The Voice of Fo Mi Chin Is Heard In the Land

[larry cottrill]

Yes, Fo Mi Chin II really did run, as mentioned by Bruno in the pulsejet meet thread on the Off Topics Forum. Starting was apparently kind of a bear, though I won't really understand why until I try it myself.

Here are some shots of Bruno and Fo Mi Chin II in action. Or at least, in position ...

Mike sent these to me, but I'm not sure who shot them.

L Cottrill

[Moderators: If you think this should have been in Off Topics with the Pulsejet Meet related threads, go ahead and move it.]

[larry cottrill]

OK, after I got home today, I finally got to quickly set up and try my Fo Mi Chin II that Bruno originally tested at Burning Grape! Yes, the engine runs very well, though it does not have a great range of fuel input that it will accept, and probably far more limited power than my FWEs (certainly less than the Smooth Lady that I just got running a couple of days ago). I think the limited range of fuel flow acceptance is due to the somewhat undersized (and perhaps inefficiently shaped) intake area.

The heat pattern on the chamber walls is not particularly smooth. There is a hot zone at the bottom and sides forward of and beneath the intake 'spout' (where it turns down inside the chamber). Heating of the front dome is more moderate. There is pretty good heating of the nozzle zone and the front end of the tailpipe.

The engine is smooth running, and starting is an absolute snap. I had none of the starting problems Bruno mentioned, at all. I have been able to start it with as low as 25 PSIG air on the starting tube, but 30 PSIG is definitely better. I can get sustaining operation within two seconds of beginning to open up the fuel needle valve. Once the shell was well heated, I was able to open the needle valve approximately two full turns before 'rich extinction'. When starting, there definitely IS a region of low fuel flow that results in rapid intermittent pulsation (as I observed in the Smooth Lady), but this is passed through so quickly as you open up the valve that it almost goes unnoticed. It is easy to overdo the fuel and kill it right after startup, until you let the chamber get some heat; then the full range is obtainable. Fuel was delivered at 12 PSIG regulated pressure.

Thanks again, Bruno, for your enthusiasm and effort in the initial testing, and to the others who helped out. I can think of no good explanation for your starting troubles. Maybe, high humidity? Too much starting air for that little intake? Hard to say, I think ...

I'll throw up some pictures here when I have them, but it will take a while.

L Cottrill

[Dave]

Larry

I thought that first picture looked familiar so I went back through my digitals from the Burning Grape to find that it was mine.

I think you will also find part of our starting difficulties are clearly displayed in the first photo. That is not exactly a needle valve you see in line with your engine. Getting small engines to run using the tank valve involves a lot of guess work and somewhat of a magical touch.

The second picture shows a bit of team work and a good regulator. That is when she really took off! Note the color of the combustion chamber.

Dave

[larry cottrill]

Dave -

More nice shots! Watch for my last ever Fo Mi Chin thread, coming soon.

Thanks!

L Cottrill

[hagent]

Hi Larry,

Can UFlow be used to model your Fo Mi Chin ? Also I'm getting close to finishing my annular? PJ The intake surrounds the exhaust, and I was wondering if UFLow can be used in this type of design because one end is totaly blocked off with two ends open on the same side.

Thanks,

[larry cottrill]

Can UFlow be used to model your Fo Mi Chin? Also I'm getting close to finishing my annular? PJ The intake surrounds the exhaust, and I was wondering if UFLow can be used in this type of design because one end is totaly blocked off with two ends open on the same side.

You can model rear-breathing engines only very roughly, by pretending the intake doesn't exist. Obviously, pretty crude, though I have found it useful. It just gives you an idea of the resonance characteristics of the basic tube, which does behave as a half-wave oscillator when viewed in isolation. You can't get any information on how such an engine will actually breathe through the intake.

The reason accurate modeling of such geometries is impossible with UFLOW1D is the '1D' part - if you have an intake that isn't in line with the rest of the tube, the problem is not describable in one-dimensional terms. Even the Reynst Pot can't be correctly modeled, because UFLOW1D has no provision for an intermediate gap communicating with the outside world.

L Cottrill

[hagent]

Thanks Larry,

So the CC end would just be left open? I guess there is no other choice.

[larry cottrill]

Thanks Larry,

So the CC end would just be left open? I guess there is no other choice.

Sorry I didn't mention anything about that. You make the CC outer end a tiny hole (like 1 mm diam) with a very flat cone forming the front end of the chamber. When you run UFLOW1D, there will be a little anomaly in the curves there (at the left end where the tiny hole is) which you can ignore. I refer to this as the 'spark plug leakage' ;-) The very flat cone will essentially reflect just like a flat end plate. Of course, I always prefer a slightly domed end anyway, so I simulate that with an appropriately sized cone and tiny hole.

If you just left the front end open, you would be analyzing a half-wave pipe, not a quarter-wave one, which would give you completely incorrect results.

L Cottrill

[Bruno Ogorelec]

You make the CC outer end a tiny hole (like 1 mm diam) with a very flat cone forming the front end of the chamber. When you run UFLOW1D, there will be a little anomaly in the curves there (at the left end where the tiny hole is) which you can ignore. I refer to this as the 'spark plug leakage' ;-) The very flat cone will essentially reflect just like a flat end plate. Of course, I always prefer a slightly domed end anyway, so I simulate that with an appropriately sized cone and tiny hole.

If you just left the front end open, you would be analyzing a half-wave pipe, not a quarter-wave one, which would give you completely incorrect results.

Larry, why can't you turn around the process next and pretend that the tailpipe was of minuscule size and analyze the events in the intake, in the same way you did with the tailpipe in the first step.

Maybe you would arrive at two sets of results that might usefully fit together.

[larry cottrill]

Larry, why can't you turn around the process next and pretend that the tailpipe was of minuscule size and analyze the events in the intake, in the same way you did with the tailpipe in the first step.

Maybe you would arrive at two sets of results that might usefully fit together.

I'm sure you can do that. I'm sure you cannot do it perfectly. One problem I have with analysis of the intake, though, is deciding what gas and pipe temperatures to use. Both must be fairly low compared to the tailpipe, in the case of my engines, anyway. Another problem is that my intakes always come in at odd angles - that's a lot harder to simulate than an inline pipe, since UFLOW1D assumes symmetry around an axis. In the classic FWE, how far is it from the inner end of the pipe to the "bottom" of the chamber? What does that pipe see as the "bottom" of the chamber? We would have to assume the front dome, I guess, and calculate the position of the pipe according to wave path length. What else could we do? Doing this, we would need a "rear plate" on the chamber at the inner end of the pipe, which means the chamber modeled would be much shorter than reality - would this be OK? I'm not sure ... there would be a lot more internal reflection than in the real chamber, but maybe that could be easily sorted out.

Yes, I need to look at this more closely and see how some variations work. I certainly have enough real-life examples now from running engines, as far as the dimensions are concerned. The Smooth Lady seems like a simple case to model - practically a straight-in Thermojet clone, as far as the intake is concerned.

When I have tried crude analysis of the intake before, it was always by artificially swinging it around to the front (using the 'Reynst Point' as a fulcrum), making a 'linear' engine with a somewhat shortened chamber. While this seemed to show reasonable results with carefully tweaked temperatures, it is not an ideal solution, because it changes the operation of the pipe as a whole. It has been suggested that lengthening the chamber by swinging the intake around the front plate (pressure antinode) would be the right way, since this makes sense acoustically - however, that alters the total gas mass considerably.

It's at least possible that there simply is no highly accurate way to model such designs with 1-dimensional analysis methods. But of course, it's also possible that some method (or combination) will be good enough!

L Cottrill