Homebuilt Fuel Tank ("Cell") Design and Fabrication

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

SVSUSteve

Well-Known Member
Joined
Aug 20, 2007
Messages
3,904
Location
Evansville, Indiana
Just in case if you decide to build ballistic proof tanks :gig: Woven fabrics for ballistic protection

LOL Don't tempt me. Although, impregnating ballistic fabrics often severely reduces or complete removes their ballistic protective properties. ;)

That said we've joked about making some more tank wall samples in a year or so (once we have a little more time to spare) and using them for target practice with a .40 S&W chambered pistol just to see how well they would perform (to use the American colloquial expression "just for ****s and giggles"). Given some of the properties with regards to puncture even by a sharp object of the elastomer we used in our current design, I would honestly not be shocked if the bullets would not penetrate.
 

Mac790

Well-Known Member
Joined
Jun 22, 2008
Messages
1,529
Location
Poznan, Poland
LOL Don't tempt me.
It's not as crazy as it might seems:), have you seen this thread Bullet Holes

Although, impregnating ballistic fabrics often severely reduces or complete removes their ballistic protective properties. ;)
Do you have any paper on it? I've seen something contrary to that.

This study demonstrates that the ballistic penetration resistance of kevlar, nylon, and ramie fabrics is enhanced by impregnation of the fabric with the compatible resin. Impregnated resin fabrics composites are shown to provide superior ballistic protection as compared with samples stacks of neat fabrics.
source http://isjd.pdii.lipi.go.id/admin/jurnal/32109714_0126-1533.pdf

I have more papers that claims same.

Seb
 

autoreply

Super Moderator
Staff member
Joined
Jul 7, 2009
Messages
10,792
Location
Rotterdam, Netherlands
Kevlar is a downright bad idea for any tank. Fuel will always creep through and trash any organic fiber (like Kevlar) in a short while, carbon and glass will do just fine there.

For energy absorption, "brittleness" or "rigidity" are meaningless terms. What is relevant is how much energy (impact due to crash loads) you can absord. A corrugated tank with carbon/epoxy is pretty optimal for that, energy absorption per pound. Rubber and similar materials are absolutely horrible at absorbing energy.
 

SVSUSteve

Well-Known Member
Joined
Aug 20, 2007
Messages
3,904
Location
Evansville, Indiana
I have more papers that claims same.

I was going with what I was told by friends who work for the US Army's soldier protection lab. They are the guys that mess around with body armor and vehicle protection so I would assume they know a thing or two about it. It depends largely on what you are impregnating the fabric with and according to the folks I talked to most of the epoxies we use traditionally in homebuilt aircraft it is likely you'll reduce the penetration resistance because you're hindering the ability of the fabric to respond dynamically ("flex") to the impact. However, I am always willing to have my mind changed so thanks for the paper!

A corrugated tank with carbon/epoxy is pretty optimal for that, energy absorption per pound.

Go take a look at some of the TU-Delfft studies on sine-wave/corrugated beams for energy absorption and think about how that would behave for a tank. It fails catastrophically and would probably not be a great idea for a wall where you want to keep something in or out. For a subfloor, it's sometimes hard to beat.

Kevlar is a downright bad idea for any tank. Fuel will always creep through and trash any organic fiber (like Kevlar) in a short while, carbon and glass will do just fine there.

I would argue that S-glass and ballistic nylon would be a much better choice than CF. In a fuel tank you're not looking so much for energy absorption as much as you are trying to minimize anything penetrating it or tearing it.
 

autoreply

Super Moderator
Staff member
Joined
Jul 7, 2009
Messages
10,792
Location
Rotterdam, Netherlands
Go take a look at some of the TU-Delfft studies on sine-wave/corrugated beams for energy absorption and think about how that would behave for a tank.
Wtf?
A beam and a tank (loaded in pressure) are like black and white, like night and day.
I would argue that S-glass and ballistic nylon would be a much better choice than CF. In a fuel tank you're not looking so much for energy absorption as much as you are trying to minimize anything penetrating it or tearing it.
This is homebuiltairplanes.com Not militarychoppersthatgetshotatwithanakfourtysevenbutwherethetaxpayercanaffordanybill.com

If anything manages to penetrate our fuselage or wing at speed, the tank will burst anyhow. No, military helicopters won't, not because they're so brilliant, but simply because they don't typically whack a wing into trees or fences at 50 kts with a load of gasoline. Drop a (full...) tank from 100 ft at a rigid boom barrier and see what happens...
Simply keeping the fuel tank from rupturing from impact is hard enough, without going to use unobtatium or materials that weaken in months if not weeks. Theorizing around is a useless venture if that very theory is just not practical. Thát in the end is what HBA is about.
 

