Uniflow tanks - a bit of physics!

A colleague suggested recently a different sort of uniflow tank for a profile model. See 'rough' diagram below. The obvious differences are that (A) the tank is conventionally vented top & bottom and (B) the outlet goes through the fuel and through the back of the tank. The fuel tube goes the 'long way round' to the NVA. He's tried it and maintains it works. He says that the pressure at point P in the outlet tube will be the same as at the surface of the liquid and will therefore be unaffected by fuel height and centrifugal force and as the tank is conventionally vented will be at atmospheric pressure, just as a standard uniflow. Is he right. If not what bit of physics is misunderstood? There is alspo a local debate going on as to whether the long tube to the NVA on the inside of the plane helps/hinders an even flow. Can anyone help or pour cold water on this design?
Adrian

> Is he
>right.

If the drawing is as it seems, then no, it does not work.

>If not what bit of physics is misunderstood?

The idea that the pressure is dependent on the depth of the fluid, in other words, the basic principle involved.

The pressure at the pickup is atmospheric+rho*X*a, where rho is the fuel density and X is the depth from the surface to the pickup, and a is the cetrifugal force - just like any other suction tank.

The pressure at the external end of the pickup tube is lower than this, in fact lower than atmospheric, since it's well "above" the surface of the fuel. But this will get recovered when you put the line around to the spraybar. The path it takes has no real effect on the static pressure. The total length of the line effects the pressure drop from flow but that's a secondary effect.

In short, of the tank is mounted conventionally behind the engines, the net static pressure as delivered to the engine is not any different than any other suction tank. Moving the tank around will change it, but simply routing the lines differently makes no difference in the static pressure.

> There is
>alspo a local debate going on as to whether the long tube to
>the NVA on the inside of the plane helps/hinders an even
>flow.

It may make some difference, not intrinsicially due to the route it takes, but because the path is longer and therefore has more resistance to flow. There are some interesting developments on the topic of fuel line resistance. Lower is better, although most of the time anything reasonable doesn't make a difference. However, there is an interesting theory under discussion about fuel line flow resistance in regards to 4-strokes, their very intermittent fuel flow pattern, and their preference for clunk tanks.

Brett, can we look at this again? I understand perfectly the pressure at the pickup point (atmos + weight of fuel etc). But, is the pressure in the fuel tube AS IT CROSSES THE FUEL SURFACE the same as the pressure of that fuel surface ie atmosphere. If it is then surely as the fuel is depleted that point will always be atmosphere (unaffected by height & centrifugal force). If so then the fuel delivery initiation point is at constant pressure. Which is the same effect as in a standard uniflow, isn't it? I hope I've explained my point clearly. The fuel pipe routing may be a completely different discussion. What do you think?

>Brett, can we look at this again? I understand perfectly the
>pressure at the pickup point (atmos + weight of fuel etc).
>But, is the pressure in the fuel tube AS IT CROSSES THE FUEL
>SURFACE the same as the pressure of that fuel surface ie
>atmosphere. If it is then surely as the fuel is depleted
>that point will always be atmosphere (unaffected by height &
>centrifugal force). If so then the fuel delivery initiation
>point is at constant pressure.

To the extent that you follow it, you are correct. But the pickup tube is full of fuel. That tube has to get back to the spraybar on the engine somehow. It comes out of the inboard side of the fuse, then (assuming a conventional profile engine mount) then has to work it's way back to the spray bar - in effect, moving to a point "downhill". Which increases the pressure just as much as it dropped "climbing" up the tube.

The net effect is that the pressure at the engine end is just exactly the same as if you ran it out the front end of the tank.

If you want to keep the "surface of the fuel" concept, consider a tank with a floating pickup. The pickup floats on the surface and picks up fuel from there. At the beginning of the run, the float is right up near the fuselage, and the fuel flows "downhill" about 1/2" to the spraybar, which results in relatively high pressure (more than atmospheric). As the fuel runs out, the pickup eventually gets to the point of being "level" with the spraybar, and at that point it's equal to atmospheric pressure at the engine. At the end, the float is all the way at the point of the wedge, and the fuel has to flow "uphill" to the engine, and thus presents lower pressure to the engine. This is exactly analogous to your "surface of the fuel" argument, and in fact the pressure is exactly the same whether the pickup moves or not.

If you don't like that one, think the the entire system, tank, pipes, filter, etc. as a single fluid field. The pressure is still equal to the depth from the surface to the outlet. But the outlet you care about is the engine end.

Which is the same effect as
>in a standard uniflow, isn't it? I hope I've explained my
>point clearly. The fuel pipe routing may be a completely
>different discussion. What do you think?

