produce the same power (torque) output. Thus the very old concern about doing compounding right and why the two valve gears have to provide different cutoffs and yet be interlocked in a vehicle. Boats and stationary use, fine; but in a car, probably not use compounding.

Andy has a point. When the inlet pressure gets too low in a compound or a triple, reheat notwithstanding, the LP cylinder just goes along for the ride. One reason to provide as high a vacuum as you can to stretch out that PV diagram. Just as long as things are hot enough that the steam cannot actually condense into water and knock out a cylinder head or bend a connecting rod.
This exact condition I have seen on Besler's engine dyno when we were testing the second airplane engine under every possible condition. You cannot argue when the PV diagrams for both cylinders are right there on the scope staring you in the face.

Perhaps a good reason not to use compounding in a vehicle and add two more cylinders for reliable starting and smooth running at light load, minimum throttle and slow speed; but to provide short cutoff and then recuperate the remaining heat in a condenser fan or draft booster turbine.
Also why so much time and effort was spent learning about all the varieties of turbines that sidestep all these frustrating problems. If only there was a good way to side step or mask that nasty 1/2-V business.

Re: "Jakuba's Compounding article in the Bulletin"
August 31, 2014 02:47PM
Exactly Andy that’s the essence of my previous post.

I am still under the opinion that steam over electric would be the best approach as I believe the steam plant would need to run at a constant load.


Re: "Jakuba's Compounding article in the Bulletin"
August 31, 2014 04:14PM
Rolly, Kerry, Andy,

I think we are loosing sight of just how a steam car is used in the real world.
Stacking two energy conversion systems in series is costly, takes up a lot of space, is heavy and is just not as efficient as one of them alone. Either you want a steam car or an electric car; but not both jammed into one vehicle. Then think of reliability over time. The steamer is enough trouble all by itself!! I refer to the now frequent electric car fires you read about. If one wants that kind of a car, go buy a Chevy VOLT.

Consider this: When the engine is designed correctly, the highest achievable and usable expansion is wanted when the car is on an open road and cruising along at say 70 mph, where it spends most of its time. So short cutoff is what you want then.
If you are over expanding or making the LP cylinder actually worthless, then you designed the engine wrong and it is back to the drawing board and the text books.
When you are hill climbing or accelerating hard, this is really a momentary condition, so long cutoff for maximum torque is wanted and the larger water rate then is really of no consequence and just doesn't matter.

Quite frankly, I think these esoteric discussions would be much more useful and productive if people had experience actually driving steam cars and not just fantasizing about them. Something that our National Meet should look into. So assist someone financially if needed to bring his car to the Meet.

Re: "Jakuba's Compounding article in the Bulletin"
August 31, 2014 06:21PM

Hybrid cars have proven to more efficient. And I have not heard of any of them catching fire. The Tesla electric is the one catching fire. Personally I do not favor either one. But the Hybrids have better efficiency running a heat engine at optimum peak efficiency and electric motors that have flat efficiency from near 0 RPM on up.

A steam hybrid could have the advantage of using many different fuels.

As I explained is detail a compound design that would run a full expansion cycle of 27 to 1 or greater expansion ratio from 20 MPH to 100 MPH or more is theatricality possible. It is the mechanical design that's the problem. A steam engine of that design it just like old time engines producing gobs of torque. So the mechanical mechanisms would have to handle extreme torque loads. That's truue in any case. But changing the stroke on the fly under load presents a challenge.


Re: "Jakuba's Compounding article in the Bulletin"
August 31, 2014 08:58PM
steamerandy Wrote:
> Jim.
> Hybrid cars have proven to more efficient. And I
> have not heard of any of them catching fire. The
> Tesla electric is the one catching fire.
> Personally I do not favor either one. But the
> Hybrids have better efficiency running a heat
> engine at optimum peak efficiency and electric
> motors that have flat efficiency from near 0 RPM
> on up.

Let's be clear here about the Tesla! I will NOT ALLOW statements about EV's to circulate about catching on fire.

