Proposed CC design

[Ogge]

Adam -

Now that we can all see it better, why not have a two-piece design with the entire front shell being a large portion of a prolate ellipsoid (back to the egg shape ;-) with one focus at the ignition point and the other focus at the exhaust pipe portal?
L Cottrill

You lost me here. Whats a prolate ellipsoid?

[hinote]

But there is another aspect. There will be 2 sucessive waves coming from the exhaust tube. Its actually the intersection of those 2 waves that I envision will initiate ignition.

.

That creates another bone of contention--namely, the ignition of the fresh charge in these engines occurs in stagnant zones along the walls of the combustion chamber.

Whether from hot metal, free radicals, or ongoing combustion--it's widely agreed this is where reignition begins. It seems you are proposing a whole new location for this to occur, and it flies in the face of existing (successful) pulsejet design.

Somebody's going to have to build your proposal--and then try to run it--before I'm going to be convinced.

Bill H.
Acoustic Propulsion Concepts

".......some day soon we'll be flying airplanes powered by pulsejets."

[pezman]

A prolate ellipsoid (or prolate spheroid) is a stretched spheroid -- a.k.a an egg.

An oblate spheroid is what your basket ball becomes if you sit on it.

[jmhdx]

Hi Adam, with regard to the picture posted by Bill(Hinote) I think I see your point.
If a flash of light originated in the spherical end of the bell, the walls were mirrored and the intake and exhaust were of equal exit area indeed more light would escape through the exhaust.
I believe it would "focus" the wave in favour of the exhaust, but what about the mass flow? We want to squeeze as much of the expanding gases out of the exhaust without having to sheild the intakes. Uflow can't model a complex chamber form much to our dissapointment.
Mike.

[Ogge]

Hi Adam, with regard to the picture posted by Bill(Hinote) I think I see your point.
If a flash of light originated in the spherical end of the bell, the walls were mirrored and the intake and exhaust were of equal exit area indeed more light would escape through the exhaust.
I believe it would "focus" the wave in favour of the exhaust, but what about the mass flow? We want to squeeze as much of the expanding gases out of the exhaust without having to sheild the intakes. Uflow can't model a complex chamber form much to our dissapointment.
Mike.

Any reason the intake and exhaust need to be equal area? I was thinking something like an exhaust at 2" diameter vs 3 1" intakes or smaller. That would give a 1.33 ratio of exhaust/intake area. Of course the larger number of intakes the smaller the diameter of each would need to be. The main reason for multiple intakes would be mixing and direction of intersecting transverse waves originating from the intakes.

There are also other tricks like crimpting the intake into a nozzle to limit the mass flow. But my main focus has been on making sure the majority of the heatwaves are propogated down the exhaust, to accelerate the cold mass there.





Another option would be to have a bigger exhaust port that gradually narrowed for a lenght like the FWE allowing a larger diameter. Unfortunatly I have not envisioned all the consequences to that modification. Im unsure how it would effect the wave propogation.

[Ogge]

Adam -

Now that we can all see it better, why not have a two-piece design with the entire front shell being a large portion of a prolate ellipsoid (back to the egg shape ;-) with one focus at the ignition point and the other focus at the exhaust pipe portal? The rear wall would still be your spherical reflector. That way, all the wave travel from the ignition point to the front shell wall would have the same (fairly short) path lengths to the pipe and all the wave travel that gets reflected off the back wall and re-reflected off the front wall would have the same (long) path lengths? It seems to me that would be a highly efficient solution, with simple forms and a single circumferential seam.

L Cottrill

Im afraid I dont know the math involved with a prolate ellipsoid. How does it have 2 focal points? It may be a viable alternative if I understand correctly. The question I would have is 'If the point source is the 1st focal point and reflect off the side wall, will it then focus on the 2nd focal point? If so, it would be perfect.

I was thinking of a 2 part myself. I was goning to find a nice cheap bell then use a parabolict bit on the dome to shape it and have the spherical section welled on. I found a large selection of bells on the net for a dollar or 2 each.

[larry cottrill]

Adam -

Yes, it is an ellipsoid made by spinning an ellipse on its longest axis. The reflected path between the two foci is the same length no matter where on the shell the outgoing wave strikes.

The paraboloid is the perfect shape for focusing an image of a point source that is "infinitely" distant. The ellipsoid as described is perfect for focusing the image of a source point nearby.

