Fiber Glass Fuel Tank

[Murray Randall]

Attached are some pics of my pretty well finished fuel tank. It is generally three layers of 7781 cloth and vinylester resin. The filler neck is substantially reworked hardware store PVC pipe sealed in with Proseal and eight 8-32 SS screws. The tank conforms to the airfoil, has 1/8 NPT taps bonded in forward, rear bottom and an option for a sight glass at the top.  I laid up flanges that rest on the top of the adjacent ribs.  I also made two fully adjustable .020 aluminum bottom support straps and one adjustable top strap.  Not sure that I like the tank resting on the ribs, but have not cut them back yet. I'd appreciate advice on the mounting. Planning such a tank I did not put two ribs inboard, only one.  Also I did not bond in a drag strut but moved out a bay and bonded in an X.  This provides outlandish tank volume.  The tank weighs in a 6 lbs.  I wanted it to weigh 5 lbs but that didn't happen. Steve recommends a post oven cure at 130 degrees F, which I would like to do, but the tank is 27" and my oven 24". The data on post cures want 4 hours.  Thats a lot of borrowed pizza oven time eh?   I thought about a 100 bulb inside but the filler is only 1.5".  Ideas again? That 7781 cloth is the nicest compound wrapping cloth I've seen.  Another thing Steve advised making an aluminum tank.  I'd say he's spot on there but I'm a terrible thin wall alum welder. 

[GriffWatkins]

Lookin' good!
If you only need to go to 130, I'd think you could stick it in a big cardboard box along with a hotplate set on low.  So long as they're not very close together, and the heat is kept indirect, it should work fine.  Maybe use a digital cooking thermometer to monitor it?
If you absolutely have to heat the inside of the tank, I guess you could suspend a water heater element down in the filler neck, but that seems unnecessary.

[Tom H]

Murray, I know how much work it takes to make a FG tank - any tank, for that matter.  It looks good.  Post-curing the tank would ensure that all chemical reactions are finished off and help fend off the evil additives in gasoline.

I am a bit concerned about the removal of the drag strut in the fuel bay (1st bay).  When the wing is flying forward, the drag forces are resisted by the drag struts, which are in compressing.  The drag forces add up and are the greatest in the 1st bay.  With no drag strut, the tank will be resisting those forces.  Don't know if this is critical, but could be.  The "X" in bay 2 does not contribute to countering drag forces in bay 1.

Is there any way to incorporate a drag strut?  Maybe put a tunnel in the bottom of the tank?

Just a cautionary note.

[Murray Randall]

 

Thank you very much for your input on the drag issue Tom.  I contemplated it a great deal myself.  A great deal. My evaluation:  Wing drag loads are primarily accommodated by the drag struts triangulating the compression struts and the spars.  This resists the wing bending load about the vertical axis applied by airflow at the leading edge.  Where we might differ is that I believe that the wing drag bending resistance does not have to be continuous over the entire span.  But maybe intermitent.  Yes that inboard tank bay of mine will see a modestly higher compressive load on the rear spar and tension on the front spar.  But the next bay with X bracing is considerable stronger than single drag braced bay and should take much of those spar load changes.

 

Never the less I did want to do my geeky analysis thing again.   I made a finite element model but could not get the math to resolve to my satisfaction due to evil interactions of my interfacing plate and beam elements.  (Singularities in the stiffness matrix.)   The shop rat hauled me away from the computer again. 

[okdonn]

I could be way off base here, but I would think that the spars mounting to the fuselage, and the longeron between the wing mounts would provide the drag strut function for the first bay???

[scottiniowa]

Great work on the fiberglass tank, I don’t know on your using the X on the next bays… Sure looks good but I also don’t know how to do those calculations.

Generally tanks are held with two straps below, and 2 on top.    You do have a interesting fit up there.

I enclosed two different tank proposals, both in XL wings,  the smaller tank is 4.9 gal and the second with a step up for the diagonal is just shy of 9 gal.  So yours I am guessing may be 10 or so?

These are just offerings of photos that may give others ideas.  Your going to like that sight gage set up.

