Ted Warbrooke's Stirling 1: Question

[Alphax]

Tom,

Turning now to your other post.

I think you are reminding me that I will respond here in due course (my words). And of course I will! That may be some time away yet, so, as I said (and thank you for reminding me), there is no need for you or anyone else to wait! That remains true, as does my comment that I welcome input from all who are interested in discussing scaling up Thermal Lag engines.

[Alphax]

@ Nobody,


Yes, I share your view that (the relevant) engineers engaged in electric motors are not complacent. I think that is not in small measure due to the fact that electric motors are manufactured in a competitive business where, amongst other attributes, efficiency is important in order to remain commercially competitive to stay in business. So yes - exactly as you say, engineers are ever striving to reduce losses (including financial ones!).

I think Tom is just drawing a parallel between electric motor efficiency and Stirling engine efficiency in order to illustrate one of his points on efficiency.

As far as I can see Tom has begun a thread on efficiency and I think that thread would be a great place to - for example - draw such analogies and parallels.

Regarding scaling up Thermal Lag engines - I have said quite a few times now that I'd much prefer to ignore efficiency considerations if at all possible (I realise that you aren't ,in fact, raising an efficiency point yourself).

Lets discuss your other points about heat engines in the next post.

[Alphax]

@Nobody

I understand the points you are making, and would agree with them (because they are essentially correct).


But I offer this little thought, based on years of experience managing commercial PVT labs - the concepts involved can appear to some folk as being as slippery as eels - i.e. hard to grasp. So whilst you get agreement from me, I wouldn't hold your breath if you are hoping for everyone else to necessarily agree, some doubtless may not.

Your real point, of course, is that knowing something about a measurement of Temperature (the 'T' in PVT) tells you nothing useful about the Volume OR the Pressure (the 'V' and the 'P') anywhere in the closed system that is (by definition) a functioning Stirling engine (of whatever breed).

So yes... you are correct.


To know anything at all useful about what is actually going on within the working fluid inside the engine at any point in the cycle at any position in the engine space is extraordinarily difficult experimentally. A conventional approach in a steady state system might be to confine the local Equation of State to those variables that you can actually instrument for - namely P,V and T (there are others, but PVT are always the usual suspects in instrumenting thermodynamic systems). Knowing just one variable (eg T or delta T from, say thermocouples) tells us nothing of any use in terms of what is happening to the working fluid's volume or pressure at that instantaneous moment. But a working engine works because it is NOT steady state, making conventional instrumentation (difficult enough) even more difficult!

This practical, real world difficulty is one of the reasons that some people do like to fall back on conceptual sketches (such as indicator diagrams, which are about as old as the field of thermodynamics itself) and familiar equations. Such things serve a very useful frame of reference in suggesting what may be occurring within the engine at particular points (power piston positions, usually) within the cycle. And, of course, if we are to discuss things then we need a common technical language (usually thermodynamics) and it is therefore helpful if that common technical language is relatively constant in meaning.

And..... of course.....small engines are harder to instrument than large engines. Ask me how I know.

[Alphax]

@ Nobody

Aha..... back fair and square to the topic!!

Back on topic:
What about a double ended pulse engine? With cold section in the middle and two hot ends and two pulse tubes. Having one piston, magnets and generator in the middle. It could be highly pressurized.

I could get really excited about that idea. Why? Because it could be made by assembling sections to form a gas-tight unit (engine) working with a better gas (than air) as the working fluid at high pressure. With a magnet in the only moving part - the central, oscillating piston. Exactly as you describe. And, of course, the different sections would be made from materials appropriate to the task - one of Tom's interests, IIRC.

Getting into a teensy bit of detail.....briefly..... the engine would have to be such that the central section (I'd be tempted to give it a name such as 'power cylinder' simply because the power piston oscillates within it) must remain cool. Specifically, it would be highly beneficial to keep it below the Curie point of the magnet(s) in/on the oscillating piston. Similarly, the coil(s) would also benefit from staying cool as their performance drops with increasing temperature - even though they would be outside the "power cylinder".


