Radial inlet

[GRIM]

Heres an idea that keeps coming back to me ,
until someone tells me why it wont work it will keep bugging me.

I am not sure theres much to be gained , maybe more symetrical combustion?

Thinking along the lines of a 3Lb size

OK guys rip it to pieces and put me out of my misery !

[pezman]

I think that there have been a few coaxial engines that have been made and I'm pretty sure a few of these have actually been successfully operated.

One problem with having OD of the inlet "annulus" match that of the combustion chamber is that you are basically maximizing the friction. It might be better to make the coaxial intake be as small in diameter as possible (e.g. if the CC diameter is Dcc, then the exhaust diameter might be Dcc/2 and the shell for the coaxial inlet might be 1.414 * Dcc/2. This gives you the same intake area with an inner wall whose surface area is a bit over 60% of that shown in your design (less interior surface -> less friction -> less damping effect).

The integrated fueling is probably good, since it is desirable for the inlet to be cool relative to the exhaust and sharing a wall with the exhaust is likely to undermine this a little. The cooling effect of the fuel seems like it would offset that, which seems a good thing.

If you can get it to work, it would be great -- the radial designs are certainly a lot simpler to fabricate.

[Mike Everman]

I agree. More surface area per area has been bad in my experience. It's a lot of drag when you want to move gas out quickly.
I don't know how many times I've sketched different approaches to that, so yes, it is compelling!

Another thing to consider is that you'd think centering isn't that important, if your mindset is Ao-Ai being the intake area. There is a boundary layer that is attached to all surfaces, it's thickness effectively reducing the area by a small amount, 1-2mm as an example scale. When you calculate the area you need for your intake, with the outer wall as your CC, you see that it is a very small annular distance, and easily lets you overlap your boundary layers, even if you accounted for them.

So, in one sense you've got a variable intake just by moving your centerbody around, but I think the narrow side will make the exact wrong kind of turbulence. Otherwise, all ideas are worth a try. you can't predict what you'll learn in the process.

[Jutte]

Would such a small annular gap of such an intake cause a toroidal
vortex ala the Reynst?But wouldn't the resulting boundary 'friction' also cause a 'choking' of the inflow from the intake resulting in only a low resultant rate of pulsating combustion? Also if you extend a secondary enclosed cylinder around the intake you sort of get what the Argus engineers were playing with.
What marvelous questions arise from what essentially are empty tubes - man these things are neat!

[Bruno Ogorelec]

I agree. More surface area per area has been bad in my experience. It's a lot of drag when you want to move gas out quickly.

Well, who says the section of the annular duct has to be the same as in a corresponding tubular duct? What you want is the same flow rate.; never mind the section.

Don't just recalculate circular area values into annular areas. Recalculate the area to give you the same cool air flow. Or, if calculation is uncertain due to unknown temperature and pressure values, fiddle with different duct sections.

Given the greater drag, you will obviously need greater duct area to maintain the required gas flow. That's probably all there is to it.

Dave Gibbel has made a superbly simple design work well.

[Mark]

I remember Tharratt did one of those capped pulsejets and the results were less than he had hoped for.
Just shooting from the hip, I imagine as air charges in and out around the sleeve all 360 degrees, that the charge has more of an opportunity to become uneven/defrayed, or on one side for whatever reason or vagary, an erratic flame front develops from several points and might dampen/baffle the feedback, moreso than a straight duct that conveys a confined charge, a pressure that can't "explore" escape routes so easily or get ahead of itself.

[PyroJoe]

The surface turbulence will probably be your weak point. If I where to approach that design, I would polish the close proximity surfaces to near mirror finish.

[pezman]

Well, who says the section of the annular duct has to be the same as in a corresponding tubular duct?

