Tanks 101

[Dick Fowler]

There are lots of discussions about fuel tank venting and the subject of uniflow always brings lots of questions and head scratching. I'm a visual kind of guy so I deal best with pictures and drawings. Here's my take on how tank venting works.

[Howard Rush]

Close.  I guess the second picture depends on how you define muffler pressure.  If it's the increment over atmospheric, then the second picture is OK.  The third picture contains an extra Pm, and the little line drawn across the tube even with the fuel level might be misleading.  The tube is full of air. 

[Dick Fowler]

Thanks Howard for picking up on the points you mentioned (cut and paste can be dangerous). I corrected the drawing. I consider that muffler pressure is additive to atmospheric pressure ( in absolute terms it is) simply because other parts of the system are exposed to atmospheric pressure as well. It would be difficult to have fuel flow if you only think of the top of the tank having a few inches of water pressure over the fuel while the venturi inlet is seeing about 34 ft of water pressure.

[Keith Spriggs]

RE: Drawing number three.

I know I am covering some ground that was discussed only a short time ago, but please humor an old man.

Is this the critical factors for a uniflow tank?

Uniflow line same elevation as fuel outlet line.
Vent capped
Uniflow line exposed to the atmosphere, I assume pointing into the air stream or at least positioned so the airstream does not create a vacumn in the tank.
I assume that the opening in the tank from the uniflow line should be far enough from the outlet line opening that there are no bubbles that reach the outlet line.

DO I have this right?   I am just getting back into C/L and when I quit before we were using hollow rocks for tanks.

[Dick Fowler]

Keith - Is this the critical factors for a uniflow tank?

Keith - Uniflow line same elevation as fuel outlet line.

Not necessarily. If you consider Fig. C, the pressure at the fuel outlet (Pout) =  atmospheric pressure (Pa) + fuel outlet head pressure (Hp) - Uniflow head pressure (Hu).

If the uniflow and fuel outlet lines are the same elevation then Hp = Hu and Hp - Hu = 0  so  Pout = Pa + Hp - Hu = Pa + 0 so the outlet pressure is equal to Pa only ( atmospheric pressure) for all fuel levels.

If they aren't the same elevation  Hp -  Hu = K ( some constant value) so then  Pout = Pa + Hp - Hu = Pa + K so now the outlet pressure equals atmospheric pressure minus some constant for all fuel levels until the fuel level reaches the uniflow outlet. At that point the outlet pressure behaves the same as a standard atmospheric vented tank.

Keith - Vent capped?

Yes. If not then the tank will pull vent air through the uncapped vent ( less resistance than trying to pull air through uniflow vent and the fuel)  and it becomes a standard vented tank again.

Keith - Uniflow line exposed to the atmosphere, I assume pointing into the air stream or at least positioned so the airstream does not create a vacumn in the tank.

That seems to be an entire subject unto itself. Everyone has a favorite that they swear works best. I have pointed them into the prop blast and even inside the cowl. Both seem to work Ok for me.... your mileage may vary!


Keith - I assume that the opening in the tank from the uniflow line should be far enough from the outlet line opening that there are no bubbles that reach the outlet line.

I don't think the bubbles can usually get in the outlet. The bubbles want to shoot upwards. I have noticed that Fox Superfuel tends to retain the bubbles in the tank more so than Powermaster. A smidge of surfactant helps.

[L0U CRANE]

Dick, nicely done, and very clear!

We do need to keep in mind that up and down, as experienced by the fuel in the tank, is different in flight. A much knocked around number for the g load due to circular flight is about 3g. Gravity always has its 1.0g, towards the center of the planet.

So, out three units (or so), and down one unit, means the fuel feels about 3.16g slanted at about 20° angle out and down in level flight. The fuel surface lies perpendicular to the "local gravity."

Maneuvers complicate this a bit, since they have the centrifugal force that comes with their shapes to add in to find how much the actual local g load is, and which way it points. Could be a fun project, if you are so minded, to work out loads at several key points in the pattern ...

How this affects tanks?

The original open, over and under vented "stunt" tanks ran leaner throughout the flight because of the decreasing fuel head, as in Fig 1. We had to find a setup that was lean enough on the front half of a flight, but that wouldn't starve off toward the end of the fuel. The key factor was a setting that could do that, somehow, dependably.

Uniflow tanks, as your Fig 3 makes very clear, only have a "fuel head" factor of the "height" from the vent air entry inside the tank to the fuel pickup in the direction of the local g load. Two things happen:


-- The setting is so much less affected by the change of fuel height inside the tank that we consider uniflow tanks to give a constant setting throughout the flight. (It isn't, really, but is so close who cares?)

-- The "height" across the angle of local g becomes more important. Did anyone really care about "tank height" way back when, compared to how much attention it gets nowadays? Oh, yeah, if it was wa-a-ay off, bad things happened. But not like needing to fine tune to 1/32" or so to get it perfectly even both ways.