The incidents which were reported about fires involved the vehicle driving over massive pieces of junk--and which tore into the floor pan which contained the battery pack. Tesla has (as a result) massively reinforced the bottom of the battery pack and can now state that it approaches limits of damage control which are nearly impossible to imagine. It's also worth noting that the fire dept. respondents didn't have a clue about how to handle the situation with the kind of energy which was stored, they made things worse and so there's a need to train the fire specialists to handle situations like this in the future.

The battery pack is a marvel of excessive complexity and control and uses a temp system which can supply either AC or heat as required to keep the "pack" within a specified range of temps. Additionally, there are almost seven thousand individual lithium cells in the pack, each of which is monitored individually for its operating condition.

I think the "street" should marvel at how well Tesla has accomplished the seemingly impossible goal of producing a high-performance vehicle with pure battery power, and with remarkable reliability at the level of sophistication they are required to achieve.

Yup!!-I'm a steam enthusiast, and an EV enthusiast too.

Bill Hinote

Edited 1 time(s). Last edit at 08/31/2014 09:07PM by Bill Hinote.
Re: "Jakuba's Compounding article in the Bulletin"
September 01, 2014 02:38AM
Bill There were three separate incidents reported of a Tesla catching on fire on MSN news. I must assume they were separate as they happened in different states. That doesn't really bother me. The real problem is the mining of lithium. It's causing a lot of health problems in China where it mined. The environmental impact of mining lithium is an issue

Re: "Jakuba's Compounding article in the Bulletin"
September 01, 2014 12:08PM

Tell that to the guy who's brand new Fisker car burned his other car and garage down. It's the battery, not the fact that it is an electric. Electrics can be quite satisfactory city cars, had one.
Try five Teslas on fire and four others too so far. Why UPS and Priority Mail now will not allow any lithium battery powered anything on their air freight flights unless fully discharged or removed.
The Fire Depts here don't squirt water on lithium because it explodes like any alkali metal and because the Chiefs don't like the possibility of their men getting electrocuted. As one Battalion Chief i work with said: "If the people are out of the car, let it burn I am not putting my men at risk."

I know the Tesla, two friends have them, it's made right here. Around here Musk is considered to be a first class snake oil salesman. Similar to some Florida outfits.
So what makes then such worshipful idols anyhow, their B.S. publicity? Teslas are no better than other electrics like the GM, Ford, Nissan and especially the Mercedes B-Cass at one third the price, etc. Teslas are noted for risky engineering that has to be revised.

Sounds like you are comparing developed hybrids to undeveloped steam cars. Why? Too many other negatives to even consider stacking them together. Go buy a Volt and change the engine if you must.
Correct, mining lithium is a serious health risk.


Edited 1 time(s). Last edit at 09/02/2014 02:14PM by Jim Crank.
Re: "Jakuba's Compounding article in the Bulletin"
September 10, 2014 07:15PM
I argued against compounding in any vehicular application in normal duty and pointed out that even in steam locomotives compounding became rare by the time railroading has progressed to the point when accurate comparisons of performance and service duty became possible, and a degree of superheat common. In the U.S. no serious manufacturer designed compounded locos after 1930. After WW2 the compounding disappeared from the drafting boards of all manufacturers worldwide, except in France. It is interesting to observe why that was so.

Unlike elsewhere, a loco engineer in France had routinely been assigned to “his” locomotive “for life,” and vice versa. This familiarity of man and machine was apparently a prerequisite for operating compounded locomotives successfully for they were known everywhere as fussy in operation. We mentioned some of the fussiness in this Phorum, items like starting, distributing power evenly among stages, complexity and cost. To acquire the skill to deal with their tricky nature necessitated that one-to-one commitment. It included also the engineer’s participation in the maintenance, repairs and rebuilding of his loco. In this arrangement, the engineer learned how to tweak the various valves and levers to be able to realize the potential of his compounding loco on his route.

Nevertheless, even in France, compounded locomotives were designated to routes requiring long runs at a relatively high and steady load. Eventually, after WW2, the new designs reverted to single expansion even in France. The glorious, steam-locomotive era ended with the ubiquitous single-expansion, one cylinder on each side, marvels of engineering. We will never know as much about steam engines as the engineers of that by-gone era did.