SVSUSteve

Well-Known Member
Joined
Aug 20, 2007
Messages
3,904
Location
Evansville, Indiana
Drop a (full...) tank from 100 ft at a rigid boom barrier and see what happens...

Have you not read my posts about the testing we have done on our fuel tank design? Granted, the highest drop was roughly 80 to 90 feet but the point stands that both of those tests (80-90 feet and 100 ft) are well above even the US Army CSDG crashworthy fuel tank testing standards. In case, anyone is wondering the actual military standards are:
-Bladder drop height without rupture/spillage (full) MIL-T-27422B US 751 standard: 65 feet
-Constant rate tear
Parallel/warp 400 ft/lb
45-degree warp 400 ft/lb
-Tensile strength
Warp 1717 lb (based on safety cell US-756)
Fill 1128 lb (based on safety cell US-756)
-Impact penetration
5-lb chisel
Parallel/warp 15 feet (MIL-T-27422B US751 standard)
45-degree warp 15 feet (MIL-T-27422B US751 standard)
Screwdriver 370.5 lb (safety cell US-756)


In our series, there were drop tests of a half-full tank and a full tank configuration on to a barrier (the concrete "Jersey" barrier was used to simulate impact with an absolutely unyielding surface), a horizontally-oriented telephone pole (to more accurately simulate a tree impact) and onto a flat surface (concrete parking lot). We also ran tests with a smaller series of tanks (same wall thicknesses but only a five or ten gallon volume rather than the 25 and 50 gallon variants in other tests) involving dropping them onto various penetrating items including a 5-lb wood splitting chisel (to simulate a metal item) and sharpened oak stakes (think the "punji" stakes from the Vietnam War minus the coating of human excrement; choosen to simulate impacting a broken tree) to validate the behaviors of various wall configurations including various combinations of Kevlar, CF, e-glass, s-glass, ballistic nylon both in traditional woven material and a harder to obtain "nylon felt" as well as some other reinforcements and several different resin options (vinylester and epoxy) and elastomer liner materials. We probably dropped or purposefully punctured over two dozen tank mock-ups in the course of our testing in preparation for the patent application we are working on. Once we get the patent issued, we are actually planning on selling the plans to the homebuilt community simply because that's what it's designed for.

If anything manages to penetrate our fuselage or wing at speed, the tank will burst anyhow. No, military helicopters won't, not because they're so brilliant, but simply because they don't typically whack a wing into trees or fences at 50 kts with a load of gasoline.

It's funny you say that because with a reasonable design capable of being put together by a homebuilder with minimal experience in composite construction (namely, me) is able to withstand the US Army's penetration tests using a screwdriver.

Simply keeping the fuel tank from rupturing from impact is hard enough, without going to use unobtatium or materials that weaken in months if not weeks.

Which one of the two materials (s-glass and nylon) that I mentioned which provoked your rant is "unobtainium" (we bought the S-glass from Aircraft Spruce and the nylon came from a supply company that caters to manufacturers of camping material. The owner actually joked that he would toss in some extra oddly shaped scraps he had around because of the "cool" factor of what we were trying to do) or that weaken any more rapidly than CF? I mean I agree that Kevlar is a bad choice of the inner layers where fuel contact is likely (we used it as an outer layer for the purposes of making the tank "attachments" in our drop tests and making the tanks easy to move during the tests).
Wtf?
A beam and a tank (loaded in pressure) are like black and white, like night and day.

A wall of a tank that is loaded in compression such as what happens as the spar is pushed back into the tank is going to buckle and shatter just like the sine-wave/corrugated beam does if impacted laterally rather than vertically. The issue is the material and not necessarily how it is being used.

CF would be good for a "hard outer shell" for the purposes of making installing the tank easier, but it's not an ideal material for content retention because a reasonable thickness of CF, especially if you're "corrugating" it, is going to buckle and fail. CF is a great material but this is one of those applications where old fashioned fiberglass is a better option because of its properties.