Best of success.
Scott

[Tom H]

Murray, I see what you are saying about the increased tension in the main spar and increased compression in the rear spar.  But, what I am concerned about is this - The wing drag is trying to move the entire wing straight aft.  The inboard rib (rib 0) is fixed to the fuselage.  The next rib out, on the other side of the tank, rib 1, is also being forced straight aft.  This is setting up a shearing force in the space between rib 0 and rib 1.  This force is trying to make the normal rectangular shaped area in bay 1 to become non-rectangular.  Rib 1 is being slid backwards related to rib 0.  The diagonal in bay 1 would resist that force, and it is not there, so the force will be resisted by the tank.

Again, I don't know if this is critical, don't know if the tank structure is sufficient to resist this force.  Seems you have the background and tools to sort it out.  I just wanted to be clear about my concern.

[Murray Randall]

Yes sir the drag loads sure want to make a parallelogram mess out of the wing but if you resist that drag force anywhere along the span of the wing with structure that can withstand the drag force and minimize deflections, my thesis is that you have accomplished the task.  You will alter the stress modestly along span. 

 

Sam measured thrust at 180 lbs once.  100 lbs per wing.   Isn't VNE 85mph?  I figure a VNE dive might go to 400 lbs.  Can't the drag force be equal lift in some conditions?   Thats a pretty serious force!   That leading edg D section specially if you put a light nose rib between the design ribs is a mighty strong element.  The wing was tested but with no drag forces. Joe spun Put Put.  Lots of interesting data. 

 

Another kinda interesting point.  When I did the finite element analysis thing.  The parallel wing struts were much much stronger than a VEE strut geometry would be.  The parallel struts took some of that drag bending helping the wing strength and also resulted in lower fuselage tube stresses.  In retrospect you could expect that but I was surprised. 

[Tom H]

Murray, I just couldn't get comfortable with removing the drag strut in the inboard bay where the tank was installed.  Maybe I was missing something.  So, I started researching some design issues with drag/anti-drag in this type wing.

Drag forces can be either forward (drag) or backwards (anti-drag).  Turns out that the anti-drag forces are larger than the drag forces.  This is due to the angle of attack lift vector.  At positive angle of attack, the lift vector is angled forward, resulting in an anti-drag force - trying to push the wing in it's forward direction along the cord line.  My reading showed that at high G pull-ups, the anti-drag force can be as much as the gross weight.  A good bit of force.

Some wings use drag/anti-drag wires instead of drag struts.  Some designs have the tanks in the wing and have tubes through the tanks to run the wires through.

These internal drag bracings are referred to as truss construction.  If you remove the triangulation, you don't have a truss design anymore.

An analogy would be to consider the main spar design.  It is composed of top and bottom caps, vertical interstitial spacers, and a web.  It is designed to carry primary forces that are aligned with its web surface.  Removing a section of the web between the vertical spacers would compromise the spar strength.

With regards to drag/anti-drag forces and their direction, I think removing the drag strut between compression struts is a similar modification of structure as removing the web between the vertical spacers in a wing strut.

Another piece of evidence is that some wing designs utilize wire bracing between the parallel front and rear wing struts to handle the drag/anti-drag forces.  These don't have the drag bracing inside the wing.

Others utilize wing struts in a V configuration to handle those forces.

I just find too much evidence that says the drag/anti-drag forces are significant and have to be dealt with properly, and I think that removing the drag strut from the wing structure without replacing its function with something else is not proper.

I may be missing something, always seeking the TRVTH.