Interestingly, this is not that far removed from one of Tom Booth's concerns, which, as I understand him, concerns trying to get an engine that works without resorting to forced cooling of the "cold end". As I understand him, Tom favours a well designed engine that automatically maintains an ambient cold end without having to maintain ambience by forcing. But, as ever, he'll correct me if I'm unintentionally misrepresenting him!

[Tom Booth]

I think this will likely be my last post on this thread, though I still have no idea what you found noteworthy about my post on the derwood thread or why you encourage additional posts, but don't want to talk about the subject. but anyway...

As far as I can fathom, you want to talk about scaling up a particular type of engine but seem to want to rule out or avoid efficiency considerations, however you continually bring up YOUR opinion that the small engine is inherently inefficient compared with the scaled up large engine, so, by your own analysis scaling up or not scaling up as a subject hinges upon efficiency.

You seem to want material from "us" (other forum members) to support some thesis you are writing.

At any rate, IMO, the prospect of "scaling up" an engine, more than anything else, is intimately tied up with efficiency. Will the big engine work as well, and if not why not?

Certainly any physical machine can be scaled up physically. All it's parts can be made bigger and be put together, what else is there? Efficiency, in one form or another.

But the subject is taboo, for what reason, no one knows.

Derwood demonstrated an engine of the type can be scaled up physically and at least run. The only remaining issue is, does it run as well (as efficiently), and if not why not?

You announce you don't want to discuss efficiency but then single out an old post of mine, squarely on the subject of efficiency. So what is it you actually want?

[Alphax]

Tom,

Certainly any physical machine can be scaled up physically. All it's parts can be made bigger and be put together,......

(my truncation)


Yes, we agree 100%. That can be done with ease.


But;

........ what else is there? Efficiency, in one form or another.

(remainder of quote)


No. This is wrong. Efficiency in one form or another is not all else there is to consider first. The priority is to establish how the various important aspects actually scale because they do not all scale similarly. This consideration "trumps" consideration of efficiency.

A good example is given by Matt Brown (his post a few posts back in this thread). He rightly says:-

Consider you have a sailboat design that you want to enlarge, but unsure as how to precede. To keep this simple, let's say that you want to double the 'size', so

2^1 is delta length
2^2 is delta sail area
2^3 is delta displacement
2^4 is delta heeling moment (tendency to heel)
2^5 is delta righting moment (resistance to heeling)
2^6 is directional stability

Matt is pointing out that some aspects of a sail boat scale as L¹, some (such as areas) as L², some (such as mass and constant internal volumes) as L³, some (such as radiated heat) as L⁴ and so on.


That is why consideration of how different aspects change when scaling up takes precedence over any consideration of efficiency - at least in the first instance. And therefore considering efficiency in preference to scaling parameters puts the cart before the horse.

Take just one very simple example - the heat that an engine loses externally by radiation alone (i.e. neglecting conduction losses and convection losses and all internal losses)per unit surface area is given by the Stefan-Boltzman equation. That quantity is proportional to the fourth power of the difference between surface temperature and ambient. Since SA/V scales as L² by L³, simply changing the length L of the engine automatically changes its radiative thermal loss disproportionately (ie not simply times 2 if the engine is made twice as big).

I know from your previous posts on this thread that you reject that notion, despite its universal acceptance in science. But it is the reason that efficiency is less important than scaling issues when substantially increasing the size of a working engine (it is much less significant when considering small changes in overall size).

[Tom Booth]

To square a measurement can be simplified.

For example, side a is 10 inches, side b is the square of a. Or 100 inches.

You imagine squaring when scaling up an object is some kind of issue but it's not, it's just shorthand.

The actual ratio of side a to side b is 1:10 period.

Same thing with cubed.

The problem you're trying to solve is an illusion. So you may carry on trying to solve it without me.

Enjoy!