Even the graphic layout of the cross-section of the ellipsoid is simple; you can do it with two tacks at the foci and a loop of non-stretchable wire hooked around them, with your pencil as a third point pulling it tight as you "circle" around them. Since the part of the loop stretched between the two foci is constant, you can easily see that what remains of the loop represents a constant optical path length between the two points via the outer point. Proof that the curved line is at the correct angle to accomplish this is more difficult, but it is, at every point on the curve.

L Cottrill

[Ogge]

Adam -

Now that we can all see it better, why not have a two-piece design with the entire front shell being a large portion of a prolate ellipsoid (back to the egg shape ;-) with one focus at the ignition point and the other focus at the exhaust pipe portal? The rear wall would still be your spherical reflector. That way, all the wave travel from the ignition point to the front shell wall would have the same (fairly short) path lengths to the pipe and all the wave travel that gets reflected off the back wall and re-reflected off the front wall would have the same (long) path lengths? It seems to me that would be a highly efficient solution, with simple forms and a single circumferential seam.

L Cottrill

Ok, refreshed my memory on elipsiod focal issues. Yes it would work like this tho I would prefer any point forward of the focal point directly in front of the exhaust not be ellisoid but parabaloid.

We want the focus of the front cap directly infront of the exhaust to have an infinite focus Not a focus at the entrance.

With an elipsoid focus, we would have to deal with transverse waves traveling down the exhaust tube instead of vectors directly parallel the axis. Since the waves would zigzag down the exhaust tube, much of the pressure concentration would be spread out or lost and the distance would be longer to travel like lightwaves down a fiberoptic cable.

I expected this for any vector reflected from the side of my CC design. Waves reflected in this maner will help to heat the mass piston for the exhaust but you will probably bleed to much energy into the walls to have any effective reflective value from the exhaust reflection. The resulting pressurewave will also be fairly distributed based on the angle it entered the exhaust tube effectiving spreading the resulting pressure wave. The wave should hit the exhaust between the front cap reflection and the rear mirror reflection.

Using an elipsiod focus for the very front cap would also be a problem on the return reflection from the exhaust. It would be very hard to get these waves to hit the same focus at the exhaust port after zigzagging down the exhuast.

So a parabolic front cap equal in diameter to exhaust port, then elispoidal to the mid point between the two focuses of the elispoid with a spherical back reflector.

Well gentlemen, I think we just layed an Egg! The only other concern I have with this design vs the bell shape is the volume of the CC. Basicly instead of putting the sphere at the halfway point, I reversed the curve to broaden the CC and used a much larger spherical radius.

[larry cottrill]

Adam -

If you're building to a fairly small scale, the difference between having a parboloid at the end and keeping the ellipsoidal part intact wouldn't matter. This is because the Rayleigh Criterion holds that the image won't be perverted by any defect smaller than 1/4 wave - you can be off that much and still have a perfect diffraction pattern image. What you would do in a small engine is you would move the ellipsoid just a hair toward the tailpipe opening, so your source point would fall between the ideal elliptical focus and the focal point of the parabola that most nearly fits. These points will be only a few mm apart, unless you're building a very large model. In surfaces the diameter of the exhaust pipe, these two curves would be practically indistinguishable, and certainly far less different than 1/4 of the wavelengths you are talking about.

This probably seems nonsensical, because you're probably thinking about the waves from the source focusing to a tiny point at the other focus, as would light waves. But that can't happen. The best "image" of a point you will get with such small surfaces, even if the curvatures are optically perfect, will be a diffraction pattern significantly bigger than 1 wavelength across, with the majority of the energy piled up in a central 'hump' maybe 1/2 wave across. That's as good as it gets, until you make the whole reflective surface many wavelengths in diameter. Even with the relatively short wavelengths you've described, you simply cannot focus to point images, as I said before.

This can be hard for people to visualize sometimes, but it can even be demonstarted with light waves. Take an index card and punch four holes side by side: a small nail hole, a thumbtack hole, a large needle hole, and the smallest needle hole you can see light through. Take it outside at night and look through each hole in turn at a distant bright street light or yard light. Obviously, the smaller the hole, the dimmer the light will look - but there's another difference: Through the tiniest hole, a clear image is impossible. It turns into a central blob surrounded by a ring [two may be visible if the light is very bright]. There is also a separation of colors. This is diffraction, and the degree of the effect is always inversely proportional to the clear aperture. The effect in your engine will be vastly larger, because the size of the "aperture" [the chamber diameter] is only a few wavelengths. The higher frequency energy will focus much better than the lower.