I do not dispute the aesthetics of the compounded models on one’s office desk as some compounding-lovers claim. But I’d classify that as a stationary application, and thus within the scope of the article’s reasoning. :-)
Re: "Jakuba's Compounding article in the Bulletin"
September 10, 2014 09:21PM
My former employer had some dealings with Tesla about making parts. Talk about champagne tastes and beer budgets!! They wanted engineering at GM levels with Studebaker resources. Knew almost nothing about building car bodies, and wanted aluminum to behave like DQAK sheet steel for panel contours. Fortunately for my old shop they could not do what Tesla wanted, and backed away before they bankrupted themselves. If I were to go electric, it would be a Volt.

A coworker bought a Prius hybrid a while back. A real tree-hugging liberal, although he looked like a stevedore. After a year he wanted his old little pickup back. Total waste of money, he called it.

Re: "Jakuba's Compounding article in the Bulletin"
September 10, 2014 10:03PM
TH Wrote:

> A coworker bought a Prius hybrid a while back. A
> real tree-hugging liberal, although he looked like
> a stevedore. After a year he wanted his old
> little pickup back. Total waste of money, he
> called it.

Toyota must be doing something right, the Prius is the best-selling car in Californiia when I last checked.

Personally, I don't agree with all of their engineering decisions here but certainly can't ignore the sales success. FWIW I drive an older Tacoma with almost 200K miles and just starting to get "ripe", no oil consumption and no failure of everyday pieces such as locks, etc. A simple but well-engineered vehicle.

Re: "Jakuba's Compounding article in the Bulletin"
September 11, 2014 06:10PM
If I had to pay California gas prices I might consider it, but his was an '08. Guess they got better.

Drove by a station selling unleaded at 2.89 yesterday. Man, I love Texas.

Re: "Jakuba's Compounding article in the Bulletin"
September 13, 2014 07:19PM

I'm not burying the compound for automotive use. That's all there is to it. And mine has reheat.

We can agree to disagree, I guess.


Bill G.
Re: "Jakuba's Compounding article in the Bulletin"
September 14, 2014 03:50AM
My fantasy engine is a supercritical compound unaflow (HP->2xLP) with reheat. So Bill G, currently, you are at three piston stages, with reheat after the first? Or, are you talking about reheat before a turbine?

Re: "Jakuba's Compounding article in the Bulletin"
September 14, 2014 11:31AM
Another advocate of a compound expander with reheat between stages is John Cozby. It looks complicated, but Cyclone style packaging with high and low pressure cylinders in a radial pattern might be an approach. The reheat loops could easily be routed back to the combustion chamber. What would the difference be between this compound engine and the uniflow? Does the Cyclone overcome the theoretical efficiency advantage of compounding with its variable clearance and cut off for the highly variable loads in a car engine?

Lohring Miller
Re: "Jakuba's Compounding article in the Bulletin"
September 14, 2014 02:04PM

Mine would use reheat after the first very high pressure stage. The "trick", as I see it, is to design the first stage to supply all of the steam to the lower two compound stages at a cruising speed for the car on the freeway. This is where the highest engine efficiency is appreciated. At full power it is only supplying about 10% of the steam that the lower two compound stages are using. This makes it a small and compact stage which is more just "tacked" onto the engine.

My engine was designed to be rather large so a smaller engine might need say 20% instead of 10%. A car takes something like 20 -30 hp steady speed on the freeway.


I don't have any good data on the Cyclone, but I believe it needs reheat and compounding. It does use some re-generation but that only works well at full power. Uniflow vs counterflow for a compound depends on the setup and the relative pressures. At low pressures use uniflow as the cylinder volume is smaller for the outcome because a counterflow has to expand more.

Mine is uniflow between the two stages with an inlet pressure of 1,000 psia @1,200 deg F. Control between the mid and low pressure stages in not an issue. At high pressures around critical and a low expansion ratio of around 2.5 -3 use a counter flow with no clearance as it ends up the smaller and easier to control. I was surprised when I got that result.


Bill G.
Re: "Jakuba's Compounding article in the Bulletin"
September 15, 2014 12:59AM
Here is what I've gathered about the Cyclone Mk V:
Bore: 2"
Stroke: 2"
Cylinders: 6
Pressure: 3200 psi
Temp: 1200 F
3600 RPM max
Boiler, tube length deduced from drawing: About 160m 1/4" tube; about 3.2 m2, 304 ss.
Efficiency: "...in the 30% range, plus".

