Uniflow Woes

Quick description: The tank is 2" long, 1.75" high and 1" wide with a 1/4" deep wedge. The fuel pickup runs from the outside front bottom corner to the back of the wedge. The Uniflow vent runs from the top outside front corner to 1/2" in front of the pickup, along the wedge line. There are no other vents. The uniflow faces directly into the airstream.

Anyway, now when I start the engine, I get a great run on the ground, but when I launch it, it manages almost exactly one fine lap, then suddenly the engine dies. No spluttering, the revs drop to zero in about 1 second.

It sounds like it's overheating, but I've tried every needle setting I could make it take off with, and it's although I got a couple of runs (set rich I think) it's never consistant over the run.

It was a windy day, and the engine was revving up going into the wind, and dying out heading downwind. Sounds rich to me! Yet the cut-out sounds over-lean verging on seizing.

Oh, and after a cut-out it restarts first flick without any priming which makes me suspect it can't be lean.

Totally confused, and would much appreciate any input!
Thanks
Jonathan

Jonathan, Try choking the uniflow vent down by one tube size. Place a piece of fuel tubing over the uniflow vent and let it extend out abour 1/2 inch and wire it in place. Then insert a piece of 3/32 inch OD brass tubing inside of the fuel tubing and wire that in place. Test fly and see if that helps. It solved my similar troubles. If it cures the problem, you may wish to take the fuel tubing off and solder the small brass tubing directly inside of the uniflow tube.
Try also a hotter plug.
There can be NO leaks for the uniflow tank to work properly. You do have an overflow vent that you cap off after you fill the tank don't you? Be sure that there are no leaks there.
Bigiron

Actually I use diesel, and I didn't put an overflow in at all. I use the spraybar as the vent for filling.

I'll try your trick though. Currently the uniflow vent is 1/8" copper piping straight into the airstream.

I wondered whether there were bubbles getting in from the uniflow vent, but it looks okay, and there's no misfire at all before it dies, regardless of the needle setting (assuming I can get it airborne)

Thanks
Jonathan

Describe how you fill the tank without an overflow. Generally, the normally positioned overflow vent exiting the bottom front and extending to the inside top of the tank is open to allow filling, then is plugged with a small length of tubing with a rivet in it.

Is it possible that you arent getting any fuel in? Without a vent, fuel will quickly blow out the uniflow or feed.

I generally suction test the tank by plugging all the lines except one then sucking on the only open one--block it off with yur tongue and it should stay attached by suction. If it wont, its leaking. The smallest pinhole will cause failure.

Or pressure test under water. One problem is that sometimes the solder flux will plug a hole until fuel or alcohol dissolves it--then leak city!

After you have the tank leakproof, start out with the tank CL about 1/8" to 3/16" above the spraybar center. (really the internal exit height of the Uniflow tube is the measurement point--above the spraybar)

This should get you in the ballpark. If not, there is something else wrong.

Also, if there is a normally positioned vent in the tank, leaving it open allows you to run without the uniflow feature to establish a troubleshooting base-line.

Fix the leaks first, then see what happens.

Hope this helps.

The theory about leaks appearing sounds like a good one. I tested mine under water by pressurising it. It was fine, but now it has leaks! BTW I use diesel, so there could be all sorts of nasties in there.

I left the vent out because I thought it would be another thing to leak!

I'll fix it up and re-try.
Thanks
Jonathan

>Hi all,
>I've built my first Uniflow tank
>(I think) and I'm having
>some very strange engine issues.
> I'll start by admitting
>though that I think there
>are leaks. There weren't
>any when I finished testing
>it, but on the plane
>I can see some small
>leaks. I guess the
>solder must have cracked under
>vibration.
>
>Quick description: The tank is 2"
>long, 1.75" high and 1"
>wide with a 1/4" deep
>wedge. The fuel pickup
>runs from the outside front
>bottom corner to the back
>of the wedge. The
>Uniflow vent runs from the
>top outside front corner to
>1/2" in front of the
>pickup, along the wedge line.
> There are no other
>vents. The uniflow faces
>directly into the airstream.
>
>
>Anyway, now when I start the
>engine, I get a great
>run on the ground, but
>when I launch it, it
>manages almost exactly one fine
>lap, then suddenly the engine
>dies. No spluttering, the
>revs drop to zero in
>about 1 second.
>
>It sounds like it's overheating, but
>I've tried every needle setting
>I could make it take
>off with, and it's although
>I got a couple of
>runs (set rich I think)
>it's never consistant over the
>run.
>
>It was a windy day, and
>the engine was revving up
>going into the wind, and
>dying out heading downwind.
>Sounds rich to me!
>Yet the cut-out sounds over-lean
>verging on seizing.
>
>Oh, and after a cut-out it
>restarts first flick without any
>priming which makes me suspect
>it can't be lean.
>
>Totally confused, and would much appreciate
>any input!
>

