New and Improved Lady Anne

[Eric]

Hmmm. Try tacking a 3.5" long straight section to the end of the tailpipe if you get time.

Looks like another very nice build.

Eric

[mk]

The build looks good, however, with the inlet that near to the head plate, it is very likely that the engines will still refuse to run. Regardless of other changes.

[larry cottrill]

James -

Yes, it is a nice build and it looks like the layout is exactly what Graham and I predicted from NUDiS and UFLOW1D respectively. My only criticism is that the intake flare appears too "abrupt", i.e. not well rounded from tube to rim. Eric has commented before on the importance of a good flare, though I think he was talking about performance, not whether the engine would run or not. Be aware, though, that an advanced design will be much less forgiving of small details than the earlier "straight tail" FWE engines.

I think the easiest thing to do now would be to use a "coffee can slider" - just a small roll of thin sheet metal or a split tin can - around the tail end rim (cinch it with a large hose clamp). Anything would work as long as it wraps around with a little overlap. This should be used to lengthen the engine a little at a time - maybe a half cm or so - until it wants to run (or you just give up in disgust).

If you can get it to run in this way, you should be able to remove the sleeve and get the engine running by SHORTENING the intake by a small fraction of that added length - one third at most. You would want to work the intake back a couple of mm at a time until you get good starting; then and only then, re-form a good flare for optimum performance. Again, this is assuming she will run without an intake flare being there at all.

If the tailpipe sleeve doesn't work in only a few cm of extension, remove it and cut off the intake flare, then use a smaller snug-fitting sleeve to gradually lengthen the intake. Again, your adjustment would be just 3 or 4 mm each time until you get it running - much more delicate tuning than at the tail end. Once you have the running length established, weld on a permanent extension with a decent flare. This is tuning in the OPPOSITE direction from lengthening the tail, of course, and several times as sensitive!

One tip pertaining to this kind of testing: I have found that an engine that is a little too long (meaning the main body) will start to sustain, but then fade out and die after it warms up for a moment; a too-short engine won't act like this. Of course, this was experienced with straight-tailpipe engines, but I don't know why it wouldn't apply in this case as well. Just something to watch for as you work the length outward. I found that a straight-tailpipe FWE that acted like this could have an inch or more cut back to get a running engine - the Lady Anne will be much more sensitive than this, I'm sure.

L Cottrill

[Graham C. Williams]

Missed this one.
Nice build again, James.
I'm sure the slightly longer induction and or exhaust pipe will got things moving.
I'm not so sure about the location of the induction pipe. That region between the linear motor and the Chinese is all uncharted territory.
It may be that in the end you'll have to move to a Chinese location.

Regards
Graham.

[Eric]

From my guestimate / calculation 3.5" should add just enough to the tail for the intake location to work well... if its built to the same size as the other lady anne's.

The intake flare would help things along, a proper flare not only improves performance but also will increase throttle range and ease of starting.

Just for the hell of it check your spark plug and make sure the gap is a decent size and maybe clean the terminal, if the propane coated it with carbon you might not be getting a good ignition spark.

Also try rapidly pulsing the compressed air, with pulses you will hit a good fuel air ratio more often than constant blowing.

Something like an old air chisel might be able to be taken apart and use the exhaust air from the chisel which will be pulsed pretty rapidly, or you could maybe make a setup to pulse the air yourself.

If you can pulse the air at 10-20 times a second you can get almost anything to run as a pulsejet or at least run on compressed air, including straight tubes.

Eric

[Graham C. Williams]

Dear Eric.
Love the pulsed compressed air thing. I can do that with NUDiS. I can run a pipe with no heat addition and just pressure rise. Interesting thing, rapid temperature rise alone in the combustion chamber is not sufficient to get things running (but pressure rise is). I think I know the reason for this, it's a 2D mixing thing. The Belfast papers allude to this but it's a bit oblique.

Graham.

[larry cottrill]

Dear Larry.
If you get the NUDiS programme up and running (See Tools) send me an email. I'll send you the model and full combustion run for this motor.

Regards
Graham.

Graham -

I don't know how to handle the RAR files. Yes, it was explained to me a few years ago, but I've never done it and am not sure what tool(s) it takes to extract from them. If I can figure that out, and after completing a few other 'projects', I'll let you know.

L Cottrill

[larry cottrill]

The build looks good, however, with the inlet that near to the head plate, it is very likely that the engines will still refuse to run. Regardless of other changes.

Marten -

But, we've seen at least one example of an engine with the intake right "up against the wall", just a few months ago. I think it was a pretty long engine, though, so Eric may have something in his comments.

I'm still betting on small changes bringing it in.

(Of course, I have been occasionally wrong before ... ;-)

L Cottrill

[larry cottrill]

Love the pulsed compressed air thing. I can do that with NUDiS. I can run a pipe with no heat addition and just pressure rise.

Yes, this is really good starting technique, especially for an unproven engine where you're not sure how much air is "just enough"!