Thanks again for the clarity of the diagrams and your descriptions of what's happening! It doesn't hurt anyone to understand more about what we do with these models. I hope a great many will see and grow from your post!

[Howard Rush]

I got to wondering about the relationship between ram air pressure and fuel head.  It's interesting.  I did some ciphering, but there may be some errors.  You guys should check it.  Ram air pressure = dynamic pressure, I think, = 1/2*rhoair*V^2, where "rhoair" is the Greek letter rho with the subscript "air"  (If I had more imagination, I could have called it Y or something), and V is the airspeed.  Assume no wind.  Fuel pressure at the outside of the tank for fuel height h = h*rhofuel*V^2/r, where r = flying circle radius.  Here's the cool part.  When you solve for the fuel height to balance the ram pressure, the speeds cancel, as they do for loop radius-- another matter.  So equilibrium h = (rhoair*r)/(rhofuel*2).  For 70-foot lines, sea-level standard air (.002377 slugs/cubic ft.), and 10%-nitro stunt fuel (1.707 slugs/cubic ft.), h=0.58 inch.

What does fuel vapor pressure do, if anything?

[Keith Spriggs]

Thanks to everybody that contributed to this thread. You have all helped this old hillbilly understand them better. I wish I had known you guys when I was having trouble getting my still to work.

Relax "Big Brother Revenoorers" I was just kidding.

[Dick Fowler]

Howard - I got to wondering about the relationship between ram air pressure and fuel head.  It's interesting.  I did some ciphering, but there may be some errors.  You guys should check it.  Ram air pressure = dynamic pressure, I think, = 1/2*rhoair*V^2, where "rhoair" is the Greek letter rho with the subscript "air"  (If I had more imagination, I could have called it Y or something), and V is the airspeed.  Assume no wind.

Looks OK so far.

Howard - Fuel pressure at the outside of the tank for fuel height h = h*rhofuel*V^2/r, where r = flying circle radius.

I think you haven’t accounted for the acceleration due to gravity in your formula. You need a more general formula, which uses the resultant vector of both the centripetal acceleration  and gravitational  acceleration to make your calculation. Lou pretty much pinned it down in his post above. I don’t see the gravitational acceleration component in your calculation.

Also the configuration of the tank will have a significant impact on the value of h. The fluid (fuel in our case) changes shape under acceleration so it can be very different for different tank shapes. Big disclaimer – accelerating liquids in containers is  very complex and I don’t claim to understand all I know about it!!!

Howard - Here’s the cool part.  When you solve for the fuel height to balance the ram pressure, the speeds cancel, as they do for loop radius-- another matter.  So equilibrium h = (rhoair*r)/(rhofuel*2).  For 70-foot lines, sea-level standard air (.002377 slugs/cubic ft.), and 10%-nitro stunt fuel (1.707 slugs/cubic ft.), h=0.58 inch.

Ok… but I guess I’m not sure why this is important. In a uniflow tank, the ram air component adds pressure over the fuel in the tank after launch which would probably increase fuel flow to the engine making it a bit richer, but the uniflow concept negates all of the head pressure variations due to positional changes if the uniflow vent is close to the fuel outlet.

Howard - What does fuel vapor pressure do, if anything?

I don’t think it matters. This is usually confined to close containers but our tanks are vented so IMHO it isn’t a factor.

This whole discussion skirts probably the biggest variation in the fuel system. The constant change in spatial relationship between the fuel tube end in  the tank and the spraybar. Makes my head hurt!

[Howard Rush]

I looked at your third picture again and was confused as to whether the H's are heights or pressures.  the .58" is a distance, not a pressure.  You're right, I omitted gravity.  That .58" would only be on the spanwise component, and would be the same for any tank shape.   Mind you, gravity and maneuvering accelerations act every which a way. 

Now I'm wondering if the uniflow regulation starts as soon as the airplane is up to speed, or does the tank level have to drop .58" first?  How does the position and outlet pressure at the engine affect this?  It might be fun to put the MEMS pressure sensor in a T on the filler tube and record the pressure in the cavity over the fuel during flight.  One could calculate the dynamics for a stunt pattern and compare the two.  Then there's Bob Reeves's supercool tank video.

This is interesting, but I don't know if there's a problem to solve.  My uniflow tank holds lap time just fine.  If lap times start changing, I only need restore the system to the configuration that's known to be good-- usually by fixing a leak somewhere. 

Your mentioning the relationship between the end of the fuel tube (or the end of the uniflow tube) in the tank and the spraybar made me wonder if an engine would run richer in maneuvers if you increase the flap/elevator ratio. 

[Dick Fowler]

Your mentioning the relationship between the end of the fuel tube (or the end of the uniflow tube) in the tank and the spraybar made me wonder if an engine would run richer in maneuvers if you increase the flap/elevator ratio.

I wonder about that myself. It could be that it does if the bottom only of the flap hinge line is taped. I'm not sure if both top and bottom are taped.