To me this is very impressive, as it is only a bit lower than a lossless (friction, heat transfer, flow) engine with 2% cutoff and 2% clearance.
Re: "Jakuba's Compounding article in the Bulletin"
September 15, 2014 01:21AM
In theory:

At a constant expansion rato power is proportional to the steam used per unit time. Power = force * distance / time
Power is then proportional to RPM
Power is proportional to displacement.
So to get more power you need a bigger engine and/or higher RPM.
Power is generally reduced at higher RPM because flow.
Power may also be reduced or increased by harmonics in the flow path. Tuned port as in Stan paper.

There are basically two methods of throttling. The common one is a simple flow restriction. The second is pressure regulating. In either case the result is a reduced pressure at cutoff. With no change in expansion ratio there is some initial (at cutoff) pressure where the expansion results in the ending expansion pressure equal to the external pressure. (To be clear I am talking about throttling down. From wide open throttle to a point of full expansion engine cycle and then continued throttled inlet pressure reduction) Continued throttling results in the ending pressure below the external exhaust pressure. This is a loss as the engine is doing work to the steam being drawn in when the exhaust opens as it is basically now having to pump that steam bsck out during the exhaust stroke. This is defiantly a problem with a high expansion compound.

From my analysis it is more efficient to reduce the expansions as you throttle. However that doesn't work because decreasing the expansion ratio means increasing the admission duration. That somewhat cancels the power reduction as you would be admitting more steam at a lower pressure work only varies slightly and in some cases even increases.

This is a problem in a single expansion engine as well. The problem is perturbed in a compound because the lower pressure stages become dead stages or worse when figuring pumping losses. Now add to that a higher expansion requires a larger engine and/or higher RPM to produce the
same power. Now you are pumping greater volumes when in over expansion for even greater losses.

The only reason I am looking at a compound is because I need the long cutoff the multistage expansion gives. Three stages of 3 to 1 expansion equals 27 expansions. My plan is to vary power using a combination of displacement, cutoff and clearance. I wont to vary the stroke so that as the stroke decreasses the clearance increases and cutoff is reduced to maintain the expansion ratio. Exhaust would be varied to maintain compression to eliminate the clearance losses. For simplification say at full stroke you have 0 clearance. 1/3 stroke cutoff is a 3 to 1 expansion. If you reduce the stroke to 1/2 about it's center then the clearance is 1/3 of the stroke displacement. with a range of 0 power at 1/2 stroke to full power at full stroke. Of course in the real world we can not have zero clearance. All three stages would be closely power balanced. The real engine has some pressure drop between stages and to avoid overly high temperature in the first stage I use a feed past system that keeps the inter-stage receivers at a minim pressure. Somebody said that would be a loss of efficiency loss. Once running the amount of make up steam (what maintains the inter-stage receiver pressures) is less then 5% of the total steam exhaust steam.

It a very complicated engine. I have used MathCad to calculate 100, 200 and 400 cycles from min cutoff to max cutoff in that mode. Actually many more cycle were calculated as it is an iterative process just calculating the stage cycles with the residual steam use. I was satisfied with 100 points as it was only marginally different of 400 points. In calculating the points I use the previous point as a starting point. Makes it run faster.The input variables calculating the cycles were stroke clearance and cutoff. Stroke being the controlling input, which then fixed initial clearance and the cutoff was found by iteration to make the ending pressure the set value. Then the power balancing would then adjust the ending pressure to balance the stage powers. For an error margin it always has the lower stages using more steam then the previous by a fixed percentage parameter.

Before I ever try to build this I need a method of varying the stroke that can stand the torque and has very very little slop. Have a working simulation in VisSim that tests against a working single stage test engine. Advancing to two sages.

Any suggestions on the mechanism are helpful. The rocker single crank duel piston engine is one I looked at. I think something like that might work only with the piston rod would slide along the driving arm. It looks awful heavy. I favor pushing up the RPM to decrease engine size. Another factor in the design. One thing I noticed in the S.E.S. tuned port charts is that pushing the peak torque higher up in RPM tends to cause the low-end torque to drop off quicker and sooner.