The routing doesn't really matter, with the exception that the tubing resists flow, and the longer or smaller the tube, the more pressure drop.

Brett, I like the word "intrinsically" in your message no.1. To raise an old chestnut, I think all your explanations here also explain why I still can see no practical difference in the LA set-ups, either venturi mounted nva or remote. Physics wise they are intrinsically the same.
Also on the same subject as this thread, another old "folk myth" concerning fuel flow and head was the trick perpetrated in vintage days of putting a loop in the fuel line, supposedly so that any variation in fuel head in one direction would be counterbalanced by the fuel on the opposite side of the loop. Brett's explanation also explains why this doesn't work.

>Brett, I like the word "intrinsically" in your message no.1.
>To raise an old chestnut, I think all your explanations here
>also explain why I still can see no practical difference in
>the LA set-ups, either venturi mounted nva or remote.

Once again, all true assuming a *static* condition. It does make a difference in the dynamic pressure. Two weeks ago, I might have assumed that the line drag was completely negligible. Now, I'm not quite so sure that's true in all cases.

I'm afraid this is not the uniflow effect at all. It is essentially equivalent to leaving both vents open, they are still seeing one atmosphere.

As the tank empties, the height of the fuel column will decrease and the pressure associated with the "weight" of the fuel will decrease accordingly. The mixture will lean as a result.

By contrast, a uniflow vent will maintain the pressure of the fuel near the outlet, and the air that bubbles in at that point will be at one atmosphere. The air pressure above the fuel top surface of the uniflow will decrease accordingly, to offset the weight of the fuel head.

[photo not recovered: 3fdc6a1f54376fcc.jpg]

It is worthwhile to make up a uniflow demonstrator with a small coffee can (keep the plastic lid), with outlet and uniflow vents soldered in it. You can cap the uniflow vent and uncap the can to compare a conventional "suction" tank..

Capping the (bottom) vent to the top of the conventional two-vent tank would come closer to making a uniflow. But the remaining outlet has its terminus too far from the outlet to be as effective as a proper uniflow vent.

I'm not sure what effect is actually being observed with the linked vent tube setup. Perhaps it is the elimination of some venturi effect on the vent ends as they are no longer seeing the direct effects of the prop wash.

[photo not recovered: 3e69bb3870f12ad5.jpg]

"An investment in knowledge pays the best interest." -Benjamin Franklin

Thanks, Larry and Brett, for accurate responses. Perhaps the original title should have been "Black Magic and Old Wives' tales".

If anybody is interested, I wrote a lengthy article on uniflow venting, etc. way back in the July 1978 issue of Model Aviation in my CL Racing column. You can see that article by visitng the MA archive on the AMA web site.

Regards,

>Thanks, Larry and Brett, for accurate responses. Perhaps the
>original title should have been "Black Magic and Old Wives'
>tales".
>

For whatever reason, tanks are the subject of more "fencepost wisdom" than anything else. and Uniflow seems to be frequently relegated to the realm of magic. No idea why a can with a pipe in it seems so daunting.

Brett:
I tried to explain how a uniflow tank works in simple terms, when I rewrote Dave Rees’s article in Stunt New.

Given your knowledge in fluid dynamics I was hoping you could answer a question I have concerning the first question. How much effect does air in the fuel have in relation to static pressure at the head of a standard vented tank. I assume it would affect it since air is a compressible Fluid and liquids in our fuels are non-compressible fluids.

Larry

LAST EDITED ON Dec-14-03 AT 10:38 PM (CST)
 
Larry (Cunningham) How's the foot doing?
Are you still in a wheel chair?

Larry

LARRY,
Hows you foot?
You In a wheel chair?

What you do DROP a Fox on your foot?

Scott(just want to be in the loop)Riese

OK, here's something to chew on... I've been thinking about this for years, too, since Dan Miles brought it up at a Combat meet in Detroit back in the 80s.

Back then, the "limpy bladder" (as Doc Passen used to call them) were all the rage - you put a limp balloon in some kind of container behind the engine on a Slow Combat plane. This, in essence, is the same thing as the now-popular Tettra tanks used in Pylon racing. People experimented with all sorts of quick-change containers (urine specimen cups were popular) since refueling these things were a pain. Also, with the large venturis used in Slow Combat, people were running into problems with the fuel head changing drastically during the run (a flat, squat container was better than a tall one, as you can imagine).

Dan experimented with a balloon that was superglued to the roof of a container (i.e. the part that abutted up against the profile fuselage) so that the fuel was essentially being suspended by a flexible sling, and supported by atmospheric pressure. He says you got very, very steady runs - no problems whatsoever with leaning out during the run. He said you could predictably set the needle based on how much the balloon was "elevated" off the fuselage (he tried different thickness of spacers) - the more the tank was elevated off the fuselage, the more you had to open the needle.

