www.pulse-jets.com

SP-06 -- The Swirl Can Engine

Al, this is the first idea I would like to explore with you. It is extremely simple but needs an introduction so that the reasons behind it become clear. Namely, no one else appears to be thinking in the same direction, so the idea is obviously somewhat strange. I don’t know why – it looks perfectly logical to me. See if it appeals to you. Can you see something where I have gone wrong? Let me know what you think and we’ll talk.

I’ve been looking at pulsejets for perhaps 30 years. Some of their curious traits look like disadvantages to me and I have tried to find ways to remove them or wiggle out somehow if they cannot be removed.
One of the things I dislike is seemingly inherent in the pulsejet – the go-stop-go motion. First we use the partial vacuum to accelerate fresh air on its way into the chamber. Then we stop it dead. After that, we use combustion pressure to move it from standstill and accelerate it down the tailpipe.
To most people, this does not look like a big deal. After all, we are talking of a puff of air that needs to change speed or direction or both. Yet, to me, it has always looked wasteful. I mean, having that puff of gas move in the proper direction is what the jet engine is all about.
But, it took me a long time to come up with a way to avoid the stopping and the change of direction of gas in a pulsejet. The original inspiration came to me from Reynst, who avoided the loss of momentum in his pot combustor by having the mixture swirl in a toroidal vortex around an internal diffuser. I am fascinated by the Reynst concept, I must say. However, the Reynst pot is an unknown quantity, really. No one has seriously researched it but the guy himself. I wanted to see if the principle could be applied to something that was closer to the conventional pulsejet – but without increasing engine complexity.
Today, the answer looks simple and obvious to me, but it took me years to arrive at. I am just not a very quick thinker.
In the end, the trick was in changing the direction of incoming gases. In most conventional pulsejets, gases enter the combustion chamber in the direction of its longitudinal axis. In the Swirl Can, they enter perpendicular to the chamber and tangential to its section. Instead of the intake and exhaust being positioned at the top and bottom of a cylindrical can, I put them on the sides, tangential to the cylinder.
So, fresh air shoots in, follows the wall of the chamber and moves in a spiral motion, like the shape of the coil spring, towards the exhaust, slowing much less than it would if it shot out into the center of the can. It is not diffused because it is kept together between the can wall and the centrifugal force.
What will such motion accomplish? Well, I hope to induce the incoming mixture to generate a single big regular vortex rather than break out in myriad chaotic small ones. This vortex will act as a kind of flywheel, preserving the momentum of the charge.
If Reynst is right, such a vortex also acts a kind of a pump. It rotates away from the intake and thus sucks additional air in. This is helped by the fact that the intake protrudes some way into the chamber.
My bet is that the can will be refilled with a greater amount of mixture than with a conventional intake. This will make up for the fact that the charge is traveling at a greater speed than normal. At greater speed, its dynamic pressure is lower. We’ll have more charge at a somewhat lower pressure.
Keeping up the speed will delay ignition, further helping the swirl pump up the quantity of the mixture. This is considered a very good thing. The most efficient valveless pulsejets have very long induction periods. The French Escopette, for instance spends two thirds (!) of the entire cycle sucking mixture in. It is just about the most fuel-efficient pulsejet around.
True, some retained gas is coming into the chamber from the exhaust and moves in the opposite direction, but we are talking of low-density slow gas from the exhaust, which – I hope – will not be too effective in braking down the high-density fast gas from the intake.
It is generally thought that the ignition in a pulsejet is initiated by the mixing of the fresh charge with the free radicals remaining in the chamber and the tailpipe from the previous cycle. Reynst has shown that the tailpipe gas is not necessary. Both the ordinary jam-jars and the ‘serious’ Reynst pots get ignition without them, using just the free radicals in the boundary layer that clings to the chamber walls. Well, this should work here, too. Certainly the walls are going to be scrubbed quite thoroughly by the mixture swirling across them.
In conventional pulsejets, the incipient ignition that starts between the mixture and the free radicals is triggered into quick propagation throughout the mixture by two things. One is the hammer wave that accompanies the full refilling of the chamber (you can imagine it as the mixture front hitting the back wall of the chamber). The other is the pressure wave reflected from the end of the tailpipe. We have both here, just as in any other pulsejet. The difference is that in the Swirl Can, they will ‘slap’ the fresh charge from the side, rather than head on. This will not impede the circular motion, but will only increase the pre-combustion pressure. In a manner of speaking, the coiled spring will be compressed, but not unwound.
Combustion will take place within the swirl. Because combustion is chaotic and has no particular direction of its own, it will not fight against the direction of the swirl, either. My bet is that at worst, a great deal of the rotating momentum will be preserved right through the process. Indeed, Reynst says that combustion within a vortex adds speed to the spin. I don’t know enough about the vortices to be certain about it, but it sure sounds encouraging.
What keeps a vortex together is the balance of two forces. One is the centrifugal force that tries to break it apart and the other is the low pressure created in the center, which pulls it together. Combustion will produce additional gas and increase the internal pressure, so that this internal force pulling the vortex together will lessen. The centrifugal force will get the upper hand. This force and the rising pressure will both be forcing the exhaust gas out of the engine through the exhaust.
As the swirling combusting mixture reaches the exhaust – while its pressure is growing fast -- the direction of its motion will be right into the exhaust opening. Again, no change of direction – gas will just shoot off into the tailpipe on a tangent, as if released from a catapult, driven by both the gas pressure and the preserved momentum.
This is it, more or less. The intake is shown as a converging nozzle. This is the form it had on the drawing I used as the template – a paper on experiments with valveless pulsejets at the Department of Aerospace Engineering at the Indian Institute of Science, Bangalore, in 1998 and 1999. It may be a good thing, for it gives the incoming charge extra speed. However, it can also be a straight pipe or a diverging diffuser, for all I know. We can experiment with other shapes. I have no clear idea for the best location of the spark plug, but I don’t think it is very important.

