For a Close Shave, You Need a Fo Mi Chin

[larry cottrill]

Monday night I welded up the bent-back intake tube and made the tailpipe with its oval slot to accept the intake. Tonight, I got the whole mess welded up. From arm's length, it's actually fairly pretty -- much nicer proportions than my crude preliminary drawing shows.

What I like particularly is the close fit of the intake lying along the tailpipe -- about 5mm gap between the tube walls! And, a nice steel flare formed with a tubing flaring tool. The intake on this one is just 1/2-inch conduit, and also a little shorter in total length than on the Elektra I. The end that penetrates the tailpipe is flattened to a shade less than .5 inch outside width, to streamline it to the exhaust flow around it. And, as Mike Everman suggested, I limited its penetration into the tailpipe cross section to about the halfway point, so it doesn't form too much of a 'pinch' for the exhaust flow rearward.

Up close, it's a different story. The welding of the pipe to the domed nozzle section of the can was some of the most difficult and frustrating I've ever done -- over 30 minutes to fully weld around that joint! Of course, this time included burning and filling in about 7 or 8 little holes in the paper-thin steel dome. The weld is probably basically sound, but is rough and cratered in spots, and the thickness of reinforcement probably varies all over the map. The fillet weld between the intake and tailpipe is much nicer, from having thicker material to work with. That was done in probably ten minutes of less.

The intake lies so snug against the pipe that you just barely see it above the chamber wall when looking dead on at the front end. So, it's really a pretty clean looking machine. Also, there are no engine mounts or frame to clutter it up, yet.

Overall length without spark plug, 26 inches. The tailpipe is 1.25 inch OD, the chamber 2.5 inch OD, just like the Dynajet. Straight side length of the chamber, about 5 inches. Nozzle dome length, just under 1 inch. Total weight right now [spark plug in place, but no fuel pipe, no mounts] is right at 16 ounces, again exactly matching the Dynajet.

Photos next week!

L Cottrill

[larry cottrill]

Well, this afternoon I tried to run it, using my variation on the Steve Bukowski mount. My first problem was that the buzzer on my Model T coil seized up. Of course, it took a lot of wasted propane before I realized I wasn't getting any spark to the plug. Fortunately, maintenance of Model T coils is simple, and I quickly got it going once I recognized the problem. Then, I got a few bangs and it quit firing again. The coil was working fine now, so I took out the plug and found a large particle shorting it out! So I carefully wire brushed that out, and starting trying to test in earnest.

What I get is basically similar to what I got on the Elektra I when I didn't have any intake flare, except this seems a lot closer to going -- I can get a good buzz started, and draw the air source back away from the intake and it gradually weakens until it just fades away at quite a distance out [a couple of inches with the shop vac, maybe 8 to 10 inches with the leaf blower].

One thing that is NOT like the Elektra is that in this case, frequency definitely does depend on the engine length. I think the difference is the impedance matches of the two machines: Fo Mi Chin is trying to support a 1.25 inch OD tailpipe with about a .6 inch ID intake, which is then flattened significantly where it pours out into the chamber via the tailpipe. Could this be too small a straw to breathe through? On the other hand, the flare I was able to form in the end of the intake tube is pretty modest compared to the double-diameter funnel on the intake of the Elektra; maybe it would respond to being built up a little. The fuel pipe arrangement is virtually identical, except that in Fo Mi Chin it is purely temporary -- the fuel tube just rests in the air tube by gravity. It seems perfectly stable, however, and seems to deliver plenty of gas.

I can always get a perfectly smooth, roaring buzz, somehow. When I tried extending the pipe several inches, it was harder to get it to roar, not easier, presumably because of the increased tail piston mass. I decided that since I was going to play with extensions anyway, I might as well try shortening the pipe a little. Shortening the engine to the length of the Dynajet [pipe plus head] raised the frequency slightly [less than expected!] but did not make it any harder to obtain a good buzz, and possibly slightly easier. So, I don't get any sense of where we are acoustically -- the whole problem seems to be dealing with the gas mass.

Perhaps I'm too impressed with the beneficial effect of the oversize flare on the Elektra, but that did make a huge difference, and I'm thinking that's the simplest thing to try next.

Any suggestions or comments welcome.