So a conversation developed around this - people were trying to figure out how this worked. How exactly did it, and how did you go about proving, through the P = rho*g*h equation, that the fuel head did not change?

Have at it...

Isky,
One indirect way to check, if not prove, little or no change in pressure head with uniflow vented tanks would be a bench schematic like Larry's coffee can sketch. As Head = Height to open surface X density:--

The 'throw' of the outlet stream reflects static head. Given care that the tube IDs are not so large that "Glugging*" of vent bubbles occurs, and that the soft lid isn't oil-canning, the stream of air bubbles vented to the tank interior should be reasonably steady. So should the stream throw distance from the "tank" outlet jet...(Both those experimental setup problems will cause a pulsing output.) You may need to add fuel tube to the outlet, so that you can place the open end at a height that WILL draw vent air to the tank interior -- It might not vent for flow otherwise...

* - Glug, an old measure of how much molasses to add to a recipe. Tip the molasses jug, and due to viscosity, etc., it releases none, then a "glug" then none, etc.

One other caution for bench, static testing: I set up something like that for a special purpose with a 'diesel' engine and fuel. Vent bubbles were not evident. The vapor pressure of the tech ether in the fuel is so low that liquid ether gassed out directly within the volume, without need to draw air from the exterior!

LAST EDITED ON Dec-14-03 AT 10:49 PM (CST)
 
Scott
No, though I have several piston and liners that almost made it out the 2nd story window last night the fits suck, way to loose. It was a lot easier when they were too tight. Does look like I was asking myself so I edited the original post. Larry Cunningham has had very serious problems with his foot and has been in a wheel chair. He was worried about loosing it. Threw my trails of recovering from back surgery, trying to recover from the winter influx of Foxes and building a Caprice I haven't stayed in touch with him like I should.

Larry (trying to get a thicker skin) Foster

Larry (and other friends),

My condition is called Charcot Foot. It is a diabetic induced disease where the neurological system doesn't work properly, and bones tend to break and fail. In my case, the ankle area has failed bone fragments. Treatment is to immobilize it in a cast for extended periods (3 months to up to a year).

My foot is improving, but is still in a cast, and I'm still using the wheelchair (at work). Although the joint seems to be fusing, I'll still have to lose the front portion of my foot, probably in January.

I'm optimistic, I expect to get a prothesis and be walking again fairly well in early 2004.. Meanwhile, my sweet wife is driving me everywhere, while I drive her.. crazy.

Thanks for your concern and friendship.

[photo not recovered: 3e69bb3870f12ad5.jpg]

"The human foot is a masterpiece of engineering and a work of art." -Leonardo da Vinci

Brett -- One of my columns explored the line losses for fuel tubing and fuel filters. I did a set of measurements at two different flow rates -- 2/3 of an ounce per minute (typical of a glow 40) and one ounce per minute (typical of a 60). The tests used Aerotrend tubing and Taffinder's 10PA fuel. The tests were done at 75 degrees F. For the 40 flow rate, 3/32-inch tubing dropped 0.12 inches of H2O per inch of tubing, and 1/8-inch tubing dropprd 0.02 inches of H2O per inch of tubing. For the 60 flow rate, 3/32-inch tubing dropped 0.17 inches of H2O per inch of tubing, and 1/8-inch tubing dropprd 0.03 inches of H2O per inch of tubing. The fuel filter, a Sullivan Crap Trap dropped 0.027 inches of H2O for the 40 and 0.33 inches of H2O for the 60.

I was surprised at how low the losses were. I started off trying to measure the pressure drop across one foot of fuel tubing. I had to use six feet of tubing before I could measure pressure drops with any degree of accuracy. Considering that a typical fuel draw for our engines is about 15 inches of H2O, I would say that minimizing the length of fuel line is probably not as important as avoiding kinks and sharp bends.

I hope that this data will be useful in analyzing the four-stroke problems.