No one, as far as I can tell. I think that there is general agreement that you would need to fiddle with the duct dimensions to get them where they need to be. I just proposed "1/4 CC cross section" as a starting point. My gut feeling is that the size would ultimately need to be adjusted upward

I guess that it comes down to how important the "inductance" of the intake is in a valveless pj, since the increased friction of an annular intake seems like it would effectively lower the "Q" of the system (quotes here to denote a degree of irony in using linear, electrical, lumped parameter concept to describe effects in a non-linear, mechanical, spatially distributed system). If the intake inductance is unimportant (or relatively unimportant vs that of the exhaust), then for sure, you could tweak the system by adjusting the diameter to offset frictional effects -- and the resulting decrease in "inductance" could probably be disregarded.

Just as a note, if I recall correctly, the mechanical equivalent of inductance is "inertance" (and the mechanical equivalent of capacitance is "compliance"?) -- but if I'm going to abuse terminology, I figure why not go the whole nine yards.

At any rate, it seems like a question that is best settled empirically, since any simulations of that geometry probably need a pretty serious CFD code.

The more I think about this, the more I want to build one. The seeming ease of fabrication appeals to me.

[Mike Everman]

Hi Bruno!
Yes, Dave can make a good motor. I wish he could play nice. And of course I was not proposing simple conversions of area, I was preaching exactly the opposite.

What do you mean, "made it work well"? Yes, it's a layout we've all considered and some of us have made, but the point is, is it really better than an equivalent minimum surface area intake?

[larry cottrill]

Note that (with considerably more effort) you could make this so the clear area could be experimentally adjusted. The trick is, once you know what length range is right for the intake cap, make a new engine just the same but with the end cap wall and chamber wall as matching cones (very narrow ones, I mean) - a tiny adjustment in cap depth can then be used to make a fairly large difference in the clear passage area between the cones, for a wide range of "tuneability".

Just an idea - probably, few builders would consider it worth the effort.

L Cottrill

[Bruno Ogorelec]

the point is, is it really better than an equivalent minimum surface area intake?

Better at what? Engineering is always a compromise. You gain at one end by losing on the other. You just make sure you're gaining where it's really important for you and losing where you don't care about it as much.

Theoretically, the best shape for a combustor chamber is a sphere. Yet, as far as I know, only liquid-fuel rockets use spherical chambers. Others come in all kinds of weird shapes -- because other important considerations outweigh the theoretical advantage of the spherical shape.

Same here. Look at the short length, for instance. Or the neat profile. No penile appendages sticking out to the side, as on the 'Chinese'. That engine, if mounted out in the airstream, would gain a lot simply by offering less drag than almost any other pulsejet.

To me, the layout looks like being ideal for supercharging, because it allows you to keep the combustor out of the forced air stream and lets it take from the stream as much air as it wants. It lets the pulsejet be pulsejet, in other words.

At the same time, it lets both the intake and the exhaust add momentum to the same gas flow.

Here's a very simple sketch.

(I have let out details of the drive to the blower because they are unimportant at the moment. The discontinuity in the duct after the end of the combustor tailpipe should function as the Bertin wave trap. It is not particularly necessary on the blower side, as the blower impeller and housing decouple the pulsation pretty effectively.)

[Bruno Ogorelec]

Note that (with considerably more effort) you could make this so the clear area could be experimentally adjusted. The trick is, once you know what length range is right for the intake cap, make a new engine just the same but with the end cap wall and chamber wall as matching cones (very narrow ones, I mean) - a tiny adjustment in cap depth can then be used to make a fairly large difference in the clear passage area between the cones, for a wide range of "tuneability".

Just an idea - probably, few builders would consider it worth the effort.

L Cottrill

As ever from you, a smart idea. I like it a lot.

[Mike Everman]

Yeah, I like Larry's idea, too. I added a movable CC head as well so if you did make an adjustment to the intake, you could bring the acoustic distance back in tune...

[Mike Everman]

Then there's one where I got concerned that the tail throat region would just melt, but had a nice length reduction as a feature...

[Mark]

One thing I remember with Tharratt's engine was that if you slid the cap down or was it up, anyway as you gradually moved the occlusion/cap, thrust decreased and as you kept moving it even more in the same direction the thrust once again started to increase as shown on some graph.