As the others have mentioned, for the uniflow to work properly, the tank can't leak. But if it does, typically, the result is just gradual leaning out as you run the fuel out. This doesn't appear to be the case here.

Sounds more like there might be some crud/solder balls in the tank that choke off the flow once you take off, or possible internal cracks in the fuel pickup.

Are you sure you are getting the tank full of fuel? I would have a hard time filling the tank as you describe.

I have set up several tanks like he described and all you have to do is stand the model on its nose to fuel it, works like a champ.

Then the problems returned. So I'm really starting to suspect it must be leaking again and that's the cause of the problems.

I don't really understand why sealing should be so critical on this form of tank, as my past standard vent tanks have all leaked like sieves!

Anyway, I can really see what the plane/engine is capable of now and I like it. So I'll definitely stick with a uniflow, and just test and seal it again, with silicon or something.

When the airspeed was up it really started to snap turns, but loops are a bit too much for it I think. I guess it just loses too much airspeed in the climb.

Thanks guys.
Jon

You will find the uniflow tank to be much better in performance, giving a more consistent run from beginning to end, but, yes, you do need to insure that you have no leasks.

Leonard Neumann

but that's what I don't understand. I figure that if a Uniflow tank leaks, then at worst you are back to a standard vent tank. But none of my standard vent tanks had any weird behaviour, so something else must be happening.

In fact at one point it was running fine on the ground, then the second it was released and started to roll, it quit again. So it seems linked to the forward motion, or the flow of air due to that motion.

You're right about the better performance though!

With a uniflow tank you will get a slight vacuum inside, and the more fuel, the greater the vacuum which equalizes the pressure. You have less, but more consistent pressure from beginning to end. In order to compensate, you will have to open the needle more while running uniflow. If you suddenly develop a leak in the system, you will suddenly have more pressure and a very rich mixture which is probably what is causing your engine to quit.

As a means of seeing what the difference is, cap off the overlow vent on the ground and start up your engine and set the mixture to where you normally run it. Now remove the cap from the overflow vent and see what happens to your engine run. Pffft!

Leonard Neumann

Bob

Oh Boy! I'll supply some basics and others will probably supply corrections and additions.
First consider the tank you are probably used to, the standard suction tank. You have a line that runs from the outside rear of the tank to supply fuel, and probably two lines on the inside front for fill and vent.
The uniflow has the line that runs from the outside rear as before. The vent line will run from the inside top of the tank to an exit point. This is capped when flying. The third line (uniflow line) is run from the outside rear, about 1/4" forward of the fuel pickup tube, to an exit point. Supplying atmospheric pressure near the pickup tube causes the engine to see only the pressure that it would have if it had the amount of fuel from the uniflow vent to the back of the tank... no matter how much fuel is actually in the tank.
The purpose is to provide consistant fuel pressure throughout the flight. It prevents the engine from running from rich to lean as the tank is emptied.
Hope this helps.
George

Bob, go down to the bottom of the Discussion Group page and click on page 12, to find "Inventor of the uniflow tank". It will tell you more than you want to know. Actually, having gone back and looked at the discussion in detail, it is not all that helpful in understanding the function of the tank.

To elaborate a little: the uniflow tank is closed except for the fuel outlet and the air inlet (uniflow tube). So long as the air inlet is submerged in fuel, the engine will think it is sucking fuel from the end for the air inlet. The engine will not know how much fuel is in the tank, and thus will run the same throughout the flight. How's that for a scientific explaination of the phenomenon?

