SpudFiles

whats the carbon fiber thing for? looks nice

matti wrote:whats the carbon fiber thing for? looks nice

Hi Matti,

That's a replacement for the grey/pink initiator chamber. This is just a test casting as I have never made any carbon fiber before, and I need to figure out the details. Notice the corrugated steel pipe is a short scrap piece.
The mold inside the carbon fiber is 3D printed and painted with water based poster paint, actually the paint is poster paint for blacklight from another project, but is very well suited for this due to a number of reasons: It is easy to apply to the 3D printed mold, it comes off easy too after is has dried so it acts as a release agent between the epoxy resin and the 3D printed mold, it will dissolve in water and in a dark room with a blacklight torch, I can easily see if I need another layer of carbon fiber sleeve to cover the mold completely.

The epoxy resin is a high temperature variant that I need to post-cure at elevated temperatures, so I also need to get some kind of baking device for the carbon fiber part before it is suitable for sound cannon use. I'm hoping that during the post-cure baking, which will eventually reach 160°C, the 3D printed mold will become so soft that I can peel it out of the carbon fiber part.

Very exiting stuff, I get around to a lot of mini-projects to get this main project done 8)

The blacklight paint works really well for spotting holes in the carbon fiber sleeve. It might just be a matter of me practicing to apply the sleeve

I did a quick test tonight to see whether I got a too rich or too lean mix when using the high pressure pure propane. It turns out, it's too rich. I used the standard size air intake (3D printed a few copies of the brass piece from the weedburner so I had something to work with) and covered 3 of the holes with tape, and nothing, no ignition at all. Uncovered 1 hole, a small blurp. Uncovered another hole, bigger blurp. Removed all the tape from all 4 holes, a small pop. Drilled the holes up in size, 2 at a time and 0.5mm at a time, bigger pop, and when all 4 holes were 7mm (standard is 6mm), I started to get some real sound. But the bigger holes does something to the flow when the gas is burning, the sound is not as high pitched, I guess it lets out more pressure at the intake holes and this is what changes the whole combustion process.

I will have to try some different hole configurations, maybe 5-6 smaller holes or maybe a change in hole geometry or something. Thank god for 3D printers

7mm holes to the right, standard weedburner 6mm hole sizes to the left.

So I printed some new air intakes with a different geometry and they seem to be working well, and are also easy to reproduce in aluminium on the lathe and mill for the final version. The idea was to have something that was easy to adjust if needed, and I can add a ring to block part of the intake on this new design if I want to richen the mix. The 3D printed PLA works fine for testing but I can see it starts to melt so I will be making the final ones in aluminium.

I have found a place outside of town where I can go for testing without disturbing anyone with the noise. It's a private property with some very nice folks, and I can come and go as I please. So I went out testing the new air intakes with pure propane in the tank (higher pressure) and I don't think I reached the upper limit for firing rate, but some of the 3D printed parts gave up from the heat before I got there. I ended up with a firing rate of 5.33Hz for a single cannon, that's 320 shots per minute per cannon, a total of 640 shots per minute with 2 cannons! Totally awesome!

It will be raining for the rest of the day today, so I got the 3D printer working on some new parts, including a holder for the propane valve and servo, so I can open/close the propane flow from the remote controller 8)

Things will be moving along quicker now that I got a permanent place to do my testing

Check out the result of my work on the ON/OFF propane valve in my other thread here:
http://www.spudfiles.com/combustion-can ... ml#p381252

More testing has been done. Best sound is right now at around 4.7 shots/sec. If I step up the firing rate, the sound level decreases and I don't want that.
I need to limit the amount of propane coming into the nozzle, if I don't, I get a too rich mix and no real boom sound. So perhaps there is more performance to gain from designing some more stuff around the air intake.
Testing is much easier now that I can use my servo operated propane valve. This is also a good way to test the construction of the valve and servo assembly.

I also managed to 3D print the initiator chamber in one piece, approximately the same size and shape as the carbon fiber version.
The carbon fiber test I made has revealed that I cannot remove the inner 3D printed PLA mold even by heating it. I'm not going to just let it sit in there. So I ordered some PVA filament, which is 3D print material that can be completely dissolved in water. This material is like magic for 3D printing, but also a bit finicky as it will absorb moisture from the air very quickly and get soft, so I need to either store the 3D printed PVA mold in a sealed container with desiccant or print it just before I wrap the carbon fiber around it. Hopefully it can stay stiff until the epoxy resin starts to become stiff. Maybe I can put a desiccant bag inside the PVA mold while the epoxy resins hardens.

I'm also waiting for some brass bolts from ebay to arrive. I hope to make a more compact and sturdy setup for the spark plug. I think I will need to embed a small steel plate into the carbon fiber so I can cut threads in that, for mounting the spark plug bolt. I cannot have the bolt head inside the chamber as I do now, because it would be too complicated to mount the spark plug from the inside when using the carbon fiber initiator chamber. The reason for choosing brass bolts, is that I can solder them to a piece of printed circuit board which I will use to attach the high voltage wires from the HV module. Just to make it sleek and easy to fit inside the RC plane.