L Cottrill

[Mark]

Well, this afternoon I tried to run it, using my variation on the Steve Bukowski mount. My first problem was that the buzzer on my Model T coil seized up. Of course, it took a lot of wasted propane before I realized I wasn't getting any spark to the plug. Fortunately, maintenance of Model T coils is simple, and I quickly got it going once I recognized the problem. Then, I got a few bangs and it quit firing again. The coil was working fine now, so I took out the plug and found a large particle shorting it out! So I carefully wire brushed that out, and starting trying to test in earnest.

Any suggestions or comments welcome.

L Cottrill

Did you sand the points on your buzz coil? And it is funny how those spark plugs can do weird things like the spark crawl along the surface instead of jumping or they become corroded and weak, they are sneaky. In a way I guess the points and the spark plug really do the same thing, they both have a spark jumping between points. Aren't the points made of tungsten or something resistant like that? I met a fellow that had some solid heavy points used in some older large aircraft for something, whatever they were for, I thought it neat to go all out and use enormous pure platinum disks for points. I can't remember how much they weighed but close to a half ounce or something. Car points are tiny bits of course.
Mark

[larry cottrill]

Did you sand the points on your buzz coil? And it is funny how those spark plugs can do weird things like the spark crawl along the surface instead of jumping or they become corroded and weak, they are sneaky. In a way I guess the points and the spark plug really do the same thing, they both have a spark jumping between points. Aren't the points made of tungsten or something resistant like that? I met a fellow that had some solid heavy points used in some older large aircraft for something, whatever they were for, I thought it neat to go all out and use enormous pure platinum disks for points. I can't remember how much they weighed but close to a half ounce or something. Car points are tiny bits of course.
Mark

Mark -

Actually, you should never sand the points -- you should use a point file, available at auto parts stores everywhere ;-) The points and the plug "do the same thing" only in the sense that they have a spark cross them; the problem with the points is that what you get is a fairly high current spark that can cause a lot of heating and damage. Any time I've seen damaged points, it was always that a chunk had broken out of one side and welded onto the other, looking like a little mountain peak sticking up in the middle of an otherwise smooth plateau. It means you don't ever get a clean separation of the points [where they suddenly quit sustaining a spark] but rather, the spark is drawn out over extra time. This greatly weakens the collapse of the electromagnetic field captured in the primary coil, and thereby weakens the spark that is generated a bit later at the plug.

What my problem turned out to be is that the moving point armature was no longer springing back away from the stationary point -- I just had to disassemble it, bend it back a little to a better resting orientation, and put it back together and readjust. It worked fine after that.

With a Model T coil, you need periodic maintenance like that every ten years or so ;-)

L Cottrill

[Mark]

Yes, my buzz coil hasn't needed tweaking in the 5 years I've had it. I have another new one just in case anything goes wrong. I love them. I thought it might be interesting to put the apparatus, the points and tiny magnet, spring arm, atop a car coil with a capacitor. Would be the same really. I've felt the magnetic effect when I used some threaded rod and run a contact over it skittering along making and breaking a car coil field, sometime it would try to stick to the all thread magnetically. I was just toying with a way to use a car coil, condenser, and battery alone to run a spark plug and making and breaking the circuit, much like a playing card in the spokes of a bicycle tire, only using all thread instead.
Mark

[hinote]

What I get is basically similar to what I got on the Elektra I when I didn't have any intake flare, except this seems a lot closer to going -- I can get a good buzz started, and draw the air source back away from the intake and it gradually weakens until it just fades away at quite a distance out [a couple of inches with the shop vac, maybe 8 to 10 inches with the leaf blower].

L Cottrill

Larry, I'm going to try and give advice about an engine that I don't (yet) fully understand:

I've done some basic research on this engine type (Chinese/Thermojet valveless); I believe it obeys rules somewhat similar to the rest of the operating valveless engines--a combination of resonant (acoustic) and gas-dynamic effects. The real trick appears to be in the intake tube, and how to relate its length and location to the rest of the engine.

If you look at the "standard" Chinese dimensions (currently being re-posted on the thread, "Nord 1500 Project"), you'll find the distance from the front of the engine to the mouth of the intake is very close to 1/3 the length of the engine (all calcs should be made to include open-end correction factor of .6R); this figure is reinforced by the Thermojet drawing that has appeared from time to time on this Forum.