Usually the uniflow ends somewhere back near the fuel pickup tube; not too close, or air may get sucked into the fuel pickup tube. The Palmer tank has the uniflow going to the front outside corner of the tank. I have built one of these and it works fine. I am thinking, because uniflow tanks work fine with the uniflow ending toward the back of the tank as well as at the front of the tank, it may not make any difference where the uniflow ends along the length of the tank. Up or down, in or out, does make a difference, of course.

If the scientific explaination given above actually reflects reality, perhaps better fuel suction would result from having the uniflow end at the front of the tank rather than the rear.

Jim Thomerson (excessively fond of uniflow tanks)

The real key is, the fuel pressure head affects how much vacuum is in the tank. The fuel exerts pressure--the more fuel, the more pressure, because it is being thrown towards the outside of the tank. With the air inlet under this pressure, it reduces the pressure of the air that is allowed inside the tank. The more fuel, the more pressure, the less air. The fuel, with a partial vacuum behind it, thus exerts less pressure forcing fuel into the engine. As the fuel is depleted, the air pressure in the tank is allowed to increase, all of which helps to keep the pressure more equal. Hence uni-flow. (one pressure flow).

Take an empty fuel jug and fill it with water. Punch a small hole (Drill a 1/16 hole) in the side of the jug towards the bottom, and, with the lid off watch the water squirt out. With a full jug you get a lot more flow than when the jug is nearing the bottom. Now drill a hole in the cap just large enough to pull some plastic fuel tubing through and allow the tubing to extend all the way to the bottom of the jug. Seal the area around the cap, fill the jug again and replace the lid making sure the tubing extends all the way to the bottom. Now watch the flow of the water coming out of the jug. The pressure will be a lot less, and it will flow a lot more constant throughout the emptying of the jug.

And, by the way, a full jug of water equals about the same pressure you get with 3 g's of pull on the lines, which is what we often get. So the effect will be about the same.

Leonard Neumann

I've run the same engine, ST G20-15 diesel, on tanks with the uniflow in the front and in the rear, and I can't tell any difference. One virtue of putting the uniflow up front is that you can solder it full length to the front of the tank.

Incidentally, the reason there is a sight vacum in uniflow tanks is that the uniflow outlet is 'below' or outside of the fuel outlet in the engine carburator. so the head in the tank is negative rather than positive while flying in a circle. I would guess it at around -0.03 psi or less for the usual set-up. Of course if you use an inboard tank you could get a little positive head.

As you said, the depth of water in your jug doesn't make any difference, just the head between the uniflow outlet and the water outlet. If you pull the uniflow tube up, the rate of flow will increase.

I'd encourage people to try a Palmer-type tank and see how it works (or not).

Jim, it doesn't make any difference that the inlet tube is outside or inside the carb. The vacuum will still be there because the air has to bubble through the water. The only difference is the positive or negative effect of the fuel head in relation to the flow back to the carb. As I remember from my Physics class, fifteen pounds of normal airpressure will hold up about 30 feet of water in a closed column. Since we have an effective 6 inches of liquid slightly lighter than water (but close enough) the vacuum will be very definitely partial. But all we are doing is compensating for the difference in fuel head from the beginning of the run until the end.

And, yes, the location of that tube very definitely does make a difference. Since the fuel is generally being thrown to the outside, we need to locate the vent tube somewhere as close as possible to the outside. I still prefer it as close to the pickup tube as possible (about 3/8 inch) but it will work half way up the wedge, also. But if you put the vent tube in the middle of the tank (between the outside and inside) you would only compensate for half of the fuel head. And the last half of the run would be the same as if there were no uniflow at all.

Try my illustration of the fuel jug. This is a "true" uniflow set-up. Try it with the cap open and watch the flow (standard vent). Try it with the cap sealed and the vent tube going all the way to the bottom (true uniflow). Much less flow at the beginning, but constant all the way. And then try it with the tubing half way down in the jug. You will have a greater flow at the beginning, and it WILL vary.

Leonard Neumann

I think you guys are arguing apples an oranges. The side-to-side position of the uniflow vent sets the pressure at the spraybar inlet. If the tube only went halfway across the tank, the pressure at the engine would be about the same as the pressure at the uniflow inlet for half a tank, and the pressure in the ullage of the tank would be higher than a full uniflow. Of course, once the fuel level got down to the vent, it would just start getting leaner.

