Proposed: Simple Cyclical Flow Engine

[larry cottrill]

Rossco [or whoever might want to try it] -

OK, here's one that just came to me today. A simple cyclical flow engine with two flat sheet sides and the rest made of simply bent sheet steel. Here's my design thinking:

The usual patent drawing for a cyclical flow engine is some variation of the following: An ejector intake feeding a combustion chamber which nozzles down into a short exhaust stack, where high velocity hot gas is picked up and brought forward through a narrow duct to drive the ejector. Back in 2000 and 2001 I ran across maybe a dozen of patents which were simply variations on this theme, and there must be dozens more I've never seen. Some of them had multi-stage ejectors, etc., but they were all some transmogrification of this basic plan.

The problem here, in my opinion, is the inefficiency of the recycling duct - lousy Reynolds number, big radiation losses, etc. etc. The ejector is fine, if well-designed, but probably in reality can't make up for the huge losses in the duct, at least in the case of small engines. What we need is a way to eliminate the long duct. Today I thought of one way to do this, to wit:

Instead of an ejector at the front, design a big augmentor. Instead of tapping into the exhaust stream, tap into the augmentor stream at a place where the velocity is still pretty high. Instead of a long hot gas duct, use a long intake duct. Instead of an inefficient "can type" chamber, make the active part of the engine a decent ramjet shape.

Build the thing with sheet metal of constant width between two flat sides, with the augmentor running along the bottom. Make the bottom sheet flat, so the top of the augmentor is bent to form the full contour needed for augmentation. Now, let this bent sheet also form the "fishbelly" shape that defines the diffuser, combustor and nozzle sections of the ramjet - this allows you to leave the top flat, too. There are only three smooth cylindrical segments needed - one for the outside of the sharp bend from the intake duct into the diffuser, and two to form the exhaust nozzle turnabout.

The first drawing shows the basic "typical" cyclical flow design as mentioned above; the second is my re-work using an augmentor to drive the cyclical flow. What do you think?

Note that the intake duct and exhaust duct seem undersized in the drawing. This is because we are not talking about circular sections here; they are full-width areas between the two flat walls of the engine, i.e. the same width across as the augmentor and chamber.

Some experimentation required ...

L Cottrill

[Bruno Ogorelec]

I like it. Of course, I have no idea what the figures would say about it, but I like it.

What I like the most is the adaptability for the airfoil shape. You can have the combustor iside the wing and only the augmenter duct sticking out below.

The simplicity is compelling. Are you thinking of building and testing it, Larry?

[larry cottrill]

Bruno -

Thanks. I'd like to try it - it's one of the few designs I've done where I could actually cut and bend the metal myself!

It does have one glaring construction problem, though: It isn't all that easy to weld an interior steel partition between two parallel sheet faces. I could probably do it by drilling a zillion little holes [imagine a row of holes where you can see the edge of the sheet partition at the bottom of each hole] and sort of blending closely neighboring plug welds together. Right offhand, that's the only method I can think of that I could actually do.

It would be fun to try this sucker, though! Maybe Rossco or Al or somebody can think of a more workable method of construction.

Glad you like it.

L Cottrill

[steve]

wow! do you think it would actually work? have any of the patented designs been tested? I would really like to see someone try it!

[Bruno Ogorelec]

It isn't all that easy to weld an interior steel partition between two parallel sheet faces. I could probably do it by drilling a zillion little holes [imagine a row of holes where you can see the edge of the sheet partition at the bottom of each hole] and sort of blending closely neighboring plug welds together. Right offhand, that's the only method I can think of that I could actually do.

Have you seen that posting in the forum about a homebuilt spot welder? That sounds like a useful tool. If you make the electrode holders long enough you could reach pretty deeply into such box-like structures. Of course, you'd have to bend a flange at the edge of the perpendicular sheet; you can't spot weld an edge to a flat surface.

I'd really like to see this thing tried.

[matt512s]

Hi;
Looks really cool! Imagine your drawing as a longitudinal section of
an annular combustor. Do you think that is feasible?
Matt

[Rossco]

EDIT, i was in a bad mood.
I deleat, and withdraw my comment. Having read this properly, with now a better view of what is here, i shall not jump to conclusions again.

I understand that you have the freedom of speach, and can put things into Paint very elegantly, although doing so with what i describe, (even if thats not where it came from) and am doing in another thread got a impulseive comment without thought.