Noel Drindak

>Brett -- One of my columns explored the line losses for fuel
>tubing and fuel filters. I did a set of measurements at two
>different flow rates -- 2/3 of an ounce per minute (typical
>of a glow 40) and one ounce per minute (typical of a 60).
>The tests used Aerotrend tubing and Taffinder's 10PA fuel.
>The tests were done at 75 degrees F. For the 40 flow rate,
>3/32-inch tubing dropped 0.12 inches of H2O per inch of
>tubing, and 1/8-inch tubing dropprd 0.02 inches of H2O per
>inch of tubing. For the 60 flow rate, 3/32-inch tubing
>dropped 0.17 inches of H2O per inch of tubing, and 1/8-inch
>tubing dropprd 0.03 inches of H2O per inch of tubing. The
>fuel filter, a Sullivan Crap Trap dropped 0.027 inches of
>H2O for the 40 and 0.33 inches of H2O for the 60.
>
>I was surprised at how low the losses were. I started off
>trying to measure the pressure drop across one foot of fuel
>tubing. I had to use six feet of tubing before I could
>measure pressure drops with any degree of accuracy.
>Considering that a typical fuel draw for our engines is
>about 15 inches of H2O, I would say that minimizing the
>length of fuel line is probably not as important as avoiding
>kinks and sharp bends.
>
>I hope that this data will be useful in analyzing the
>four-stroke problems.

I think that the crux of the issue is that the flow is not constant, but "bursts". They don't take much fuel, so the average flow rate is low, but the actual time it's drawing is only a small fraction of the time that it's running. I would guess that a two-stroke is close to a continuous flow.

Like I mentioned, I would have thought the whole thing was negligible myself two weeks ago. But the extremely dramatic improvement in 4-stroke run characteristics from running a clunk tank (note: a standard Veco T-21 tank IS NOT VERY GOOD FOR 4-STROKES; if you have one, run, do not walk, to the hobby shop and get a Sullivan SS-4 clunk tank and set it up like Godzilla says to), coupled with the apparent dislike for uniflow (and associated continuous, but low, pressure) is caused by something other than PFM.

>Like I mentioned, I would have thought the whole thing
>was negligible myself two weeks ago. But the extremely
>dramatic improvement in 4-stroke run characteristics from
>running a clunk tank (note: a standard Veco T-21 tank IS NOT
>VERY GOOD FOR 4-STROKES; if you have one, run, do not walk,
>to the hobby shop and get a Sullivan SS-4 clunk tank and set
>it up like Godzilla says to), coupled with the apparent
>dislike for uniflow (and associated continuous, but low,
>pressure) is caused by something other than PFM.
>
> Brett

Brett,

I would have thought Godzilla effectively made above point to you months and months ago.

What happened two weeks ago? If you're just slummin' with Uncle Jimby and his ongoing efforts with a Saito 56, never mind.

But I am intrigued by your sly comment...

>Brett,
>
>I would have thought Godzilla effectively made above point
>to you months and months ago.

A really smart Aussie got me to looking at this flow rate thing more than 6 months ago.

There is more than one way to skin that cat...

>>Brett,
>>
>>I would have thought Godzilla effectively made above point
>>to you months and months ago.
>
>A really smart Aussie got me to looking at this flow rate
>thing more than 6 months ago.

Precisely. The very first thing I thought of when I saw the dramatic improvement in the Satio 56 at the 2002 NATs from switching tanks, was that there was some difference in the flow that the 4-stroke cares about, and the 2-stroke doesn't (or at least doesn't care much about). The "start/stop" nature of the fuel draw, and brief but very strong actual vacuum on the 4-stroke somehow matters. The start/stop nature suggest a higher flow velocity when it's sucking, and dead stop in between.

Then, I find that there is empirical evidence that using a regular tank with larger-than-normal tubing significantly improves the situation. The key feature of larger tubing is reduced flow drag. This sort of fits with suction working better (higher absolute pressure), and sort of with the clunk tank. I'm not sure I have a complete story, but I wouldn't discount the observation.

BTW, can anyone with Fox 35 instructions (you know the little paper thing that most people throw out immediately) confirm for me that it says to use a tank with 1/8" I.D., vice O.D., tubing?

Brett,

Do I have to come down there and spy on you? You have failed to answer the question; indeed, your reponse is too clever by half.

Regardless of what has prompted all this concern over fuel tanks for four-strokes, are you really going to cause me to re-plumb my plastic clunk tanks with larger tubing?

>Brett,
>
>Do I have to come down there and spy on you? You have failed
>to answer the question; indeed, your reponse is too clever
>by half.

You didn't read Brad's response. I have heard that Brian Eather, Godzilla's "really smart Aussie" managed to be able to use hard tanks successfully by using larger-diameter plumbing. It's still more than a little vague why it would be better, and what makes it different from a 2-stroke, but it's an important observation.

>Regardless of what has prompted all this concern over fuel
>tanks for four-strokes, are you really going to cause me to
>re-plumb my plastic clunk tanks with larger tubing?

It would be an interesting experiment. But the clunk tank seems to address whatever issue it is all by itself.