PS -  h is the height that would be used to calculate head pressure, and it seems to me that h is determined by the tank and Ra (the resultant acceleration vector)  "Local gravity" as Lou put it.... great term Lou, mind if I borrow it?

[L0U CRANE]

To the very informative discussion from Reply #7 on...

Stray thoughts:

-- For some engines set fairly rich in 4/2 mode, on a briskly windy day there is noticeable richening of the setting as the model crosses the upwind half of the lap... This can be reduced, but not eliminated, by sleeving the forward facing uniflow entry vent to a smaller ID.

-- Vapor pressure does not seem to make problems, agreed. But we're presuming standard glow fuels, right? I also fly diesels, at times, and just recently have come to believe that the greater volatility of the ether fraction of the fuel and uniflow venting are not nice to each other. I've had problems gettng a diesel - i.e., an engine that supposedly can be set to run the same, regardless,  by prop, fuel blend, compression and mixture needle settings - to run consistently and well with a uniflow tank. The heat and vibration conditions, apparently, get the ether "all shook up."

Jim T.'s solution is a one-way valve to prevent the gasifying ether from pumping fuel out the uniflow tube. For him and his ST 15D it works well. I've tried similar, and had less happy results. Recently, in benching a diesel, I saw muchos bubbles in the fuel line while the tank was rigged as uniflow. Shut it down, reconnected the plumbing so that it was single, free-venting type, and Vwah-lah - no bubbles, and the diesel didn't care about a static fuel head change from level to the NVA to over 2" lower - RPM crept up only a hundred or so over 6 minutes!
 
CONCLUSION: IF vapor pressure in a uniflow vented tank becomes a problem, it will "relieve itself" out the uniflow entry tube.

-- Another thought to keep in mind: The hydraulic layout for the tank, fuel-line and NVA should be viewed in regard to the resultant ("local gravity") direction. The fuel in the length of the fuel line is pretty much at the same "height" from the pickup opening until it turns "up" to the fuel nipple. Yeah, there're pipe friction losses, but they'd be there in a "static tank" condition, too, so they might well cancel out of consideration.

I'll read over these fine posts a few more times, f'r sh'r... Thanks!

 

[minnesotamodeler]

Your mentioning the relationship between the end of the fuel tube (or the end of the uniflow tube) in the tank and the spraybar made me wonder if an engine would run richer in maneuvers if you increase the flap/elevator ratio.

I wonder about that myself. It could be that it does if the bottom only of the flap hinge line is taped. I'm not sure if both top and bottom are taped.

I was doing pretty good until this.  Now you're just getting silly.  Or else I'm an idiot.

--Ray

[Howard Rush]

My guess is that Dick thought I was kidding about the flap/elevator ratio.

[Dick Fowler]

My guess is that Dick thought I was kidding about the flap/elevator ratio.

Come on Howard... you have been known to "jest" on occasion. My comment was a bit tongue in cheek. Maybe ( and highly probable) that I don't understand what you are suggesting... assuming you are serious.

The way I see it is the required forces to make a same radius turn for a given airplane are pretty much the same and all produced by lift. If I understand this aerodynamics stuff correctly, changing the flap ratios will have an effect on the attitude of the model with more flap allowing the nose to hang a bit more outside of the turning radius of the CG of the model. So if the turn forces are the same for all configurations then in my mind the fuel levels and head pressures are different only by the subtle change in attitude ( about 3 deg diff. on an average stunt wing with 15 deg and 30 deg. comparative deflections) with a  very small corresponding change in head pressure. I don't know if it's big enough to detect. Empirically, when I've been jacking flap/elevator ratios around I haven't seen any change in engine sounds etc., but then again I'm certainly not a world class stunt guy by any means.

So if you were serious maybe you could explain. Thanks.

I like the elapsed time picture of the corner. It tells a bunch. Think he could use a bit more flap?

[Howard Rush]

That's what I was thinking.  The change in angle of attack would make a different amount of head,  hence fuel flow. at the engine. 

[Dick Fowler]

 Lou, your experience made me think some more about the effects of vapor pressure in a "semi closed" uniflow tank and I think I've changed my mind. The ether has about 6 times the evaporation rate of methyl alcohol and nearly 4.5 times the vapor pressure. Ether's vapor pressure is more than 1/2 Atmosphere ( 8 p.s.i.)!

Glow engines probably consumes the fuel at a rate that keeps up with the phase change over the fuel. Diesels have two problems... significantly more volatile fuel and low consumption. With the lower consumption rate of a four stroke, I wonder if some of the four stroke uniflow problems might  be vapor pressure related. Muffler pressure would help to reduce the evaporation rate of the methyl alcohol in that situation but standard venting seems to work better (and  that makes sense under the above assumptions).

You made a comment about a restriction helping but not eliminating the engine going a bit rich entering the upwind quadrant. Do you think the engine unloading at that point might be why you can't get rid of all of it?