[Murray Randall]

I agree with absolutely everything you are saying Tom.   100%.   And I like your pointing out that the wing spars with the drag and compression struts are effectively an I beam w/ the C and D struts acting as a spar web. Somebody once said on the site that the web is only there to keep the caps in proper position.  Thats a big truth.    The Eagle designs handle the drag and anti drag loads and I have modified that important aspect of the design.  I removed the inboard drag strut, put in a X in the next bay.  The question remains, is whether that X is sufficient compensation for removal of inboard drag strut.  Maybe I should go back to the finite element analysis and work the thru the math problems I first encountered.  I can do that.  But the analysis program was making me work hard, I'm lazy or happier in the shop than at the computer.   But lets go back to the spar and web comparison that you brought up. What I've done is to leave a one bay gap in the web.  Isn't that like putting the lightening holes in the wing spar webs?  Another change in my XL is to make a four longeron tube fuse rather than the design three tube fuse, with a steel tube X brace in the center section. This might add to the drag/antidrag capability.  Then I put thin nose ribs between each design nose rib to stiffen the D of the nose and that D contributes to the drag/antidrag capability.   If the above doesn't make you feel comfortable maybe I'll go back to the computer.

 

[GriffWatkins]

Murray,
I agree that the spar analogy is a good one.

However, I think the 'lightening holes' comparison you've made is a little bit off.  The X design of the drag struts in the wing is already analogous to the lightening holes in the spar- the web forms support and connecting material between the two caps on the spar, with some material removed to improve weight; the drag and anti-drag struts are just like the wood left around the lightening holes.

A better analogy to what you're proposing would be to take a spar with lightening holes already in it and cut out a one foot section of ALL of the web between the two caps, and then assume that the caps and material that remains on either side of your cuts would be substantial enough to support the plane in flight.  The results should be pretty easy to figure out, even without failure modelling.

I believe I'd go back to the original design; the increased range you get from that tank might only extend the time that you're in the air wondering whether your wing will collapse in a strong gust.

[stevejahr]

The wing LE is one structure unto itself and will carry all kinds of loads: drag/anti-drag, lift, and torsion.  Typical model practice uses larger LE D tube and omits the drag/anti-drag structure entirely.  An interesting experiment is to build a D-tube without the spar shear web.  It is all twisty and easy to add washout.  Then you put in the first shear web and everything is locked in place (so get it right before putting that web in).

Then there is the structure formed by the front and rear spars with the compression and drag/anti-drag struts (or wires).  Vertical deflection of the spar caps is controlled by the shear webs.  Longitude deflection of the front spar caps is controlled by the LE and the rear is then tied to the front via ribs and compression struts.  Torsional deflection is controlled again by the LE with an assist from the ribs.  That leaves parallelogram deflection and if you tie the spars together anywhere along their span they will not be able to move that way.  If the fuse mounting is rigid then one simple X in the outer bay would suffice.

Further the X is stronger than the basic design since the two members can prevent their own deflection.

Not to mention the not insignificant shear capability of the tank and its mounting.

I see no doom here.  

[Murray Randall]

The strength of the XL wing spars in vertical load bending is real close to 80% the caps and 20% the web.  Not uncommon for aircraft wing spars and I beams in general.

But I'm getting motivated to make a simple finite element model to compare the XL design to the "skip the first bay put an X in the 2nd bay concept".  I say a simple model and compare.  Not a model to evaluate vertical or any torsional loads, only the drag.  And the objective would be to compare the two designs not provide absolute third decimal place stress level accuracy.   Thats dependent upon the old Windows XL machine upon which the Stardyne analysis program is installed cooperating.

[Vince Carucci]

I believe Tom H is correct. The stress resulting from drag forces increase as it approach the fuselage. What ever load was being borne by the inboard compression struts is now going to be distributed to the spars in an axis they were never intended to carry, by the fabric whose strength is negligible, the .8 mm plywood over the nose of the wing (also negligible) and to the fiberglass fuel tank. If the fuel tank wasn't designed to replace the compression strut, it probably won't.

The attached image is a representation of what the wing will experience due to drag. It is effectively a cantilevered truss design experiencing a distributed load.  Without the first diagonal support, the whole structure is likely to collapse. Please reconsider what is in the first bay.

Vince

[scottiniowa]

Vince

With all due respect, what you showed in the drawing is not what he showed us via photos,  of what  he did .  Clearly he showed a   Full X brace diagonals.  starting in rib bay #2.  I don’t know if this is right or wrong, but I do know the drawing and what he did is different.

Just sayin’