Rossco

[matt512s]

Sorry.

[Rossco]

On a more positive comment.
Larry, your ejector exhast stream would have to be possitioned on the outside of the augmentor wall. As in the confuguration depicted, i would say a lot of the exhasht stream would be injected back into the engine.

What i am working on is a separation of these, by way of the coanda effect, sicking the hot flow along the wall to exit with minimal mixing with the cold.
I am working it in steel tho, so its position is physicaly tested, not just to look good, which nicely centred i admit does.

Rossco

[larry cottrill]

Rossco -

Man, I'm sorry I missed it! I was eating dinner at the time ;-) If it got really nasty, I forgive you sincerely and profusely for whatever was put forth.

Needless to say, I did not intend to give away your cards. You kept talking about an 'ejector' rather than an augmentor -- I thought you were basically working on some sort of refinement to something like the front end of my first drawing. Going back to look, I now see that you said,

A small flow through the intake and out the ejector mouth creates a great volume of air to flow, at some small pressure around the clean air intake piping growing exponentially with input velocity.

... so, indeed, you were describing something [a "point of novelty"] very much like what I ended up illustrating. And, it did not occur to me until after I had read your post as well as studying my first drawing [i.e. the "typical" cyclical jet pattern]. Where you said "around the clean air piping," it did not register with me that you were describing an air entrance, and I really didn't understand why this mention was an essential detail in your description.

So, let it be declared herewith that as to tapping off mixed air at high velocity from the output of an exhaust augmentor for the purpose of regeneratively aspirating the source combustion process, Rossco shall have precedence. Have it, and welcome.

Ross, believe me, I am seldom very hard to deal with.

I apologize for revealing something that you felt was worth playing close to the vest for a while, and especially for the appearance of taking it over and presenting it as my own stock in trade, an action which would most certainly make me look bad [and should have, if it had been intentional]. I know it is not the first occasion on which I have unwittingly revealed someone else's "secrets".

L Cottrill

[Rossco]

Thanks Larry, its unnnessasary.
Such a sincere appoligy and forgiveness is seldome given for something that was not, at all your fault.
I had one of those knee jerk reactions that was quite inappropriat.

My post was not really nasty, more rude, as i hadnt even read your input. You didnt miss anything.
I looked at the pics (very elegant btw) and the first line or two, then bashed in my abnoctious assumption that youd taken "my" project to a new thread.

No hard feelings, and i will be cirtain to let you know of the progress. Im not keeping anything to my chest, i just havnt got anything out on paper as each step needs to be varified and optimised as i go.
My descriptions are usualy a ramble, so what is a definative point could be anywhere in the middle.

You put a smile back on my face on an otherwise crappy day, there is some decent people still around after all.

Rossco

[larry cottrill]

As in the configuration depicted, I would say a lot of the exhaust stream would be injected back into the engine.

Ross, I don't think it matters as much as you might think. The mixing action of an ejector basically takes place in the gap between the exit face of the driving stream and the throat, not within the throat itself. The other pertinent fact is that the flame from good lean burning has a lot of unused oxygen left floating around in it [making afterburning possible, for example].

This reminds me of some useful things told to me by a guy who designs ejectors for a living [and these would presumably apply to what we call augmentors as well]. First, ejectors are not usually designed by formula, because that can yield only approximate results; instead, each company keeps tables of empirical data on designs they have tested, and they refer to those for guidance on new designs! Second, the absolute number one most important design variable is the distance between the spout face of the driving stream and the front end of the throat section. Here's how it works: The larger the gap, the more entrainment from the surrounding fluid you'll get, and the slower the exit velocity will be from the device as a whole. In other words, as you increase the gap, you trade off final velocity for the increase in mass flow. All other design parameters pale in comparison to this one dimension.

What this means to us is, it would be wise to experiment with a configuration where this gap could be easily varied as you go!

L Cottrill

[Greg O'Bryant]

Larry,Rossco;
Do you think that the ejector would have enough power to get the air moving in the intake to above supersonic speeds? And if it did would you be able to make a supersonic intake, perhaps with a little restriction in it? If that could be done I think you would really have something. We all know tht ram jets work their best at supersonic speeds. I love the idea and have been thinking of something similar but with a pulse jet as the combuster instead of a ram. I was thinking of having a pulse jet turn back on it'self and then act like a ram jet to harness the radient heat loss as well as maybe adding a little ram preasure to the engine. Why spin the metal in a turbine when you can spin the air?

