# Composite fuel tank question

### Help Support HomeBuiltAirplanes.com:

#### BJC

##### Well-Known Member
HBA Supporter
I suspect that ethanol fuel would turn your Glasair wing to rubber in less than a week.
I know of one Glasair wing that was scrapped due to auto gas with ethanol causing the VE resin to become gummy.

I realize that that is contrary to the common wisdom wrt VE resin.

edit: Would also note that not all VE resin is the same.

BJC

Last edited:

#### rv7charlie

##### Well-Known Member
Hmmm. The thin stuff folks are talking about scares me a bit. You folks had better have absolutely no way to make much pressure in these tanks. Let's lay out some conditions:

You want to know that:
• a blocked vent and a fuel pump drawing from the tank at max vacuum will not collapse the tank enough to cause it to rupture someplace and dump fuel;
• a blocked vent and a transfer pump filling the tank at max transfer pressure can not expand the tank enough to rupture someplace and dump fuel;
• [snipped]
• a full tank at your crashworthiness limits can not be ruptured.
I am sure that you can come up with a few more. Establish the ones that apply to your plane, then check that you have enough tank wall strength and stiffness. Maybe you do or maybe you don't... I like SolidWorks for this stuff.

Billski
The 3 items remaining (above) in that list cannot be met by *any* aluminum small aircraft fuel tank I've ever seen. Or heard about. Or even dreamed about. And while I have little experience with 'glass' tanks, I'd be pretty surprised if they're met by any existing 'glass' tanks, either.

#### cluttonfred

##### Well-Known Member
HBA Supporter
On the VE resin and alcohol experiments, that is disturbing. It sounds like some sort of alcohol-resistant fuel tank sealant would be a good idea no matter what type of resin you're using.

#### Geraldc

##### Well-Known Member
Hmmm. The thin stuff folks are talking about scares me a bit. You folks had better have absolutely no way to make much pressure in these tanks. Let's lay out some conditions:

You want to know that:
• a blocked vent and a fuel pump drawing from the tank at max vacuum will not collapse the tank enough to cause it to rupture someplace and dump fuel;
• a blocked vent and a transfer pump filling the tank at max transfer pressure can not expand the tank enough to rupture someplace and dump fuel;
• a full tank and max g in positive and negative directions will not overload the tank with head pressure through tank height;
• a full tank in a spin will not overload the tank ends through centrifugal head pressure through tank spanwise dimension;
• a full tank with aero load drawing on the outside skin of the tank can not expand the tank enough to rupture someplace and dump fuel;
• a full tank at your crashworthiness limits can not be ruptured.
I am sure that you can come up with a few more. Establish the ones that apply to your plane, then check that you have enough tank wall strength and stiffness. Maybe you do or maybe you don't... I like SolidWorks for this stuff.

Billski
In my plane the surrounding structures support the fuel tank. I have not built the tank yet so I might change to an aluminium one.
As for a blocked vent I will try this and see the result.

#### Pops

##### Well-Known Member
Log Member
On the first wings for the SSSC I built fiberglass wing fuel tanks using this epoxy. https://www.aircraftspruce.com/pages/cm/resin/polyepoxy.php

With the SSSC, I got some auto fuel from a pump that was not labeled for ethanol as required by law. Topped off the right fuel tank and a few days latter was going flying. Noticed that I had a small fuel leak on the right tank and small white things floating in the fuel. I stopped using the right tank and since I was wanting larger fuel tanks for more range, I built a new set of wings with larger aluminum fuel tanks.
My first set of fiberglass fuel tanks was in 1980 and have built several, and never had any problems until they started putting ethanol in the auto fuel.

In DC, money will do wonders.

#### rv7charlie

##### Well-Known Member
Aluminum tanks (at least any that are light enough to be used in a small a/c) will collapse if the vent is blocked and fuel is pumped out at the pressure differential that a typical fuel pump can achieve. 14psi (atmospheric over a vacuum) is an incredible amount of force, when spread across a 6-8" square area.