The length of the intake tube itself is subject to some variation, but (again using the "standard" Chinese as a benchmark) the measurement appears to be L/5 (use an end-correction factor for both ends on this one). This figure is at odds with the L/4 spec used on the Thermojet drawing, but I would tend to "copy the Chinese" because of its greater success.

I couldn't find any specs on your drawings to make comparisons with, but my "eyeball-micrometer" says your intake tube is located way too far back on the engine. The intake resonance has to coordinate with the overall engine resonance, just as in the L-H family, so this is critical to get correct.

BTW if you get it to run (sustain) it will probably be somewhat weak because it won't get a magnified primary signal from a tailpipe megaphone. Again, refer to the classic Chinese for an example.

Please take these comments as a possible useful observation, NOT the opinion of an expert (which unfortunately I'm not). Does anyone else have comments that may reinforce my observations?

Bill H.
Acoustic Propulsion Concepts

[larry cottrill]

Larry, I'm going to try and give advice about an engine that I don't (yet) fully understand:

That's OK, Bill, I don't fully understand it, either.

I've done some basic research on this engine type (Chinese/Thermojet valveless); I believe it obeys rules somewhat similar to the rest of the operating valveless engines--a combination of resonant (acoustic) and gas-dynamic effects. The real trick appears to be in the intake tube, and how to relate its length and location to the rest of the engine.

If you look at the "standard" Chinese dimensions (currently being re-posted on the thread, "Nord 1500 Project"), you'll find the distance from the front of the engine to the mouth of the intake is very close to 1/3 the length of the engine (all calcs should be made to include open-end correction factor of .6R); this figure is reinforced by the Thermojet drawing that has appeared from time to time on this Forum.

Well, in approximate terms, that's about where it is on the finished engine -- erring somewhat closer to the center point, I'm sure.

The length of the intake tube itself is subject to some variation, but (again using the "standard" Chinese as a benchmark) the measurement appears to be L/5 (use an end-correction factor for both ends on this one). This figure is at odds with the L/4 spec used on the Thermojet drawing, but I would tend to "copy the Chinese" because of its greater success.

Again, I don't know how close you'd have to 'call it', but that's not a lot different from what I actually ended up with. See photos in my 'Shape of Things ...' thread.

I couldn't find any specs on your drawings to make comparisons with, but my "eyeball-micrometer" says your intake tube is located way too far back on the engine. The intake resonance has to coordinate with the overall engine resonance, just as in the L-H family, so this is critical to get correct.

I guess I could bite the bullet and make some actual measurements and do a dimensioned drawing, so we'd all have a better idea of what we're really talking about. Maybe next week ...

BTW if you get it to run (sustain) it will probably be somewhat weak because it won't get a magnified primary signal from a tailpipe megaphone. Again, refer to the classic Chinese for an example.

I'm sure you're right on that. Cone building just isn't something I've had any compelling urge to do, so I vote for simplicity rather than optimal performance. However, the difficulties of actually building this engine are sufficiently high that rolling and welding a long cone out of some galvanized sheet or something probably wouldn't greatly increase the burden.

Please take these comments as a possible useful observation, NOT the opinion of an expert (which unfortunately I'm not). Does anyone else have comments that may reinforce my observations?

Bill H.
Acoustic Propulsion Concepts

Sure enough -- your opinions are probably sounder than most. The only comment I can make in rebuttal is to reiterate that it takes very little high speed air to make this thing buzz really well, over a variety of lengths [via simple extensions]. It comes so close to sustaining that I really can hardly imagine that it's very far off in basic geometry. It is true, of course, that the intake air path is pretty long; the other weirdness is having the intake partially 'blocking' the tailpipe cross section right at the most critical area.

If I get around to doing the Elektra II [where the octagon box is set up in a 'diamond' rather than a 'square' orientation], what I intend to do is set the outlet of the intake pipe as close as possible to the tailpipe entrance without actually being within it. That could prove to be better geometry for that critical region of the engine.

Thanks for the 'golden ratios' you have revealed here! I'll do some simple measurements, and let you know how close to these I am right now.