For all reading this - it's a curiosity as to *why* it works, but going from any hard tank with regular-size plumbing, to a clunk with supplied plumbing, is the most dramatic improvement you can make for a 4-stroke. Gutless overheads go away. Second most important is to get rid of the stock carb and get a Pat Johnston venturi setup (available from UHP). I point this out because it directly contradicts some of the "early adopters" observations, and a lot of that advice is still floating around.

I have heard that Brian
>Eather, Godzilla's "really smart Aussie" managed to be able
>to use hard tanks successfully by using larger-diameter
>plumbing. It's still more than a little vague why it would
>be better, and what makes it different from a 2-stroke, but
>it's an important observation.

Undoubtably, he is really smart isn't he?

>>Regardless of what has prompted all this concern over fuel
>>tanks for four-strokes, are you really going to cause me to
>>re-plumb my plastic clunk tanks with larger tubing?
>
> It would be an interesting experiment. But the clunk tank
>seems to address whatever issue it is all by itself.

There is another really smart Aussie by the name of Joe Parisi. He has tested almost every version of fuel tank you can imagine with the Saito 72 (with Brian's help) HE is planning on writing his results in SN, so I have promised not to steal his thunder and blab.

For all reading this - it's a curiosity as to *why* it
>works, but going from any hard tank with regular-size
>plumbing, to a clunk with supplied plumbing, is the most
>dramatic improvement you can make for a 4-stroke. Gutless
>overheads go away. Second most important is to get rid of
>the stock carb and get a Pat Johnston venturi setup
>(available from UHP). I point this out because it directly
>contradicts some of the "early adopters" observations, and a
>lot of that advice is still floating around.
>
> Brett

I can say with quite a bit of confidence that the venturi arrangement draws fuel much better than a fixed carb at the same opening and RPM. I have done some bench testing by making the engine draw fuel various distances uphill, and the carb is not as good.

BTW there is a very definite point where the engine will no longer richen. Once the fuel has been dropped low enough below the level of the spraybar no matter how much you open the needle the engine is lean. THE ENGINE IS STARVING and is unable to draw the fuel from the needle valve any longer.

> It would be an interesting experiment. But the clunk tank
>seems to address whatever issue it is all by itself.

Sort of.

POP QUIZ FOUR STROKERS:

Everyone seems to like my fuel YS 20/20. Why?

>Sort of.
>
>POP QUIZ FOUR STROKERS:
>
>Everyone seems to like my fuel YS 20/20. Why?

I can guess why you *think* it's better in relation to this thread - lower viscosity? That be unrelated, or only partially true, however.

>
>>Sort of.
>>
>>POP QUIZ FOUR STROKERS:
>>
>>Everyone seems to like my fuel YS 20/20. Why?
>
> I can guess why you *think* it's better in relation to
>this thread - lower viscosity? That be unrelated, or only
>partially true, however.
>
> Brett

Really... I am all ears.

LAST EDITED ON Dec-15-03 AT 05:11 PM (CST)
 
Some comments.
1)The needle valve and the spraybar hole are much smaller than the tubing cross-section, so I find it hard, from a physics perspective to understand how the tubing size can matter that much. If it did, why not just not open the needle more?
2) Yes the 4 stroke "sucks" on one entire stroke out of 4, but I think it is also using less fuel than the 2 stroke (that's why you can use a smaller tank), so I don't think the instantaneous flow is twice that of a 2 stroke--maybe not entirely equal either.
3) I think one item is that a clunk tank with standard venting has less of a differential head than a uniflow tank with the uniflow vent plumbed down to the far end of the tank. What I have gathered over time (from all the posts) is that the 4 stroke seems to have much less tolerance than a 2 stroke for mixture changes. This seems to manfest itself as "lacking guts" in the vertical maneuvers. I interpret this to mean that it is leaning out too much in the climb coupled with the higher load.

Note that I don't dispute the observations that the 4 strokes run better on standard venting, I just don't completely buy the reasons.

For those of you who wish to understand the physics of the uniflow tank, I offer my direct translation from Edme Marriott's original manuscript written in the mid-1600's. "The engine she think she is sucking the fuel from the end of the uniflow tube that is submerged in the fuel."

This is a drawing I did to help understand how a uniflow works.

Larry

Brett wrote:

> I think that the crux of the issue is that the flow is
>not constant, but "bursts". They don't take much fuel, so
>the average flow rate is low, but the actual time it's
>drawing is only a small fraction of the time that it's
>running. I would guess that a two-stroke is close to a
>continuous flow.

Hmmm.. If non-constant flow through the venturi is the problem, how about using a large chamber between the venturi and the intake valve proper (I've heard these referred to as a "plenum", but I won't swear it's the correct term) to smooth out the flow? This was one of the ways to "cheat" at F2D Combat when the 4mm venturi rule came out (along with sintered brass venturis, etc.), but they shut that loophole in a hurry.