A typical 1/4" vent line that is 4-6 feet long (common on 2 seat RVs) will supply enough back pressure that a high-flow transfer pump can create over 10 psi of *internal* tank pressure if the tank is already full, *even if the vent isn't blocked*. Again, 10 psi will destroy a typical a/c tank.

Charlie

#### cluttonfred

##### Well-Known Member
HBA Supporter
I checked one online supplier today who wanted about $400 for one 4' x 8' x 1/8" extruded vinylester/fiberglass sheet cut into four 4' x 2' sections plus another$75 for shipping, so I contacted another for a quote and that one wanted $450 +$100. That's just for the one sheet, I would still need resin and FG cloth for the curved decking/tank top and FG tape to put it all together. :-( So it sounds I'd end up making the sheet myself on a plate of glass.

If going down that road, I wonder if I'd be better off making a foam plug (less the 3/8" spacing above the deck per the plans) on which to form the bottom and ends of the tank including the flanges for gluing to the tank top/decking. I could see tacking the "funnel" on the main plug to form the sump at the same time. Then I'd add the internal baffles and glue it all together, at least, that's the theory.

Last edited:

#### TFF

##### Well-Known Member
The problem with using the sheet is all joints are mechanical bonds. If you really don’t have the ability to do layups, whatever the reason is, it’s a substitute. Using a mold lets the corners and edges be part of the rest of the tank instead of scabbed on. With a mold, making the mold with one bond line at the top gives the best shot of low leak probably. A slightly fancy mold could have all corners integrated and just a top plate. Maybe removable. May not make the prettiest tank from the outside, but making the mold like a foam ice cooler and glass the inside. Carve back foam to x dimension, then glass the outside. Make the top similar.

#### cluttonfred

##### Well-Known Member
HBA Supporter
Well, in this case the fuselage top decking forms the top and sides of the tank in one continuous piece that goes from the top of the firewall to the top of the forward spar bulkhead/instrument panel. So I am not sure the "cooler" approach is the way to go here. Over in the riveted aluminum alternative thread (link in the first post in this thread), FritzW suggested fore-and-aft rather than side-to-sided bulkhead/baffles and I could see that working well.

#### dog

##### Well-Known Member
Earlier in another fuel tank thread,there was a link to a company making military helocopter fuel tanks,that company was very generous with pictures,and descriptions,and details of testing.
They were making bladders for in there tanks.
Now I havnt done much work with bonding butyl
or rubber but it isnt rocket surgery.
So work up a process for that and then fill the bladder and lay up the composite structure around
it.Of course I like like learning new processes and that part of that is spending three days building almost perfect scrap.Test tanks could be small.

#### cluttonfred

##### Well-Known Member
HBA Supporter
I hear you, dog, nobody likes to think about an in-flight or post-crash fire, but sometimes you just have to take reasonable precautions and hope for the best. The late, great Phil Bolger (speaking about boat safety) said that carried too far you “end up by locking everybody in padded cells where they'll be quite safe.” There are proven aviation, auto racing, and military vehicle solutions out there, but they are often extremely expensive and add substantial weight. At this point, I’m willing to go with a single-structure tank like just about every other light aircraft out there.

#### dog

##### Well-Known Member
I hear you, dog, nobody likes to think about an in-flight or post-crash fire, but sometimes you just have to take reasonable precautions and hope for the best. The late, great Phil Bolger (speaking about boat safety) said that carried too far you “end up by locking everybody in padded cells where they'll be quite safe.” There are proven aviation, auto racing, and military vehicle solutions out there, but they are often extremely expensive and add substantial weight. At this point, I’m willing to go with a single-structure tank like just about every other light aircraft out there.[/QUOT

I will also likely stay with the aluminum tank that came with my Christavia project.
That said ,the wheels never stop turning in my head
and I have thought up an idea: that there might be a way to inject something into a tank of fuel in an emergency that would dramaticaly lower the flamability and or flash point of gasoline.
One of those mini co2 canisters with an(anti)
charge of ?? something.Must be stuff that is
poison to gas,an anti oxident,wonder if vitamin C
would work.
Say a prayer for my eyebrows.