What about the impedance of having such a small intake pipe [pinched as well!]? You didn't mention whether you had any feelings about the limited cross section and volume of such a small tube. Again, study the photos to get a feel for it. The tube is 1/2-inch EMT [conduit] -- the actual ID is pretty close to 0.6 inch. The pics will give you a good idea of the intake flare I was able to produce; what's shown is about the maximum diameter I can get with the tool currently available.

L Cottrill

[larry cottrill]

If you look at the "standard" Chinese dimensions (currently being re-posted on the thread, "Nord 1500 Project"), you'll find the distance from the front of the engine to the mouth of the intake is very close to 1/3 the length of the engine (all calcs should be made to include open-end correction factor of .6R); this figure is reinforced by the Thermojet drawing that has appeared from time to time on this Forum.

Well, in approximate terms, that's about where it is on the finished engine -- erring somewhat closer to the center point, I'm sure.

Well, I wasn't real close on this one. I measure the distance from the apex of the front 'plate' (dome, really) to the intake mouth right at 10.5 inches. Using the original 26 inch overall length, that puts it at
10.5 / 26.0 = 0.403L - call this point datum 1
or at about the 40% point. How bad is that, do you think?

The length of the intake tube itself is subject to some variation, but (again using the "standard" Chinese as a benchmark) the measurement appears to be L/5 (use an end-correction factor for both ends on this one). This figure is at odds with the L/4 spec used on the Thermojet drawing, but I would tend to "copy the Chinese" because of its greater success.

Again, I don't know how close you'd have to 'call it', but that's not a lot different from what I actually ended up with. See photos in my 'Shape of Things ...' thread.

The straight section running parallel to the taipipe centerline is 3.75 inches from the outermost weld seam to the edge if the flare. It is no longer possible to measure what goes down into the pipe, but a reaonable estimate is 1.75 inches, for a total of 5.5 inches. However, this is the maximum possible length to consider, not the median length on the CL of the pipe. We can probably deduct about 1 ID to correct. The ID of the tube turns out to be 0.65 inch, so 5.5 - .65 = 4.85 inch path length in the tube. I assume that the .6 ID correction gets added to each end (right?) so the effective length would be 4.85 + .39 +.39 = 5.63 inches. Now,
5.63 / 26.0 = 0.217L or about 22% - call this point datum 2
That one's pretty close.

I couldn't find any specs on your drawings to make comparisons with, but my "eyeball-micrometer" says your intake tube is located way too far back on the engine. The intake resonance has to coordinate with the overall engine resonance, just as in the L-H family, so this is critical to get correct.

I guess I could bite the bullet and make some actual measurements and do a dimensioned drawing, so we'd all have a better idea of what we're really talking about. Maybe next week ...

Again, it is impossible to accurately measure exactly where the intake starts, but it can be deduced as follows: The apex of the front dome to the starting edge of the tailpipe is about 5.8 inches. The 'prow' of the intake flattened zone is about .25 inch short of that, at about 5.55, which is at the station
5.55 / 26 = 0.213L or 21%
but, what we really want is probably more like the centerline position. I measured the longitudinal axis [meaning parallel to the engine centerline] of the flattened tube at approx. .9 inch; adding half that to the above gives us 5.55 + .45 = 6.0 inch, so the station is
6.0 / 26 = 0.230L which is the 23% station - call this one datum 3.

Now, we could see what a couple of simple changes would do to our three data points:

(a) Lengthen the tailpipe 6 inches, leaving everything else alone:
Datum 1 (intake mouth station): 10.5 / 32 = 0.328 = 33%
Datum 2 (intake tube length): 5.63 / 32 = 0.176 = 18%
Datum 3 (intake port station): 6.0 / 32 = 0.188 = 19%
Seems like this would be better, bringing the first two in pretty close to the 'Hinote criteria'. You didn't put a number on your third criterion, but philosophically, we're headed in the right direction.

(b) Shorten the intake pipe 2 inches, leaving everything else alone:
Datum 1 (intake mouth station): 8.5 / 26 = 0.327 = 33%
Datum 2 (intake tube length): 3.63 / 26 = 0.140 = 14%
Datum 3 (intake port station): 6.0 / 26 = 0.231 = 23%
In this case, the first is identical to (a) and the remaining two data are significantly worse, according to the criteria!