I am using Larry's drawing here to try to explain how the uniflow tank works.

[photo not recovered: 3fde4f127b5aa768.jpg]

If you look at the picture on the left, the air above the fuel on a standard vent tank is at atmospheric pressure. So the fuel pressure at the fuel inlet tube is that of atmospheric pressure plus the weight of the fuel. As the fuel is burned down, (middle picture) the weight of the fuel diminishes, and it continues to do so until just before it runs out it where it will be very nearly at atmospheric pressure. Thus the fuel pressure constantly diminishes from being above atmospheric pressure to equalling atmospheric pressure, and the engine grows constantly leaner as the fuel is burned off. I might add that the "weight" of the fuel producing the fuel pressure is multiplied by the "g" force induced by the tethered flight, so it may be three times the actual weight that would be normally felt if it were merely a tank tipped on end.

Looking at the uniflow tank (third picture), we again have the air pressure in the tank plus the weight of the fuel adding together to get total fuel pressure. In this case, however, the tank is not vented directly to the air, and a partial vacuum is formed inside the tank as fuel is consumed. The air inlet is totally submerged, and, with the air in the tube being at atmospheric pressure, no air can escape into the tank until the pressure from the weight of the fuel plus the air pressure in the tank falls to just below atmospheric pressure. Then air is let in until the pressure at the bottom of the fuel head (which is equal to the weight of the fuel plus the pressure of the partial vacuum) is at atmospheric pressure. As fuel is consumed so that the pressure from the weight of the fuel becomes less, air is let into the tank at greater and grerater pressure until, once again, just before the fuel runs out, the air in the tank is at full atmospheric pressure.

Thus in the standard vented tank the fuel pressure is atmospheric pressure plus the weight of the fuel. In the uniflow tank, the fuel pressure is at atmospheric pressure throughout the entire run. This may explain partially why a 4-stroke engine, with its poorer fuel draw, does not like a uniflow tank. And since the 4-stroke doesn't seem to speed up as much as it gets leaner, you don't notice the difference as much in flight.

>I am using Larry's drawing here to try to explain how the
>uniflow tank works.
>

Bravo Leonard! As a card carrying professional physics type person, I will say that I could not have explained it any more clearly. I always just tell people that the air at the uniflow tube is always at atmospheric pressure, and it therefore helps "pin" the fuel inlet pressure to atmospheric, irrespective of fuel column height in the tank. Your explanation makes it clear *how* it does that. Nice job.

The only embellishment I can add is that to truly ensure atmospheric pressure in the uniflow tube, be sure and point the open end forward into the airflow. This will ensure that the flow stalls to zero speed in the uniflow tube, and that there are therefore no Bernoulli effects to change the uniflow vent pressure. In other words, it keeps the uniflow tube from trying to act like a perfume atomizer. And contrary to popular belief, the pressure in the tube will not be higher than atmospheric. It will be exactly at the ambient atmospheric pressure, because that's the pressure at the other points in the fluid, at the same height as the vent inlet, where the air is at rest.

Andy

On uniflow tanks which are open vented, ie, neither muffler nor crankcase pressure, I do not point the uniflow into the airstream. I also run a KK one-way valve on the uniflow to do two things: 1. prevent syphoning, and 2. protect the tank from variations in atmospheric pressure due to wind or whatever. I have one tank which has the uniflow pointed into the airstream. This tank syphons while the engine is running, in spite of the one=way valve. I do not understand it.

Len, Andy, and Jim T,

EXCELLENT description of the uniflow setup. I would only differ on possibly a fraction of a psi, the amount needed to pull air down the vent tube to its 'wet' end. If there is NO pressure difference, how does the slight pressure drop inside the tank shell hold back fuel from flowing continuously?

It is easy to demosntrate that the fuel stops flowing out of the tank on its own if the end of the flex fuel tube height is just right. Most of my own home-built tanks drip from the copper outlet, and not flow continuously, with the flex tube at slightly lower or even with the vent 'wet' end height.(Tank filled, capped, held level with vertex down.) If pressure difference in the tank shell doesn't "support" the fuel's gravity head -- density X height -- I can't see what does.

Minor point, and I'm very happy with the ease of grasping Len's explanation of Larry F's sketches.

>Len, Andy, and Jim T,
>
>EXCELLENT description of the uniflow setup. I would only
>differ on possibly a fraction of a psi, the amount needed to
>pull air down the vent tube to its 'wet' end. If there is NO
>pressure difference, how does the slight pressure drop
>inside the tank shell hold back fuel from flowing
>continuously?

Good point.