#### cluttonfred

##### Well-Known Member
HBA Supporter
Or you could just install an automatic fire bottle.... ;-)

#### wsimpso1

##### Super Moderator
Staff member
Log Member
The 3 items remaining (above) in that list cannot be met by *any* aluminum small aircraft fuel tank I've ever seen. Or heard about. Or even dreamed about. And while I have little experience with 'glass' tanks, I'd be pretty surprised if they're met by any existing 'glass' tanks, either.
I wonder what pressures you are thinking of and what your criteria are. Design information, test reports, etcetra would be nice backup. I do suspect a failure of imagination here. I am competent and experienced in panel and pressure vessel calculations as well as composite material mechanics and have been doing FEA since the mid 1980's. I prefer to KNOW that my tanks will stand the system failure modes we can expect without dumping the fuel in the cockpit. So, I designed my tanks to stand the things I can reasonably expect that they could see.

I have designed and am building a header tank for my ship, 8.6 gallons, about the same width and hieght, less fore-aft length. It is 2 BID on each side of 1/4" H45 Divinycel. I added baffles until it passed all of my checks. I could have also made the grade with thicker cores and/or thicker facings, but the design cited above was the lightest that passed the analytical checks. The design currently has one internal vertical baffle running athwart-ships and two running fore-aft, all running from top to bottom of the tank and with high and low spots cut for liquid and vapor communication. My transfer pump to the tank is a Facet cube pump with a dry lift capability on the suction side of 36" and supply pressure of 4.5-6 psi. According to both hand calculations and the FEA in SolidWorks, the entire tank will stand full pump pressure with a FOS well above 2.0, and so appears to meet the blocked vent and pressurization from the transfer pump. In the FAA 19 g up and back crash pulse, it has a factor of safety somewhat above 2.0 here as well. It is also quite secure in the 26g forward crash pulse. I do not know the vacuum capability of my high pressure fuel pump, but I would doubt it would draw more than half an atmosphere - these things were never intended as vacuum pumps and I can not believe a vane pump running in gasoline would make one. So, the tank stresses under this vacuum case would be about the same as the pressurization case. This tank appears to stand the set of failure modes I posited with the pump set I am planning.

I have since taken that model and changed the material to 0.050" 5052-O and it is close on stresses - it will yield in places. I would have to add a couple more baffles to make it stand all of the load cases without yielding. Even with yielding, this tank may preclude rupture as is. It appears that a safe tank in soft aluminum could be created to meet all of the criteria I set out and with the pumps I have in mind. It can be done in this scale - I strongly suspect that a larger tank could be schemed up that would also survive this sort of abuse.

Billski

Last edited:

#### wsimpso1

##### Super Moderator
Staff member
Log Member
Aluminum tanks (at least any that are light enough to be used in a small a/c) will collapse if the vent is blocked and fuel is pumped out at the pressure differential that a typical fuel pump can achieve. 14psi (atmospheric over a vacuum) is an incredible amount of force, when spread across a 6-8" square area.
I have three different pumps in my shop designed as vacuum pumps. All are piston type, one a oil bath, the other dry pumps. No one single stage vacuum pump in my shop will exceed about 24" Hg or about 12 psi. Go to Facet-Purolator's website, and their cube pumps will do a dry lift of 36" to 48" of fuel dependent upon model chosen. Lifting gasoline 36" is a touch less than 1 psi of suction. Their DuraLift pumps, designed specifically to self-prime with pumps well above the tank can lift 120" of fuel, that is a touch above 3 psi of suction. I suspect that these common pumps drawing from the tank are not capable of the vacuum RV7 is suggesting as these numbers are way lower than the 14 psi mentioned.

I am be concerned that tanks be designed so that they will not dump fuel on people and engines and electrical systems. This means we should preclude rupture. I never said that you have to preclude tank yielding or collapse, although rupture is easier to prevent if the tank does not yield or collapse.