Good Luck:)

[larry cottrill]

Do you think that the ejector would have enough power to get the air moving in the intake to above supersonic speeds? And if it did would you be able to make a supersonic intake, perhaps with a little restriction in it? If that could be done I think you would really have something. We all know that ram jets work their best at supersonic speeds.

Greg, you might really have something anyway, even at subsonic speeds. The renowned "efficiency" of pulsejets at supersonic speeds is only one advantage - to me, the real issue is the "zero maintenance" simplicity inherent in the design. You can afford poorer fuel consumption if you reduce maintenance far enough! Of course, that ignores environmental impacts, etc. ad nauseum.

I doubt that the ejector as we know it will ever be a supersonic device. Remember, what comes out is always slower than what goes in, because of the mass entrainment. Maybe you could design one that's optimized for high speed, and get something like you're talking about. But, it would be quite a bit different device from the simple ejector used at static speed. It's really hard to say how far you could take it.

I love the idea and have been thinking of something similar but with a pulse jet as the combuster instead of a ram. I was thinking of having a pulse jet turn back on itself and then act like a ram jet to harness the radient heat loss as well as maybe adding a little ram preasure to the engine.

My first grab at the cyclical flow idea was back in the late 1960s; the idea was basically to create an add-on gimmick to my Dynajet that you would flip down into place to use the exhaust to boost intake air. I had no idea of what the resonance effects might do, of course, though I was too afraid of that kind of problem to waste money trying to build it. Even then, I knew that heat losses through a long duct would be extremely deleterious. But, it you have a more sophisticated idea that keeps the pressure wave in sync, why not? Go for it!

Why spin the metal in a turbine when you can spin the air?

That's the basic idea behind all cyclical designs; it's what I like to call the Hot Gas Flywheel. In its essence, it's no different from picking up exhaust with turbine buckets to spin up a compressor - but with a LOT less moving mass. Of course, like any other humanly designed gimmick, it will have problems and inefficiencies of its own.

Look how far the gas turbine and its derivative propulsion engines have come. The first practical gas turbine was by the French team of Armengaud and Lemale in 1905. It had three separate compressor stages to drive the turbine, weighed who knows how many tons and was about the size of a locomotive. Of course, the whole point was shaft power. When they fired that baby up, it managed to deliver ONE PERCENT of the energy consumed as usable shaft torque! If you had poured millions of francs into that effort and were told it had that kind of performance, would you bet millions more to go on and perfect the technology? 100 years later, lightweight gas turbines have flown people around the world [non-stop in one very well-optimized scenario].

L Cottrill

[Bruno Ogorelec]

Why spin the metal in a turbine when you can spin the air?

That's the basic idea behind all cyclical designs; it's what I like to call the Hot Gas Flywheel. In its essence, it's no different from picking up exhaust with turbine buckets to spin up a compressor - but with a LOT less moving mass.

Some things depend on perspective. Look at the flywheel on a piston engine. It does no work – it just spins. The whole point of flywheel is ‘spinning the metal’ as it were.

What do you get by reducing the flywheel mass? Quicker response. The engine revs up and down more quickly. What do you get by increasing the mass? Smoothness. A four-cylinder engine with a flywheel works as smoothly as an eight-cylinder without one. A single cylinder engine will not work without a flywheel. The working cycle is too uneven and something is needed to help the engine bridge the gaps between the power bangs. So, the point is not in whether to have the spinning mass or not, but how much to have.

Now, back to jet engines. A cyclical flow engine can do without a massive flywheel because its operation is smooth. It is an 'eight-cylinder' engine in terms of smoothness. A pulsejet is very rough because its operation is rough by design. It is a single-cylinder engine in terms of smoothness. A flywheel would do it much good. But, it doesn’t have one.

What it means in practice is that you have to create a perfect set of conditions for the pulsejet to work. It will do so only within a very narrow set of parameters. Move only slightly out of that set and the engine dies – instantly and without recovery. A pulsejet either works or it doesn’t.

So, wouldn’t a flywheel be nice to have, huh? That’s exactly what I have been working on for months now, on and off. A pulsejet flywheel.