I think this may speak to what Brett was saying. Intermittent flow may exacerbate the problems with the 4 stroke since the pumping strokes are further apart (pumping frequency half as fast).

>>Len, Andy, and Jim T,
>>
>>EXCELLENT description of the uniflow setup. I would only
>>differ on possibly a fraction of a psi, the amount needed to
>>pull air down the vent tube to its 'wet' end. If there is NO
>>pressure difference, how does the slight pressure drop
>>inside the tank shell hold back fuel from flowing
>>continuously?
>
>Good point.
>
>I think this may speak to what Brett was saying.
>Intermittent flow may exacerbate the problems with the 4
>stroke since the pumping strokes are further apart (pumping
>frequency half as fast).

Well, I don't claim to have a fully-formed theory. But pulling in fuel for a brief time suggests high flow velocity - and the pressure drop is strongly dependent on the flow velocity.

I'm also not entirely convinced that you don't get "reverse flow" during the exhaust stroke on uniflow. I can easily imagine that the area under the cowl is higher than atmospheric pressure, and the delivery is less than or equal to atmospheric. I see no reason that you can't get air pushed backwards into the lines. Of course, the next time the intake valve opens, some of the already brief time spent sucking in fuel is actually spent sucking air. Opening the needle would only work to a point - because it meters the air flowing backwards, too, and the more you open the needle, the more air gets in.

I wonder how much distance each stroke advances the fuel in typical "medium" silicone tubing... (analysis pending). Does anyone know the intake duration (degrees) of, say, a Saito 72?

It's my observation that everything that you do to improve the fuel delivery (suction, thin oil, muffler pressure, venturi vs. carb) the better it runs. A pump looks better and better all the time.

LAST EDITED ON Dec-17-03 AT 06:31 PM (CST)
 
> It's my observation that everything that you do to
>improve the fuel delivery (suction, thin oil, muffler
>pressure, venturi vs. carb) the better it runs. A pump
>looks better and better all the time.
>
> Brett

Roger. Over and out...

>>>The needle valve and the spraybar hole are much smaller than the tubing cross-section, so I find it hard, from a physics perspective to understand how the tubing size can matter that much. If it did, why not just not open the needle more?<<<

Allan, I think you see it too “static”. I do not think it is about fuel vent, I think it is about uniflow vent, which do not have any NVA.

The pressure depends on acceleration and height. And acceleration is permanently changing, thus uniflow must reflect it.

In steady work, the uniflow air sub pressure over surface of liquid has exactly value of head difference in uniflow tube and in tank. So that sub pressure cancels its difference. But it is true in gravity field. In-air condition is, that the pressure depends on net acceleration, not on gravity only (either prop or centrifugal), so:

1/ if the acceleration is lower (you fly over head, means centrifugal acceleration minus gravity), the sub pressure of air must be lower, thus some extra volume of air must run in to the chamber over liquid. Thicker tube will allow quicker normalizing.

2/ if the acceleration is higher (you fly back down, means only centrifugal acceleration thus plus gravity compared to overhead), the sub pressure of air must be higher, thus some volume of fuel must run to the tube to expand the air over the surface (means reach lower pressure until that amount is burned out). The liquid has more friction, thus thicker tube will also allow quicker normalizing, but there is another problem, thin tube takes less volume and length of filled tube brings some extra fuel pressure. The thicker tube takes more liquid and thus 2x thicker tube with 4x more crossection will at the same acceleration difference make only 1/4 of pressure difference.

The
>path it takes has no real effect on the static pressure.

Quite right, the fuel going down will balance the fuel going up, the same way a siphon hose works. Referance Pascals law.

Giffy

In an earlier thread, it was asked what Fox said to use, tank wise, with its .35.

I have a 50s era instruction phamphlet that covers everything from a .19 to a .59.

Here is the quote:

"The fuel outlet tube and fuel line must have an inside diameter of at least 3/32 of an inch; otherwise the restriction will cause large fluctuations of motor speed. In time corrosion will partially clog the fuel outlet tube. Run a wire in it occasionally to keep it clean".

Snip, snip, cut...

For all reading this - it's a curiosity as to *why* it works, but going from any hard tank with regular-size plumbing, to a clunk with supplied plumbing, is the most dramatic improvement you can make for a 4-stroke. Gutless overheads go away. Second most important is to get rid of the stock carb and get a Pat Johnston venturi setup (available from UHP). I point this out because it directly contradicts some of the "early adopters" observations, and a lot of that advice is still floating around.

Brett

I am bringing this to the fore again, both to emphasize Brett's points and add to them.

The standard "Hard tanks work okay, wire the carb open and fly," advice was bogus back then, is even more bogus today what with there being excellent alternatives available.