The tank GMA shows on his website, and the Nobler plans, is kind of this way. The vents are about 3/4 of the way to the inside edge of the tank, so the fuel pressure is constant for the 1st 25% of the fuel burn, then leans out for the lst 75%. I hate to differ with Mr. Aldrich, since he's not here to argue back, but the shape of the tank makes *no* difference as far as fuel delivery pressure goes. The internal position of the vents is the determining factor.

If anyone wants to experiment with this sort of "partial uniflow" tank, it's easy to make an adjustable uniflow vent. Just make a tank with a movable vent.Use telescoping tubing and a piece of fuel to seal it up. Solder a short piece of the larger tubing to the side of the tank, and let the longer inner tube slide back and forth.

Leonard Neumann

>If fuel did in fact flow
>straight out to the wedge,
>it would make no difference
>whether the inlet were in
>the front of the wedge
>or the back. But
>the fuel will be angled
>a bit in flight--mosttly out
>because of centrifrugal force, but
>somewhat angled back because of
>other factors.
>

I'm not sure why you think it goes to the back. The only other forcing function in steady level flight would be the yaw angle, which would ideally be 0 and hence have no effect, or more usually, a little bit nose-out, which would make the fuel go to the front of the tank*.

Various accelerations make it move around, of course. On takeoff, it surges to the back. On a maneuver entrance, it flies forward due to decelleration from drag, and then back as the speed recovery is taking place. And various directions of gravity of course have minor effects, too.

In general steady motion, unless you do something to the tank, like tilting the back end out, or making it tapered with the wide end in the rear, there's not much reason to think the fuel has a tendency to go anywhere fore-and-aft.

Brett

*this is of course why inside cutoff loops work better than outside - the nose yaws out a little, which moves the fuel to the front in addition to the bottom.

That's why I like the design of the tanks you make, Brett. With the tank wider at the back and on the wedge side, there is no doubt that the fuel will move to the outside and back, and there is far less tendency to uncover the uniflow vent until you get pretty close to that last drop of go juice.

Leonard Neumann

Towards the front, from the decelleration of maneuvering, followed shortly thereafter by the back as the speed gets recovered, disregarding gravity. It also sloshes to the outside of the turn, towards the top or bottom of the tank.

The acceleration in Xbody is probably not all that high, in either direction, but can match up with the natural slosh frequency, and it's for sure lightly damped so it goes back and forth for a while.

The acceleration in Y is more or less constant and about 3 g's, and in Z it's dependent on the how tight the turn is, but can be around 10 g's. The angle of the fuel head in level flight is about 70 is degrees relative to the Z faces of the tank, but in the 10 g corners it's only about 16 degrees in steady state.

Of course, the actual fuel motion is not a static problem but instead involves a lot of dynamics that we have never come close to solving on satellites with unlimited dollars, much less on model airplane money.

To absolutely guarantee that the fuel stayed at the pickup for the whole flight, you would have to make the Z faces of the tank (or the "wedge angle", if you prefer) less than

atan({centrifugal acceleration along Y}/{maximum acceleration from maneuvering along Z})

Of course, this is what causes the cutoff loop to work at all - the fuel goes toward the bottom of the tank, and at low levels it pulls away from the pickup.

Most two-stroke stunt engines ride over a few bubbles pretty easily, since they speed up when the mixture gets lean. 4-strokes are very sensitive to cutting off from bubbles. Not coincidentally, if you look at Beringer's 4-stroke tank, it has a very shallow wedge to minimize this problem. It would work great on a 2-stroke, too, but the wedge shape wastes too much space on a standard stunt plane, and you'd have trouble getting enough capacity.

>
>That's why I like the design
>of the tanks you make,
>Brett. With the tank
>wider at the back and
>on the wedge side, there
>is no doubt that the
>fuel will move to the
>outside and back, and there
>is far less tendency to
>uncover the uniflow vent until
>you get pretty close to
>that last drop of go
>juice.
>
>Leonard Neumann

The taper does work pretty good. I can;t take credit for it. My old Lexington flying buddy Ed Robbert made them like that, and I think Brian Eather picked it up from him. More taper works better, and taper with a sump/chicken hopper works even better.

I'm going to make a "backwards" taper tank, with the wide part in front. The idea would be to cut the fuel head variation from X acceleration.