(c) Lengthen the intake pipe 1 inch, lengthen the tailpipe 8 inches:
Datum 1 (intake mouth station): 11.5 / 34 = 0.338 = 34%
Datum 2 (intake tube length): 6.63 / 34 = 0.195 = 20%
Datum 3 (intake port station): 6.0 / 34 = 0.176 = 18%
This seems to hit all the criteria pretty well, though the engine is significantly longer than the original. We're getting pretty far away from my "ideal" target, an engine of basically Dynajet dimensions. This is not a bad form factor for the large models many hobbyists are comfortable building today, though. The engine could fit some scale models with a fuselage length of only 1.5 x the engine length [i.e. 51 inches nose to exhaust port], which is not an outrageous size, even for U-control scale if lightly built.

So, it looks like approaches (a) or (c) would help get the engine running, with the former favoring compactness and the latter possibly coming closer to optimal resonance, according to the Hinote Criteria.

What do you think?

L Cottrill

[larry cottrill]

Bill -

Of course, I've forgotten the fact that I have one engine that starts and runs well enough to be considered successful in some sense, the Elektra I. And, the ability of the design to resonate at its final dimensions has been demonstrated by having it independently duplicated and tested. So, we should apply the Hinote Criteria to that design, and see how close or not so close it hits:

Again, it is not possible to accurately measure the exact station of the intake port. However, we can estimate it from the photos of the finished chamber [before the near side plate gets welded on] which both Steve B and I have provided for all to see. The box itself is almost exactly 3.5 inches in length, so I'm going to call the station of the port at the centerline of the intake tube 2.25 inches from the 'front plate' [the inside of the front edge of the box].

Now, for purposes of calculating the ratios to compare to the Criteria, we need to imagine the intake pipe laid out parallel to the tailpipe, a la Thermojet. The intake tube, including the intake flare, is almost exactly 6.5 inches long at its centerline. So, the station of the intake mouth would be 2.25 + 6.5 = 8.75 behind the front plate.

The engine in its final form [35 inch tailpipe] has a total length of almost exactly 38.5 inches. So, calculating the ratios to put against the Criteria, we derive:

Datum 1 (intake mouth station): 8.75 / 38.5 = 0.227 = 23%
Datum 2 (intake tube length): 6.5 / 38.5 = 0.168 = 17%
Datum 3 (intake port station): 2.25 / 38.5 = 0.058 = 6%

According to the 'Hinote criteria', this engine must be considerably longer than it really needs to be! Of course, this is based purely on the demands of acoustic resonance. The Criteria don't tell us anything about the impedances of tubes or the masses of gas in those tubes. Both of these factors are very oddly set up in the protoype Elektra I, which has a very low tailpipe / intake volume ratio compared to most valveless designs [the intake tube is approx. 0.75 inch ID with the tailpipe only a hair over 1 inch]. Of special interest is that the operating frequency of the engine was observed to be independent of the engine length, over a wide range. Nevertheless, resonance was definitely achieved, and at an overall length much larger than intuitively expected [and now, predicted by the Criteria].

Let's see if we can hit the criteria closer with a simple adjustment -- cutting 6 inches off the tailpipe:
Datum 1 (intake mouth station): 8.75 / 32.5 = 0.269 = 27%
Datum 2 (intake tube length): 6.5 / 32.5 = 0.200= 20%
Datum 3 (intake port station): 2.25 / 32.5 = 0.069 = 7%

This hits pretty close, a bit shy on the intake mouth station but right on the money for intake pipe length. Perhaps I should go back to something along these lines, trying a bit harder to get a sustaining run!

L Cottrill

[hinote]

So, it looks like approaches (a) or (c) would help get the engine running, with the former favoring compactness and the latter possibly coming closer to optimal resonance, according to the Hinote Criteria.

What do you think?

L Cottrill

Larry, I can only say, "try it!" and see what happens. Given that it won't run right now, you've got to do something.

I'd try your method (c) because it matches the Chinese most closely.

It's the same old story--lacking the mathematical ability to calculate the necessary dimensions, you can only resort to cut-and-try.

Bill H.
Acoustic Propulsion Concepts