However, I will point something out that if one were willing to make a more complicated "multi-part" tank which is then enclosed, including a corrugated "wall" or beam inside the tank would probably make an excellent reinforcement technique to help maintain shape and minimize fuel sloshing. The issue would be the application of the concept.
 

autoreply

Super Moderator
Staff member
Joined
Jul 7, 2009
Messages
10,792
Location
Rotterdam, Netherlands
Have you not read my posts about the testing we have done on our fuel tank design? [...]
In our series, there were drop tests of a half-full tank and a full tank configuration on to a barrier (the concrete "Jersey" barrier was used to simulate impact with an absolutely unyielding surface), a horizontally-oriented telephone pole (to more accurately simulate a tree impact) and onto a flat surface (concrete parking lot).
Drop a tank on a telephone pole and it'll fall straight through it. Been there, done that, but I'm afraid I don't have pictures or anything else to show my findings.
A wall of a tank that is loaded in compression such as what happens as the spar is pushed back into the tank is going to buckle and shatter just like the sine-wave/corrugated beam does if impacted laterally rather than vertically. The issue is the material and not necessarily how it is being used.

CF would be good for a "hard outer shell" for the purposes of making installing the tank easier, but it's not an ideal material for content retention because a reasonable thickness of CF, especially if you're "corrugating" it, is going to buckle and fail. CF is a great material but this is one of those applications where old fashioned fiberglass is a better option because of its properties.
Sorry Steve, I give up :speechles
 

SVSUSteve

Well-Known Member
Joined
Aug 20, 2007
Messages
3,904
Location
Evansville, Indiana
Drop a tank on a telephone pole and it'll fall straight through it. Been there, done that, but I'm afraid I don't have pictures or anything else to show my findings.

Correct me if I am wrong, but don't you Europeans use a "crash-safe" version of telephone poles that's designed to crumple or fracture? I remember seeing something about it on TV. The ones we use here are solid tree trunks soaked in creosote or something similar. Even with a car hitting a telephone pole, the pole often wins. The 350+ lb 50 gallon tank did crack it partway through it but the tank had a big semi-circle crushed into it.

Sorry Steve, I give up

I am not trying to be difficult. My take is just that we can't march in lock step if we wish to move forward. That is a growing source of frustration for me (and you apparently) because so many have the mindset that the only thing that will work is their approach (be it your love of carbon fiber, my support for technology as an additional layer to prevent/lessen the frequency and/or result of pilot shortcomings or some others and their mindset of "training is the only answer necessary" for the stall reduction/prevention discussion). You feel like you're beating your head against a wall but remember, that I often feel the same way in reverse. ;) Just keep in mind that I don't harbor any ill will or think you don't have things to offer. I learn more from people I disagree with and can debate with than I do from those who take what I say without question.
 

Vigilant1

Well-Known Member
Supporting Member
Joined
Jan 24, 2011
Messages
8,384
Location
US
Kevlar is a downright bad idea for any tank. Fuel will always creep through and trash any organic fiber (like Kevlar) in a short while, carbon and glass will do just fine there.
If the other properties of Kevlar (i.e. toughness, superior resistance to abrasion which might end up being handy as the exterior surface of a tank in a wing, etc) made it worth using, I'd think it might be okay in the >>outer<< "hardcase" (function: prevent foreign object intrusion and crushing in a crash, provide mechanical anchoring of the tank) surrounding the rigid foam (function: impact load spreading) and the inner fuel tank (function: fuel resistant in daily use, prevent breaching of the tank due to crash-induced localized hydraulic pressure). But, yes, there could be better fibers out there.

For energy absorption, "brittleness" or "rigidity" are meaningless terms. What is relevant is how much energy (impact due to crash loads) you can absord. A corrugated tank with carbon/epoxy is pretty optimal for that, energy absorption per pound. Rubber and similar materials are absolutely horrible at absorbing energy.
The failure mode of the material can be important. If, say, Kevlar/aramids tend to "hold together" after localized failure of their resins but CF "shatters", then use of the former in the outer case might offer better protection from a big fuel spill than the later. (BTW, I'm not vouching for this "shatter" or "hold together" idea, but this is a common characterization of these composites, though I've read informed opinions that indicate it is just BS). Similarly, a flexible membrane layer inside the "hardcase" or outside the inner "tank" is there to provide a way to contain fuel in the case of crushing of the rigid structure--it provides an attribute (the ability to contain the fuel despite very large changes in its shape) that the rigid tanks don't posses. I think your preferred approach is to simply avoid, as much as possible, these changes in shape, and that's certainly another way of attacking the problem.