I'm sorry for the poor quality video, I used my smartphone and zoomed in, because I went far away for the first portion of the shots. I was something like 10-15 meters from the cannon, and at this distance, I don't need ear protection. The video doesn't do the sound level justice. When I got closer, I can hear that the microphone is saturated with sound, it just can't handle the abuse from the sound pressure. I was trying to show that there is a pretty good recoil from the cannon

Sorry, video deleted during cleanup

I tried some new air intake designs today, not much of a difference to hear.

I think I'm at the limit of how fast the initiator chamber and corrugated tube can be filled with air/propane mix using only the venturi effect from the propane nozzle stream. I think I will need to use a radial fan or compressor to blow air into the air intake to see if this will help. I can't use an axial fan, like a PC cooler fan, as these can't really create any pressure, and it's pressure I need to overcome the resistance from the system.
If it helps to force air into the air intake, and by this, I mean I can fully open the propane supply and still get good sound, I might be able to benefit from this when the system is installed in the RC plane by letting the air intake on the plane (not the air intake on the propane mixer) be a lot larger than the gun barrel, this way a higher air pressure will be present at the mixer intake, due to the speed of the RC plane.
I have some electronic pressure sensors I could mount in a dummy air intake and then measure the pressure when flying the RC plane.

Futterama wrote:I think I'm at the limit of how fast the initiator chamber and corrugated tube can be filled with air/propane mix using only the venturi effect from the propane nozzle stream.

I doubt it. Lockwood-Hiller pulse jets of the "normal" dimensions, for example, cycle at around 300Hz (dependent on size), and they're basically a tube, no moving parts, and fed only by the compression / decompression of an explosion. Other designs of valveless jets running on propane have been made that run at kHz+ rates.

Moving parts around deflagrations is likely to result in parts that no longer move as they should. I'd look into the kadency effect and acoustic phenomena.

mrfoo, thanks for your input. The pulse jets have a more unrestricted airflow than my cannon. Remember, the reason for the loud sound in my cannon is the turbulence created when the gas burns down the corrugated tube, this increases the speed of the flame. So I'm sure the turbulence will also be present when trying to push or suck fresh air into the tube, creating resistance.

I also see a big flame coming out of the air intake on my cannon, melting the 3D printed parts if I fire it to many times in a row without a break to cool down. This is also present at those pulse jets, but I still think my cannon design is too far from an "acoustic resonator" as the pulse jets are, simply due to my dimensions.

Turbulence is your friend, not your enemy. Helps with mixing. Entrainment of air by venturi and nozzle will be largely laminar flow, that doesn't get you a good fuel / air mix. For a really good example of that, try an "air amplifier" fed on propane - in theory it should give you a near stoichiometric mix of air and propane, but in reality it gives a big "rich" orange flame and a load of waste air. Mixing is probably why you're needing corrugated tubing, too. I can get a deafening bang from a drop of methanol in a length of capped 10mm tube, shaken, and a match applied to the end (you have to be careful, though, it'll take your eyebrows off if you're too close)

As for resistance - it's unlikely to be a problem. If Tesla's Valvular Conduit doesn't work on pulsejets (and it doesn't), a bit of turbulence is not going to hurt.

mrfoo, I get the feeling that you know what you are talking about and could be a great help to me. Thanks for chiming in, any help is much appreciated

As for mixing, I have been thinking about it before, and currently the mix could potentially be really bad (as in fuel/air not mixed together properly, and not the ratio) in the initiator chamber, but might be better in the corrugated tube due to the turbulence in the tube, but honestly, when firing the thing 3-5 times a second, I don't know how much of the tube is filled with fresh mix if any, maybe it's only the initiator chamber that fills with fresh fuel/air mix, I simply can't tell.

Regarding resistance, notice the red part on the exit end of the corrugated tube, this is what I call the "barrel" as it is the part that will be visible on the RC plane as a gun barrel (I will of course make the part from metal on the final version). This part reduces the ID of the corrugated tube from around 16mm to 13mm and thus must present some kind of resistance to flow. I think I also mentioned this in a previous post, that the barrel part actually increases sound volume when shortening the corrugated tube, why it does this, I'm not sure but it could be something with a higher pressure at the outlet due to the reduction in ID, but it's very welcome, as it makes everything easier to fit in the RC plane when I can shorten the corrugated tube.

If you look at the attached picture, the fuel-stream pulls in air in the orange "mixer" as I call it. In the connecting chrome color tube, the only thing that could promote mixing in that tube is the step in diameter from the orange mixer ID of 8.6mm to the ID of the chrome tube of 13mm (same as barrel ID). And maybe there will be some mixing when the flow enters the initiator chamber as the ID then increases again. There is no "obstacles" to promote mixing. Maybe I could get better performance by making something to help mix the fuel and air.