I haven't read Stan's argument against Compounds. But I have to say I agree with Stan on using a compound engine in a variable power application as broad as an automobiles, unless you go to a design that doesn't go into over expansion and keeps the stages balanced. Over a huge power band. I have harped on the power range a lot. Jerry posted that it's really not a cubed relation. But I am not going to change and I hope Jerry is right for using the cube relation puts me on the good side having more speed range than I figured on is much better then having less. Not that I didn't already know what Jerry was talking about. I didn't think it was that much.I know rolling resistance drops with speed increases. I am not so sure Jerry is right on the reason for his test results. Modern cars are dynamically tuned and their drag coefficient isn't constant. They can be tuned to drop off a bit at highway speeds. Some vehicles have active air foils. The Mitsubishi 3000 GT for example. Increasing down force in corners and at speed improves stability.

Re: "Jakuba's Compounding article in the Bulletin"
September 15, 2014 12:01PM
Hello Andy,

I hope the world is finding you well.

Andy wrote:
<Then the power balancing would then adjust the ending pressure to balance the stage powers. For an error margin it always has the lower stages using more steam then the previous by a fixed percentage parameter. >

It looks like your compound engine needed make up steam for the lower pressure stages. Did I get that right? Mine did not. You posted an analysis of that several years ago for, I believe, a three stage compound engine. Was that with re-heat also?

I am going to work out some more lower power output parameters for my engine design. I didn't see that it would have any of the troubles being talked about, even with re-heat. Mine was also multi cylinder so power distribution between compounding stages wasn't given any consideration as each stage alone is balanced enough. But that is a truck engine. For a smaller car 4 mid pressure and 4 low pressure or 2 and 2 might be it, and work output per stage might be needing balancing.

I'll try to get to it soon.

My best to you,

Bill G.
Re: "Jakuba's Compounding article in the Bulletin"
September 15, 2014 03:30PM
Hi Bill

My analysis is over a output power range. Many many cycle calculations. The power range is like a 125 to 1 variation. Actually I am looking at the work output over a 25:1 range. It is a compression cycle. A compression cycle has two stam parts that make the unit steam going through the cycle. The residual (or recycled) steam part and the fresh inlet steam. My engine concept is varying the ratio of inlet steam and residual steam. Call them R and F. F is the fresh admitted steam part and R the residual part recycled from the exhaust not fully evacuating the steam. Of course the clearance space will always contain some steam. Weather compressed, partially compressed, over compressed or not compressed it is residual steam. Full compression would be to the admission steam pressure.

A long long time ago in my distant past. Seams ages ago. I learned about the Williams engines and the controversial compression cycle. Jerry sent me his paper on the Williams engine. Or maybe it was Tom. Not really sure now. Anyway Jerry had used the ideal gas formula in that paper. I thought it would be interesting to calculate using steam properties. I found the ASME steam properties code online. I used it to write my first steam cycle calculator. The problem in calculating a Rankine cycle including compression and clearance is it must be done iteratively. The steam at cutoff when expansion starts is a mix of the residual steam from the clearance space and the fresh inlet steam. At cutoff there is a mixture state that must be calculated. You do not know the state of the residual steam in the beginning for it depends on the compression if any. And that depends on the state of the residual steam in the cylinder at the end of exhaust. Again that depends on the state at the end of expansion dropped to the exhaust pressure. The end of expansion state depends on the steam state at cutoff. And we are back needing the mixture state again. Not a big problem for a computer program. You simply start with 0 residual steam in the mix and calculate the cycle finding a residual steam state. Now you have a residual steam state and can calculate a mix state at cutoff. But you are not finished. You run through the calculations and find a second residual state from the mixed cutoff state. You are still not done. You do it again and again until the cutoff state stabilizes to some specified precision. As the cutoff pressure, in the static cycle, at cutoff is a constant parameter. Pressure and enthalpy determine a state point so I used enthalpy as the property in determining the cycle calculations conclusion. As time progressed better tools appeared. MathCad is a program that provides a readable document with live mathematical formula in mathematical notation. It is kind of like a spread sheet in operation. You change a value and the document updates the calculations. It not only does numerical calculation it also does symbolic, vector and matrix calculations and … You can write function. For engineering it is a far better tool then a spread sheet. I found a steam properties plug-in for MathCad. And used it to implement a compression cycle calculator function in MathCad. Actually several of them taking different parameters to deturm the state points of the cycle. For example one took the expansion and compression ratio. Another took the end of expansion pressure and compression pressure. There are various ways of determining a state point so there were various cycle functions taking different parameters to determine the state points. Also various opinions on how to determine the residual steam state at the start of compression. Armed with MathCad and the cycle function I could easily analyze cycle over a range. For example a throttled engine with fixed expansion and compression ratios over a range of throttling. The cycle function output an array of state points each one is an array containing steam properties pressure, temperature, specific volume, specific entropy, specific enthalpy and quality. The first column of the cycle array was cycle parameters, the rest were state points as described. The second and following array are inlet state point, cutoff state point, end of expansion state point, exhaust state point, and compression state point. The residual steam part was in the first cycle property array. So it was then easy to calculate cycles over a range of throttling for example. MathCad is a very powerful tool. So much more then a spread sheet.