But for such a small airplane like Sone*x you may or may not have a problem with your C.G, if you decide to mount the fuel tank behind your back rather that in front of you, especially if it was designed that way. Of course you could redesign it... This is another reason, why the best place for the fuel is in the wings, as close to C.G as possible.
Yes, the fuel in the stock Sone*x is well forward (between the firewall and the instrument panel) and even so the plane is a bit tail-heavy when the lighter (VW-based) engines are used. Fuel in the wings aft of the spar would cause a bit of a problem, and fuel behind the seats even more of a challenge (hey--a good [-]excuse[/-] reason to go with a heavier, more powerful engine! But wait, that means carrying more fuel, and more aft CG issues)

Resins, ethanol, etc.
I really don't know why someone would like to use epoxies, for their separate fuel tank instead of Vinyl esters resins. I could understand it, if someone was going to build composite wet wings, all those issues with second bonding of Vinyl esters, etc. but for separate fuel tanks, I really don't get it. Vinyl ester resins are much better for chemical applications, than epoxies (unless we are talking about exotics one, that I haven't heard of), not only because I'm saying it, Orion was saying it also.

If it's still not enough, take a look at those specifications, (ethanol page 13) DERAKANE Chemical Resistance guide pdf free ebook download from k.b5z.net
Thanks for that information. It looks like DERAKANE is resistant to most present fuels, though it doesn't like ethanol concentrations above 10%. That's not a problem today, but could be later. But, it does seem better than most epoxies.

One approach is to try to use (for the inner surface of the cell, the one in direct contact with fuel) whatever is being used in most certified aircraft for fuel tanks or for sealing these tanks. The (possibly flawed) idea being that whatever fuel/additives/good ideas the idiots in DC dream up, hopefully it will at least be compatible with the present tanks in the GA fleet. If we use something simialr to the materials used by Cessna, Piper, etc we should be okay.
Another advantage of VE is HDT (high distortion temperature), even for room cured resin, for some +100 C deg. So in short, if you put a composite fuel cell in an aluminum wing, you could paint that wing with whatever color, without worrying that it might become really hot inside of it (wing).
A great point. I'm thinking about an AL wing now so this is a useful consideration. Even for an epoxy tank inside a fuselage, one would need to be careful to cover up the plexiglass and assure good ventilation when parked outside on a sunny day.
Assuming that you are in very early stage at the moment (still deciding what to build), probably I won't be far from true, if I say that you won't fly your airplane in the next 4-5 years. Make tests now. Call guys at Ashland (derakane producers), ask them what is the best option for your application. Buy samples of resin (even here I can buy 1kg without any problem at all). Make specimens, and put them into a jar (turn jar upside down, to be sure is properly sealed, and nothing is going to vapor quickly) with different petrols (you could replace petrol every couple months or so, to make those test even more realistic). If you make samples big enough, you could cut them in half, put one half into jar with petrol, other into for example a cardboard box (hide it from UV). And compare it after a year, two, etc. Thanks to that you will have first hand experience, even before making molds for your tank.
Good ideas--cheap, easy, and could save a lot of grief later.
Thanks to all for the continuing ideas and comments.
 

Vigilant1

Well-Known Member
Supporting Member
Joined
Jan 24, 2011
Messages
8,384
Location
US
For energy absorption, "brittleness" or "rigidity" are meaningless terms. What is relevant is how much energy (impact due to crash loads) you can absord. A corrugated tank with carbon/epoxy is pretty optimal for that, energy absorption per pound. Rubber and similar materials are absolutely horrible at absorbing energy.
More on energy absorption: The composite engineers have looked at this a lot, specifically "impact loads" or what they term "high-strain-rate" applications. To quote from one text (Agarwal, "Analysis and Performance of Fiber Composites" ed 3, p 395) "The suitability of a composite for such applications therefore is determined not only by its static strength considerations but also by its impact behavior or energy-absorbing properties. Frequently, an attempt to improve the tensile properties results in a deterioration of impact properties." The authors go on to say that high-modulus fibers are more brittle and absorb less energy than lower modulus glass fibers.

My (overly?) simplified example of the issue: HDPE (the material we often use for hand-carried gas tanks) and standard bottle glass have about the same tensile strength (30 MPa) and the glass has about double the compressive strength (6000 PSI vs 3000 PSI). Looking just at these numbers, if we made a 5 gal "jerry can" out of 1/8" thick bottle glass it would be much "stronger" (tensile and compressive strength) than one made out of a similar thickness of HDPE. Now, fill both with water and drop them onto concrete. So, I think we can agree that impact strength may not be well correlated with static strength measurements.