What I did try at a very early stage, after finding that the weed burner geometry works well, was to shorten the chrome tube, to use less space, but that wouldn't work, apparently there was then too little mixing to even fire the thing. I have also wondered if a longer tube would be better, but I haven't tried it, as I cut the weed burner tube right before it starts to bend (and I will have a hard time sealing a bend pipe onto the initiator chamber).

I have also tried to redo the mixer, so instead of threading it into the chrome tube (which is how it is done on the weed burner) I made a mixer that mounted on the outside of the chrome tube, but that wouldn't work either, and I was thinking it was because that step in diameter I mentioned was missing, not because it was the only thing promoting mixing, but I was thinking it had something to do with some back pressure when firing, as if the step was missing, then too much of the pressure (upon firing) from the initiator chamber would escape through the air inlet (the flames from the inlet I also mentioned) and this would prevent a good flame acceleration in the corrugated tube. But maybe I was just missing a good mixing.

When printing the new mixers, I tried connecting one to my air compressor, just to get a feel of how much air it actually outputs. If I block the air intakes on the part, I can hardly feel any air coming out, even when putting the output to my face (more sensitive around the lips and nose). This must be due to the fact that the only air escaping the output, is what is coming out of the 0.3mm nozzle, but this is expanded to the 8.6mm ID of the "mixer", making the air speed very low. When unblocking the air intakes, the venturi effect really kicks in and I can easily feel the air stream with my hand, even at a distance.

I have an anemometer/wind speed meter, just a cheap one from ebay, and I was thinking I could make an adapter to fit between the anemometer and the 13mm tube, and then measure the air flow speed of just the mixer, and then the barrel, and compare these to see if there is much of a difference - I think this could be a way of checking whether there is any significant resistance in the system, just to convince myself of what you are saying

So how can I proceed with testing?

I have some ideas:
- Force air into the intake by powerful radial fan or air compressor, and see if that will allow me to fully open the propane flow (maybe not directly related to the mixing but to any resistance).
- Make a longer chrome tube, see if this improves anything.
- Add mixing promoters in the chrome tube, maybe a coil at the wall like the ones Shyasuna uses.
- Exchange the chrome tube for some corrugated tube, this should improve mixing, see if it improves performance.
- Test the "step-less" mixer again, but this time add another mixing promoter instead of the ID step, any of the above solutions if they turn out to work.

I'm in no hurry to get the system installed in my RC plane, winter is coming (as the Starks would say) here in my country, so I would rather have a very reliable system that is well tested and optimized than rush it into my RC plane. Also, I love all this testing stuff, it's very educational 8)

Another update: I'm getting help from my local RC plane forum with the carbon fiber stuff for the initiator chamber, and I have ordered some braided glass fiber sleeve and carbon fiber rovings, and the plan is to have 2 layers of glass fiber sleeve in between 2 layers of carbon fiber sleeve. This way, the initiator chamber will be stiff from the carbon fiber and very tough from the glass fiber. The rovings should be used to wrap around the braided sleeves to keep them close to the shape of the mold and to get the sleeves to form around the corrugated tube, or tubes, if the intake tube ends up being corrugated too

Very exiting stuff, this project

EDIT: I just thought of something. All Shyasuna's tests were made with a gas burner with piezo ignition, so the ignition point is close to the burner outlet. Since I started using the HV modules to generate the sparks, I have just placed the spark point in the middle of the initiator chamber. Would it be worth trying a different location, or maybe multiple locations?

You know what I'd try?

Straight tube, say 12mm id, with a reed valve at one end, a metered, intermittent, injection of propane or methanol, and a spark source. A very small, intermittent, valved pulse jet. It'll (hopefully) be too small to sustain, and deliver shit all thrust, but you're after single shots and noise, right? Start with your "designed" length, if you need more, you can bend it like a trumpet.

I'd suggest something super simple, something like the "craft jet" (see http://www.pulse-jets.com/download/craft.zip) but with no expanded combustion chamber, effectively take a length of aluminium rod the same OD as your tube, turn down say half of to the ID of your tube, (can be done easily on the poor man's lathe), drill centrally and flare the hole at the outside (fat) end, cut the turned down end at 45° or so and make it super smooth, take a piece of thin spring steel to fit over the 45° end, attach with one small machine screw, sort out fuelling and spark in the tube itself and you're done.

I'd strongly suggest using methanol as fuel. Much less fussy than propane.

also - http://www.pulse-jets.com/phpbb3/viewto ... ?f=2&t=995

mrfoo, I did try a straight tube, though without a valve, and that did not give me any loud boom sound.

I don't think the pulse jet approach will work, but that forum you linked to, has a lot of information on Pulse detonation engines, and maybe there is help for me to get in that forum, I could try to make a thread in there and see if I get any response.

pulse-jets.com is pretty dead these days, there's occasional traffic and the old timers are still lurking. An interesting question will almost certainly get some cogent answers. Eventually...

I'd avoid the PDE stuff, though. You're not going to get detonation at all (and you don't want to) unless you use acetylene and oxygen (and you don't want to do that either).