This is greatly simplified with my "starting air tubes" - the best pulsejet idea I ever had, and simplicity itself, but NOBODY else ever tries to use it. It has worked for me in various forms on a variety of my engines. It gives really good starts once you experiment a little with each engine to find the right air pressure setting. Mostly, it frees you of having to "aim" the air input yourself, so the pressure is the only variable. The only time it ever failed for me was when I was trying to impress Ben and half a dozen of his friends.

Interesting thing, rapid temperature rise alone in the combustion chamber is not sufficient to get things running (but pressure rise is). I think I know the reason for this, it's a 2D mixing thing. The Belfast papers allude to this but it's a bit oblique.

That's absolutely amazing - I have suspected exactly the same thing. As the engine re-pressurizes via Kadenacy breathing, everything is in place to reignite the new charge of fuel-air mix - but it doesn't happen until the pressure reaches a certain point. Perfectly timed ignition, with no electro-mechanical assistance required. And what you're saying here is that (at some level) it depends on an effect that 1D simulation is inadequate to describe. This might have a lot to do with the effectiveness of the flat end plate (SHARP intake-to-chamber transitions) in the Kentfield design, for example - a near-ideal convergence of re-pressurization and vortex-induced turbulence.

L Cottrill

[Mark]

Although not very popular, the Logan pulsejet is stunningly quick to start, zero to full grease on methanol by just lighting the tail. At least the two sizes I have crudely put together rev up almost before you think you could have possibly lit it, like a firecracker with no fuse. It toys with your mind.
Mark

[Graham C. Williams]

Dear Mark.
Can you sort out a free rar unpacker for Larry?

Graham.

[Graham C. Williams]

I found this one. Don't know if it's any good.

http://www.majorgeeks.com/download.php?det=4696


Graham.

[larry cottrill]

James, Graham et al -

Interesting thing, rapid temperature rise alone in the combustion chamber is not sufficient to get things running (but pressure rise is). I think I know the reason for this, it's a 2D mixing thing. The Belfast papers allude to this but it's a bit oblique.

Well, I just can't get away from this. Here's a hypothesis as to why this configuration isn't working as well as predicted. See what you think:

First of all, I believe there is nothing wrong with the wave mechanics of this geometry. Secondly, there would be nothing wrong with the flow mechanics IF the pipe behaved like the NUDiS and UFLOW1D simulations. However, these are one-dimensional, and leave out a lot of small-scale multi-dimensional behaviour that might be important. Here's what I think could be happening:

The power of the FWE pipe depends on getting good combustion somewhere in the vicinity of the front dome - the rest of the chamber is basically a nozzle to take advantage of ongoing gas expansion. As Eric has pointed out, if the combustion region becomes suboptimally located, things aren't going to work well - basically, a lot of the eventually realized gas expansion will be wasted effort. Conceivably, if the combusting charge were located badly enough, "growling mode" under forced air might be all we could realize.

What if the entrance of air into the chamber is actually too smooth and unimpeded? Consider the scenario at that moment: (a)The pressure and density in the chamber is low; (b) The velocity of incoming air is fairly high; (c) There is no sharp break (a la Kentfield) in the incoming flow path, except at the rearmost edge region that is common to the intake duct and the chamber.

What we have is now almost exactly what Eric shows in his earliest drawing on this thread: The air blasts in, flows across the dome and is funneled rearward, still at fairly high speed. Possibly MOST of the fuel air charge is actually on its way out through the nozzle zone before the reflected pressure wave even makes it into the chamber! We've basically violated one of the important flow principles for pulsejets: We haven't braked the incoming air to a near standstill.

If you agree that this seems like a realistic scenario, what we need is something like this: A "bluff body" (as suggested to me by Graham, years ago) - say a fair-sized flat washer welded into the chamber corner so it projects down into the incoming flow stream as it turns into the dome! The plane of the washer should point toward the mid-front chamber region, so it will act as a more-or-less streamlined 'vane' for the blast gases routed out through the intake. See revised drawing below.

Such a device will not be destroyed by the combustion gases, because of effective cooling via the incoming air blast. It will be denounced as a hokey gimmick, of course, but might be exactly what's needed to make the design workable as dimensioned, which would allow the basic design geometry to be proven out.

L Cottrill

[Eric]

After seeing the drawings again I think a dome spark plug would also be less than perfect for starting. The spark plug position can have a pretty dramatic effect on starting, especially if you have a high spark rate that will ignite the fuel instantly.

If you blow fuel and air in and have it immediately ignite in the dome before building up a sufficient charge the engine wont be able to "sync up" and take off very easily.

I dont think the flow is too smooth, the flow will splash outward, and in the process there will be a lot of turbulent mixing, and combustion will not be terribly different from a linear engine.

If we had a video clip of a starting attempt diagnosing the problem would be a lot easier.

Eric

[James D]

Thanks for the comments everyone,

The intake flare is probably a tad sharp, I will make an extended aluminium flare that slides onto the intake pipe, so I can try adjusting the length.
I should be able to test it again, and record some video, in a couple of weeks. if it still doesn't work I'll try moving the spark plug.

If all else fails I can always just cut the intake off and put it on the front of the combustion chamber;)