Bruno

Hi Bruno,

Here are the sizes for the test:

Intake 5/8" ID X 3" L
Chamber 2 7/8" ID X 5" L
Exhaust 7/8" ID X 17" L

This compares to Sudarshan Kumar's as 1/5 area scale and 1/2 length scale, except for the intake which is 1/2.5 area and 1/3 length scale.

I used MAPP gas and a stinger for the initial test.
Some pops then a few " hoots " and then resonance.
With a few adjustments to fuel and air, and a warm system, the air and spark were shut off and the system sustained.

At low fuel rate, the sound is quite pleasant, and did not require ear protection. Near full throttle, the sound abruptly transitions to a snappy/snarly sound with no change in frequency but a definite increase in thrust ( but still not requiring ear protection ). At a slight increase in fuel flow beyond this point, the system abruptly quits.

Re starts are instantly accomplished ( when warm ) with spark, fuel, and a puff of air.

Your hand can be placed in the exhaust and intake flows within 8" with no discomfort; the intake having a greater apparent flow rate than the exhaust !

Notice the interesting heat pattern, opposite the intake, on the combustion chamber plating.

Al Belli

Congratulations, Bruno and Al! I look forward to further reports!

We can probably whip one up when you're out here, Bruno, if there is something you want to try.

very original! I'm impressed.

one nice thing about it is that it requires virtually no fabrication, like something larry would design.

I'm tempted to try an acetylene fueled micro version!

Hi Bruno,

I made a smaller version of Your SP-06 today, and ran some tests using MAPP fuel.

The sizes are as follow:
Intake - 0.260" Id. X 2" Long ( occluded by a 1/8" Od. fuel probe
Tailpipe _ 5/16" Id. X 8" l. plus a 3/8" Id. extension X 4" 12"
Comb. - 1 3/16" Id. X 2 3/4" l.
The thick wall end caps are to allow press fitting for assembly.

The system would not self-sustain with the 8" L. tailpipe, but ran fine with the added 4" section.

The running characteristics are similar to the larger version, including the " snarley " sound near maximum fuel input.
I could put My hand within 3" of the exhaust and intake with no discomfort.

See the attachment.

Thanks,

Al Belli

Hi Bruno,

Here's a really lousy picture of the small one all warmed up, in the dark. Sorry about the quality, or lack thereof.

Visually, the chamber is glowing medium red, with the exhaust a duller red; and a barely visible blue flame about 3/4" or so long from the non glowing intake.( lower center on picture )

Obviously, the camera does not like lots of infrared !!!!!!

Al Belli

Steve, looks like Al already beat you to the micro!

yea, but it can still go smaller!

Have you been able to get any estimations on thrust? perhaps a larger version running on propane would be more sutible for this.

Al, thank you very much for posting this. Thank you much more for building the thing! Two of them, in fact. It was a hugely enjoyable departure from just thinking about engines. A great favor. I remain forever grateful.

I was soooooooo excited while it was happening. This is the first complete engine from my feverish brain that was actually built since 1999 or 2000, I think (when Gary Robinson in Australia had a Unicone self-sustain). I was incredibly nervous when the time came for testing. I was mortally afraid I'd led Al on a wild goose chase.

No one happier than I when the message from Al arrived, saying the thing worked practically right off the bat!

OK, back to earth. Frankly, I am far from certain that anything important has been accomplished in terms of pulsejet progress. My hopes are high for such a basic-looking device but its function remains to be analyzed. Some smart people have said that my swirl in this can in largely imaginary and that it actually breaks down into pretty much the same kind of turbulence as in a Lockwood. They may be right. We’ll see. Only the numbers will tell the true story, I guess.