Here are the impact strengths of some common fiber/matrix compsites (Agarwal, p 408)
--- Graphite (Modmor II) - epoxy: 54 ft-lb/in^2 (or 114 kJ/m^2)
--- Nomex nylon - epoxy: 55 ft-lb/in^2 (116 kJ/m^2)
--- Kevlar - epoxy: 330 ft-lb/in^2 (694 kJ/m^2)
--- S-Glass - epoxy: 330 ft-lb/in^2 (694 kJ/m^2)
--- 6061-T6 Aluminum alloy: 73 ft-lb/in^2 (153 kJ/m^2)
The figures above come from "Charpy notched impact tests"

So, graphite (carbon fiber) may have some shortcomings in this application, having just 17% of the impact strength that S-glass has. .

Steve-- I don't know how ballistic nylon differs from Nomex nylon, and the above info obviously relate to fibers embedded in a matrix, so I'm not sure if it can applied to what you've been working on.

Seb, none of the composite layups cited used a polyester matrix, but it was a short list.

The text also says that "hybrid composite" layups are often used when both static strength and impact resistance are required. In our current fuel tank application a mixture of CF fibers (high modulus, high ultimate strength, but brittle, poor energy absorption, poor impact strength) and S-glass (lower modulus, lower ultimate strength, better energy absorption) all together in a fuel-resistant matrix might give us the best mix of properties.

There are a lot of variables here. I suspect the rate of onset of the loading may be important (i.e. ballistic rates of onset may not tell us much about what happens at those typical of a 120 MPH (50 m/s) crash). There's probably no substitute for gleaning what we can from the engineering studies, then building some tanks and testing them in a realistic manner. Plus, the testing will be a lot of fun--"look out below!"
 
Last edited:

autoreply

Super Moderator
Staff member
Joined
Jul 7, 2009
Messages
10,792
Location
Rotterdam, Netherlands
More on energy absorption: The composite engineers have looked at this a lot, specifically "impact loads" or what they term "high-strain-rate" applications. To quote from one text (Agarwal, "Analysis and Performance of Fiber Composites" ed 3, p 395) "The suitability of a composite for such applications therefore is determined not only by its static strength considerations but also by its impact behavior or energy-absorbing properties. Frequently, an attempt to improve the tensile properties results in a deterioration of impact properties." The authors go on to say that high-modulus fibers are more brittle and absorb less energy than lower modulus glass fibers.

My (overly?) simplified example of the issue: HDPE (the material we often use for hand-carried gas tanks) and standard bottle glass have about the same tensile strength (30 MPa) and the glass has about double the compressive strength (6000 PSI vs 3000 PSI). Looking just at these numbers, if we made a 5 gal "jerry can" out of 1/8" thick bottle glass it would be much "stronger" (tensile and compressive strength) than one made out of a similar thickness of HDPE. Now, fill both with water and drop them onto concrete. So, I think we can agree that impact strength may not be well correlated with static strength measurements.
Two very different topics. Energy absorption is simply that. The amount of energy a material can absord. Puncture strength (a point load in the material) is orders of magnitude lower and obviously, stiffer materials in general don't perform as well, because they can't spread the puncture energy over a larger area of the material as softer materials do. But then puncture load isn't terribly important per se. Nobody is shooting at my fuel tanks, or trying to stick a knife in.

That's the reasoning behind a corrugated tank. It will be able to deform a sizeable amount, spreading out possible puncture loads, it will spread out inertia (hydraulic shock) over a much longer deceleration and thus improve puncture resistance (a bit) as well. In the end it doesn't behave that different from tanks made of more flexible materials. But it is far lighter (or stronger) for a given strength (or weight) ;)

It's always the combination of shape and material that dictates the behavior. If you've ever tried to hammer a standing bear bottle to the ground (don't), followed by hammering one that lies flat, that is clear enough, the first one takes a massive amount of impact energy from the hammer, the 2nd doesn't. In fact, try it with an empty soda can, but be careful during the preparations (science requires confirmation of experimental data...)
 

Vigilant1

Well-Known Member
Supporting Member
Joined
Jan 24, 2011
Messages
8,384
Location
US
That's the reasoning behind a corrugated tank. It will be able to deform a sizeable amount, spreading out possible puncture loads, it will spread out inertia (hydraulic shock) over a much longer deceleration and thus improve puncture resistance (a bit) as well. In the end it doesn't behave that different from tanks made of more flexible materials. But it is far lighter (or stronger) for a given strength (or weight) ;)
Okay, I'm following you now. I had thought the corrugations were for enhanced rigidity (as they are in a corrugated cardboard box) which is why I suggested another set running 90 degrees to the first ply. You're going for the "accordion effect" to some degree. I'd still think FG might be the preferred fiber reinforcement, esp if this flexibility is desired (with strain-but-no-break an implied necessary attribute), but I've certainly got no testing to back that up.
 
Top