There are two design reasons for the makeup steam. I am trying to keep all stages putting out close to equal power. I am always running close to full expansion cycle. Expansion of each stage is to a set pressure slightly above the external or inter-stage receiver pressures. In most cases a I used a 5 to 15 psi drop. Pressure is need for flow. Using an inlet pressure of 1500 PSIA and 850F And a final exhaust pressure around 18 PSIA The MathCad program calculated 100 multi-stage cycles. I played with the inter-stage pressures to balance the stage power manually looking at plots of the stage work outputs. It is also an iterative process figuring the stage displacement.

The make up steam provides reheat so the inlet temperature is with in reason and helps maintain the power balance between the stages. It's really not a constant expansion ratio. The inlet state and last stage exhaust pressure are constants. The ending pressure of the first and second stage can vary a bit.

That is the where I am on analysis using static cycle calculations. I hope to have a full functioning VisSim simulation running soon. With VisSim I can do time dependent simulations. Flow and heat transfer can be simulated. Right now my steam property plug-in for VisSim is on the back burner. I had to make a major change so it would install. Wasn't problem I knew about until I sent it to VisSim for testing. I had made my steam property units dialog a separate DLL (Dynamic Linked Library). The units dialog DLL was larger then the calculations plug-in DLL. Anyway I had to combine them into one DLL and in the process I broke something. I was in need of new glasses(contacts) at the time and that may have contributed. I must of fat fingered something. I have tests cases from the IAPWS documents that are now returning wrong values. They were the first thing I got working. They are debug blocks that output the values from the raw formulations.

I have new contacts and can see clearly again. I have the new multi-focal hard contacts. Really great once we got them dialed in. I do not need to use reading glasses and can see distance just fine. They are fairly new and took a bit of trial and error to get them right. This is like the fourth pair. The first pair at night made a white out effect with street light. We had to go to a larger diameter to remedy that problem. My eyes didn't seam to like the material and mucus build up was a problem. I was the second person my optometrist fitted with hard multi-focul lenses. I tried soft lenses years ago. They stuck to my eyes. Lost count of the times I scratched my eye ball getting them out. I am far sited so the soft do not let oxygen thorough as well as hard lenses. I had the last pair for 8 years. One finely got broke. But I needed a new prescription anyway.

Right now I an working on a different programming project. Hope to get back to finding my VisSim problem soon. Sometimes it's good to take a break. Coming back I find problems easier.

Re: "Jakuba's Compounding article in the Bulletin"
September 17, 2014 11:36AM

From Bill G re. Cyclone:

It does use some re-generation

I think you are referring to the heated clearance space. No?
I thought regen is bleeding off steam after an HP stage and using it to heat feedwater until it condenses, then put the bleed water back into the feedwater.
The Cyclone doesn't do that. My understanding is that it super duper heats the compression steam beyond admission temperature - it can do this probably in part because the clearance valve is slow, partly because the clearance tube is thick. Is that right, and would you still call it regen? Isn't the proper term a hothead engine?