L Cottrill

[Mike Everman]

Larry, well put, I like the diffraction explanation.

Adam, you so need to burn some fuel, dude. ;-)

[Ogge]

Ok, I conceed. You win this round.

[Ogge]

Larry, well put, I like the diffraction explanation.

Adam, you so need to burn some fuel, dude. ;-)

Would love to! But until I get a job, I have time for hobbys but no money!

[larry cottrill]

Ok, I conceed. You win this round.

Adam -

I should emphasize again that I'm not trying to be intentionally perverse or discouraging, just realistic. I still think the design would make a wonderful engine, as a true pulsejet or as a wave engine. As a pulsejet, it would really be unlike anything we've seen before, and who knows what the behaviour might be!

Of course, I have no idea how you're going to form that ellipsoidal nose, but when you can, I say, go for it! I think it would be a beautiful thing to build and test, if you can get the necessary precision in the curved forms. Unfortunately, this one doesn't lend itself very well to one-dimensional analysis, so UFLOW1D won't model it well at all.

L Cottrill

[Ogge]

Well it looks like mark may have solved the CC chamber construction issue. He got a hookah on ebay that is basicly the proposed shape. The top isnt exactly spherical and the very bottom is flat but it may be close enough.

Guess we will see when he gets it going.

[Ogge]

Im gonna have to retract my agreement about the egg shaped chamber. It wont work out the way I expected.

The design REQUIRES that the radius of the spherical mirror be to the focus of the front elipse/parabala. This is required to get the correct timing between successive waves traveling down the exhaust and for them to be correctly timed on the reflection waves to drive added compression at the focus. By placing the mirror at the wrong distance, the timing of the intersecting waves would drift thereby negating the desired effect.

In order to do this with an elipse and a spherical mirror, the radius of the mirror would be way to small and the reflections off the side walls could not enter the exhaust port due to being focused at a point alot farther down the exhaust port. Waves could not cut the corner of the exhaust port mouth. The only way to do this would be to have the mouth of the exhaust port much larger in diameter then taper down or go back to my original design of allowing the side walls to flare outward for a greater distance.

This could be done by making the front section of the CC a half elipsoid with the 2nd focus at the exhaust port mouth or even a little inside it. Once the halfway point is reach, the eliptical curve will need to reverse to widen back out some before intersecting with the spherical mirror to place the mirror at the correct distance. This will make the spherical mirror/sidewall boundary a sharp angle.

The other option would be to make the area between half elipse and spherical mirror a partial conical section. This would make the overall shape of the CC a rounded nose on a cone with a spherical back mirror which shouldnt be too hard to fabricate.

Making the exhaust port mouth a gradually narrowing cone like the exhaust side of the FWE CC design may even be a good idea due to the inablility to focus the waves as accuratly as I would like for smaller geometry engines. This will also allow the area of the exhaust port mouth to be much greather then the intake port area helping to direct the mass flow due to the so called "constant pressure expansion". This will not be required for large geometry engines tho I dont think it would hurt any.

"Constant pressure expansion" is used in thermodynamics to represent an average or under the conditions that combustion is a continous process like in rocket, turbine jet or boilers where combustion is an ongoing process that generates a steady pressure at the exhaust. This is NOT the case in the wildly varrying pressures in a pulsejet CC.

Due to the pulsating nature of combustion, we cannot ever assume reaching a steady state pressure in the CC. We have reflective waves from expansion areas (intake/exhaust), variations from below atmospheric to hi pressure due to combustion cycle. We also have waves reflection off the walls generating a continously varying pressure fluctuation at the intake/exhaust port mouth.

So the goal here is to harness these fluctuation to get them to due work in an orderly and predictable way successivly adding energy in phase with the reflective waves. Thereby adding energy in sinc with the oscillations allowing them to grow to max patiential.

A good example of this is the way Nicola Testla caused a manmade earthquake in New York city with a small handheld oscillator. He later claimed he could bring down the Empire State building with a similar device the size of his hand. All the device did was to induce a vibration into the support colums of a building and triggered another pulse on the reflective return of the pulse continually adding a little more amplitude to the oscilation each cycle. It built up until windows were breaking for blocks away and people thought an earthquake hit the city.