[Alphax]

Bye Tom,

Thank you for your input.

Now where were we...........

Ah yes. Nobody's idea...... what about a double ended pulse engine?


On the basis that at least one world expert in Stirling Engines (Allan J. Organ) has explained that Thermal Lag engines are really Pulse Tube engines, and on the basis that the inventor of the Thermal Lag Engine (Peter Tailer) disagrees, it seems safe to assume the following statement may be broadly correct:-

Thermal Lag Engines (including Warbrooke type variants seen on youtube) and Pulse-Tube Engines may be very closely related or even one and the same thing for all practical purposes here.


So.... the terms "pulse-tube" and "Thermal Lag" and "Warbrooke type" may all be fundamentally the same thing. I wouldn't die in a ditch to defend that statement, but it could be useful for now if no-one has any objections.

[Alphax]

@ Nobody

In view of the previous post, I'm happy (as the OP of the thread) to have some discussion about your double-ended pulse tube idea, given that it is essentially two Warbrooke Stirling 1 type engines "back-to-back" (or should that be "cold end to cold end"?)


As you say, Nobody, the advantages (over a single Warbrooke type) are that the common piston (one shared between two hot ends) can be pressurised without needing exotic seals on the piston. The working fluid can be exotic too, as well as being pressurised. And the working fluid pressure differential at either extreme of the stroke may be small compared to the working fluid pressure itself, resulting in minimal "cross talk" leakage between opposing faces of the piston.


Scaling remains a problem that I am interested in, as I was (am) with a single Warbrooke type.

So I suppose the obvious question is this - have any been made and described? Does anyone out there know of any examples???

[Alphax]

So this is a (very!) rough sketch of what a generic Thermal Lag/Pulse Tube/Warbrooke engine consists of:

[Alphax]

And this shows the same generic type with equivalent mechanical processes to illustrate the self-returning (no flywheel) behaviour of the piston.

They comprise two equivalent mechanical processes - a spring (the gas-spring effect of the working fluid) and a dash-pot (the viscous damping effect of the working fluid).

[Alphax]

And putting two together.......

gives a double ended pulse tube engine with only one moving part - the shared piston which oscillates between the two out of phase (by 180 degrees) hot ends.

The clever thing is that the 90 degree phase shift is already inherent in the single Thermal Lag engine and this "displacerless" timing is also automatically available to the back-to-back double ended version as Nobody suggests.

[Alphax]

So, now, I think the same question that I asked about the single Warbrooke engine also arises here - will it scale up?

A supplementary question might be would the double ended pulse tube engine (I think of it as a double ended Warbrooke) actually work?


My opinion is that it would work if one was built. So far I haven't found any examples, so I don't actually know, but I have some confidence that a working model could be built (and, of course, one may yet turn up with some deeper digging on the internet).


And it seems to me that the same general principles of scaling up that might be considered to apply to the single Warbrooke type would also apply to the double ender.

At least that is how it seems to me, so I'd be very interested in any comments. And, of course, my thanks to Nobody for suggesting it.

[Alphax]

Although not strictly the same idea, the liquid piston fluidyne engine shares some of the features of a double ended pulse tube engine (one common oscillating piston shared between two "ends". However, that usually only has one hot end, not two.

[Alphax]

Scaling up the Warbrooke can be tackled mathematically, for those so inclined.

The advantage of a computational method is that individual issues (such as surface area to volume ratios changing and other non-linear scaling features) are handled at the first-principles level by calculation. This allows a genetic approach in which successive generations of solutions are "born" and then "selected" by a neural network whose job it is to selectively breed the next generation of solutions. The selection criterion is increased power output as the design variables are shuffled and the code run again and again to select the most powerful mix of design variables.

Since the design variables are conventionally dimensioned, this provides one way of scaling up the Thermal Lag/Warbrooke type of engine and eliminating non-linear scaling issues.

If anyone is interested, Alborzi et al have published details of how to go about it:-