What bothers me is that this advice is not just still floating around, the originator is still spreading it far and wide.

Although I shouldn't do so, I'm going to put a fine a point on this issue, noting than in certain cases we all must be very careful when a self-proclaimed sport flier, Bob Zambelli, is issuing advice to the competition-oriented modelers.

Nowhere has this been more true over the last several years than with four-strokes and their proper handling.

The really odd thing about this whole four-stroke deal is how few top-level fliers are using them. Seeing this in Nationals or Team Trials results, one can shrug and say, "Well, they just don't work; or they are too hard to make work well." And then we see the performances of Paul Walker and Gordan Delaney, and they don't have anything super-trick or even high-zoot up their sleeves.

Something is not computing here at all.

I kinda think Brett has identified a big factor in all of this, even though he was polite enough not to say so. At least not in the terms, nor with the specificity, that I have.

Tough. I'm not a very polite person at times...

LAST EDITED ON Dec-16-03 AT 04:42 PM (CST)
 
POP QUIZ FOUR STROKERS:

Everyone seems to like my fuel YS 20/20. Why?

The City Smasher

Brad,

"Your fuel," the YS 20/20 blend, is 1) Not as messy as SIG 10% Champion; 2) Readily available at two local hobby shops; 3) Seems to run fine, even if not demonstrably superior in output to 15% Champion; 4) Has been confirmed by PW to give more than adequate lubrication; 5) Used to be the fuel of choice of several RC Pattern guys I know, still is used by a number of others; 6) You told me to use it and I want to be just like you (Oh, wait--Got my "Walkers" mixed up); 7) Is reasonably priced; 8) No outrageous Haz-Mat shipping charges, see #2; 9) No burned castor on the valves or valve stems; 10) Causes the locals to think I'm a pro; 11) It's not *green,* no offense.

I'm thinkin' this is a trick question. Let me have it, don't spare the pleasantries.

>POP QUIZ FOUR STROKERS:
>
>Everyone seems to like my fuel YS 20/20. Why?
>
>The City Smasher
>
>
>Brad,
>
>"Your fuel," the YS 20/20 blend

Sorry ***(cough)***(cough)*** the fuel I have recommended then...

All good reasons Dirt. Especially the fact that it makes you more cool to use what the "pros" use. I personally like the smell too...

All good reasons indeed, but not the reason that I think makes YS 20/20 (in particular YS 20/20 made by Powermaster) superior for our use to any other blend I ever tried (until recently).

You can think its the nitro, or the all synthetic blend, but I am pretty darn sure that is not it. I have, in fact, made my own fuels from scratch as you have done. I even made a batch of all Klotz 20/20. To say the results were poor would be an understatement. The homemade fuel was definately down on power especially in the verticals and overheads. Why??? What went wrong? It synthetic oil right? The nitro content is the same, right? In fact, Rich's Brew YS 20/20 is a fine blend, but NEVER made the power of Powermaster. Why?

Do you have a syringe?

Take a syringe. Take some Sig 15% Champion and pull a few ounces from a jug. Did you pull hard on the syringe?

Now take some PM YS 20/20 and repeat the same thing. Would you say the force to draw the fuel is ABOUT HALF?????

Powermaster YS 20/20 is a blend of two synthetics. 2/3 of the oil is the same synthetic oil found in GMA and 1/3 is a very low viscosity high lubricity synthetic oil. The oil in YS 20/20 is about HALF the viscosity of the standard "stunt blend" oil package of 1/2 Klotz and 1/2 castor which are both VERY high viscosity compared to many of the more modern oils (just feel test here---I have no lab test data---even though Powermaster has data).

Which would you say would be easier to draw through a small tube under G's?

If you want a real SUPER SHOCKER try the Powermaster heli blend fuel. It is 23% ALL low viscosity oil (the same one found in YS 20/20). In fact, repeat the same syringe test. The heli fuel is like drawing WATER compared to a MILKSHAKE.

Are we making more power or simply NOT STARVING THE ENGINE????

I told you all there was more than one way to skin that cat....

I have been waiting for someone to allude to viscosity ever since the "Eliminating the Variables" thread. No one ever went there, so I guess there it is.

>The standard "Hard tanks work okay, wire the carb open and
>fly," advice was bogus back then, is even more bogus today
>what with there being excellent alternatives available.

I don't know about "bogus". It will run that way. It's just that it runs better other ways. I really just think it's a matter of having had 3-4 years more engineering development on it. Time moves on, and hindsight is 20/20.

Brett,

Contradict me, will ya?

I am sorely tempted to edit "bogus" out of my posting, leaving you hanging.

Probably ought edit Zambelli's name out of it as well. I have since decided my comments were a little too harsh.