One encouraging detail is that the engine remains relatively cool (in pulsejet terms). After more than 10 minutes of operation it will barely glow -- invisibly in daylight; you can only really tell it glows when it's dark. Remember, it works on MAPP gas, which produces a hotter flame than propane!

I hope it means that the heat remains in the gas, rather than being transferred to the casing. That's probably the result of the swirling mix partially insulating the casing from the heat. However, if the heat is retained in the gas, why isn't the exhaust gas hotter? Where the heat goes is anyone’s guess at the moment. Of course, the wild hope is that it all goes into exhaust gas speed, making the Swirl Can more efficient than other pulsejets, but that is just a dream at the moment.

Another encouraging detail is that there's barely a flame coming out of the intake. That means, I hope, that having the gas swirl away from the intake works at least to some extent. Maybe just a slight puff of ram pressure would eliminate this completely. (That’s the next step, folks – the self-ramming pulsejet. I'm serious!)

Several steps in the project remain to be made.

Having both ports at the same end of the can might be a worthwhile experiment. It just might increase compression, but it might also increase intake backflow. There ain’t no such thing as a free lunch.

I’ve been thinking about having the ports flattened, rather than circular in section, the better to conform to the desired pattern of gas flow. The tailpipe would start flattened and then transition into round.

Maybe the ports could also be rectangular in section – flat rectangular tubes.

Lots of thinking to do, lots of work and – I hope – some measurements. Stay tuned!

Thanks again, Al!

steve wrote:I'm tempted to try an acetylene fueled micro version!

Steve, frankly, when I saw your micro-Lockwood, I was sorely tempted to go for that size! But, reason prevailed and I opted for more conventional dimensions. It was Al's idea to go a size lower with the next one. He didn't even tell me ahead of the time -- just sent the pictures!

You might want to try the micro-size with the (somewhat more complex) self-ramming version. Given that the small engines are trickier than big ones, the ram effect might make it easier!

More about it after the Pinot Flambé Conference.

Bruno -

Yes, congratulations are definitely in order - nice work!

I would like to see this done with a "hollow disk" chamber, rather than a long cylinder. Let me know if you ever want one done up out of a pair of smallish torque converter housings! My feeling is that the core space in the middle of the swirl accomplishes little [as long as you still provide sufficient expansion space] so that a toroidal space might get you the same action more efficiently [or, maybe not].

Don't forget about the elusive property called "moment of inertia". Essentially, IF you work from a rigid central shaft, high moment of inertia is against you; this is why a skater in a spin [I don't remember the technical term] speeds up by pulling the arms and legs tighter in. This would suggest that you might try a cone with the intake coming in tangentially at the large end and exiting tangentially at the small end, for example. The large circular end of the cone could be replaced with a washer and inner cone to take up the dead space, i.e. you would go from a slow toroidal flow to a fast, more compact "cylindrical" flow before exiting.

On the other hand, there is an interesting thing when you roll solid shapes down a ramp: for a given ramp, ANY solid disk will beat ANY sphere, and ANY hoop or torus will beat ANY disk! This is the "other side" of moment of inertia, i.e. propulsion FROM THE EDGE, and this might more closely resemble your swirling tract of expanding gas. So, maybe the cone should be the other way round, with the intake at the small end near the supposedly "low pressure" center and the exhaust at the expanded end [which could still have an inner cone to occupy the supposedly "dead" zone, if needed or desired].

Of course, I do realize that the gas behavior is a far cry from a spinning machine with rigid swinging arms. So, perhaps "none of the above" makes any sense at all. You might learn something by considering how those big "cyclone" dust precipitators work that they use on a lot of industrial plants. Of course, that air isn't heated and rapidly expanding, and it isn't self-propulsive, but the swirl is certainly there.

Anyway, the future possibilities for research seem almost limitless, which makes something like this seem really exciting ... along with the fact that building the basic design can be pretty simple, as Steve has said already. Lots of potential discovery ahead!

L Cottrill

I was thinking of something more gradual, keeping the combustion chamber more or less linear to the exhaust, but introduce the tail pipe and intake at an angle, kind of like a section of spiral stair case affixed to a verticle shaft.
Mark

To me, the most amazing aspect of the whole thing was the builder, Al Belli. I was sending this concept to him in installments, building up to a dramatic finale. I was just getting into the stride, warming up to the task of explaining the fundamentals of the idea, the various ramifications and implications, when he up and built the first prototype. I had barely collected my thoughts on it -- I still have unfinished drafts -- when he build another, smaller one. Al builds faster than I talk. I am totally shocked.

Larry, interesting thinking.

I have just scratched the surface of possibilities, concentrating solely on one aspect -- preservation of speed; keeping the vectors aligned, as it were.

Vortices are interesting beasts and deserve study. There have been designs in the past that took note of some of the things you say -- I remember one with the hollow core, for instance.

Unfortunately, I know very little. (That's why I kep to the basics.) Much learning is in order.

We really need a sound recording.

and video if possible!