From Kerry:

Not speaking to anyone in particular here- just making sure we don’t have issues with semantics. With a simple reciprocating steam engine, expansion ratio is the difference in cylinder pressure between the moment the intake valve is fully closed and the moment the exhaust valve or ports begin to open.

Expansion is volume increase. V2/V1 is the expansion ratio that I use.

(volume when exhaust port opens) / (volume at cutoff)

Calculating an unaflow with % of stroke, where:

cl = clearance
co = cutoff
ep = exh. port opening

ER = (cl+ep) / (cl+co)

CR = (cl+ep) / cl

I might be wrong but if so, I would like to know how this odd use of the term "expansion ratio" came about.
Re: "Jakuba's Compounding article in the Bulletin"
September 17, 2014 12:16PM

You are correct. Expansion and compression ratios are volume ratios.


Edited 2 time(s). Last edit at 09/17/2014 12:26PM by steamerandy.
Re: "Jakuba's Compounding article in the Bulletin"
September 17, 2014 12:55PM

The Cyclone uses/used a tight fitting coil wound around the cylinder, to cool the cylinder and to pick up heat from the exhaust. Feed water ran through the coil. A feed water heater in itself is a form of re-generation using the exhaust from the engine. Picking up heat from the cylinder is a different type of re-generation. This takes heat from the steam as it is expanding. The two processes are bundled together in the Cyclone engine. That is the re-generation that I was talking about.

Some turbines used a coil inside of the turbine housing that ran from stage to stage to heat the feed water. This should have been a very effective form of re-generation. For a piston engine to do something similar the coil would be wound around the cylinder. This cools the cylinder so that lubrication is possible at higher inlet temperatures.

Harry mentioned that the inlet temperature was 1,200 deg F. and that further temperature increase happened in the clearance space to 1,500 deg F. In this sense then it is also a hot head engine.


Bill G.
Re: "Jakuba's Compounding article in the Bulletin"
September 18, 2014 12:43PM
Bill, I do not think compression as you describe would make it a hot head engine. A hot head
engine is heated externally. Usually the boiler is built around the head and cylinder. It would
not be unusual for compression temperature to exceed inlet temperature. Eben a una-flow engine.
Note the temperature chart in Stumpf book the Una-Flow steam engine. The temperature is quite a bit
above inlet steam temperature at the end of compression.

The is a lot of argument about the exhaust process. Should be treated as isenthalpic or isentropic
change. The text book cycle (not having compression) treats the exhaust is isenthalpic. I made my
MathCad calculator so it could do a mix of the two. A parameter between 0 and 1 specifies the part
of steam residual steam treated isentropic. 1 is 100% isentropic 0 is 100 % isenthalpic..5 is half
and half. Gus what. 100% isenthalpic exhaust produces the compression temperature seen in Stumpf

I would not call it reheat as it is part of the in engine cycle. All descriptors of reheat, in
thermodynamic books, I have seen occur outside the engine. I think we should try and keep to the
traditional terminology that has been established in many thermodynamic texts and courses. No
matter what. It makes it very hard to discuss the subject when we are not talking about the same

Re: "Jakuba's Compounding article in the Bulletin"
September 18, 2014 04:25PM

I agree that we should keep the terminology as accurate as possible. It was bad enough when I
misunderstood that the Stanley was not a uniflow.

I see the exhaust process as isentropic for the residual steam in the cylinder as it is expanding and doing work pushing the exhaust out. I don't know what the end results would be for wet steam though. With wet steam in the mix it won't re-compress in reverse of the way it expanded. It ends up as superheated steam with suspended water droplets in it fizzing away.

Re-compression does not make a hot head engine. As I understand it, though, the Cyclone engine has its clearance volume in a tube out in the combustion chamber getting heat from there and that would make it an, at least partial hot head engine. A good idea.

Best Regards,

Bill G.
Re: "Jakuba's Compounding article in the Bulletin"
September 18, 2014 06:36PM
I agree that we should keep the terminology as accurate as possible.

Gee Bill do we have to.
My friend wants to use the electrical term PWM to control engine speed in place of a throttle.
Pulse Width Modulation.
Same as some of the early mill engines like the Corliss and the Greene engines.

Sorry, only registered users may post in this forum.

Click here to login