A typical 1/4" vent line that is 4-6 feet long (common on 2 seat RVs) will supply enough back pressure that a high-flow transfer pump can create over 10 psi of *internal* tank pressure if the tank is already full, *even if the vent isn't blocked*. Again, 10 psi will destroy a typical a/c tank.
I would sure like to see a citation on this one too.

Pressure drop in the tank due to fuel coming out and air drag to fill in is easily calculated and second year engineering students get taught this in the introductory fluids class. Just to get some numbers, let's say this typical homebuilt has a takeoff fuel flow of 15 gph. 15 gallon/hour * 231 in^3/gallon * 1 Hour/3600 seconds, that is 0.96 cubic inches of gas leaving the tank every second. A 1/4" ID tube has a cross section area of .049 In^2, so the air flow in is at 19.6 inches per second, or about 1.6 ft/s. Go to Introduction to Fluid Mechanics by Fox and MacDonald. I always ran this easier in metric, so rho is 1.225 kg.m^3, mu is 1.8e-5 N*s/m^2, V is 0.50 m/s, D is 0.00635m, L is 1.83m. I get Re number of 215, so the flow in the tube is deeply laminar - we do not need to worry over tube roughness. Friction factor comes out at 0.297, head loss is 10.6 m. Air is low density stuff, pressure drop in the tube is 128 Pa. One atmosphere is 101,600 Pa, so the head loss and resulting suction is a tiny fraction of atmospheric pressure 0.018 psi. Nope, our suction in the tank due to fuel being pulled out and friction on the air in the vent line is pretty darned close to zero when using the standard engineering methods.

Some might be concerned about recovering ram pressure in our tanks from the vent line. At 60 knots, 100% pressure recovery would add 0.002378 slug/ft^3/2*(101.3 ft/s)^2 = 12.2 lb/ft^2 = 0.085 psi. At 240 knots, it is still only 1.36 psi.

I would not worry over pressure in the tank in either direction while the vents are working. But if your vent is solidly blocked, your pumps can do things to the tank. I would sure want to make sure that the tank does not rupture under those circumstances and spill fuel where it can be easily lit or asphyxiate the crew.

Billski

Last edited:

#### JohnBouyea

##### Member
Russell, Question: Is it possible that chemicals other than the labeled Ethanol caused the problem? I ask because a friend did a similar test after experiencing damage to his tanks. Read: http://jeffsplanes.com/Cub/MOGAS/mogas.html

#### Russell

##### Active Member
John, I cannot say with certainty that ethanol is the villain here. The damage in your friend’s tank looks very similar to what I have seen in other vinyIester tanks that have had ethanol in them. I suppose the only way to find out for sure is to test with all the different chemicals in your gas against all the components of the tank.

I suggest to always test before adding something new to your tank. The test for ethanol is a simple 60 second test. If done carefully you can even determine the percent of ethanol the sample contains. If building a new tank test the products that will be used in the construction.
Russell Sherwood

#### JohnBouyea

##### Member
Hi Russell. Is it possible that something other than the labelled Ethanol caused the problem? The reason i ask is a friend had issues with chemical additives in the "alcohol-free MoGas he was running. You can read about that and his testing here; http://jeffsplanes.com/Cub/MOGAS/mogas.html

#### Hot Wings

##### Well-Known Member
HBA Supporter
Log Member
Earlier in another fuel tank thread,there was a link to a company making military helocopter fuel tanks,that company was very generous with pictures,and descriptions,and details of testing.
As wsimpso1 noted above rupture is the ultimate failure mode, not deformation. In keeping with this thought I'd really like to find the video I referenced in this old post of mine:
The process was simple and well within the abilities of the average home-builder. The materials used and process controls are the only variables that need to be worked out before we can start building really robust fuel tanks.

Edit:
For entertainment value only. I find it interesting that the back stop for the flaming arrow is what appears to be a fuel storage tank.

Last edited:

2