Aluminum wing spar strength

[Michealvalentinsmith]

A D-tube spar typially consists of a main vertical member at the wing's thickest point with a relatively thin web and thicker members at top and bottom (I-beam, or something like that). This takes the bulk of the wing's bending loads. The thinner material around the wing leading edge (the curved part of the "D") closes it into a hollow tube so it can take the torsion loads. A moment (torsion is independent of the location, so it doesn't matter that it's that far forward. Yes I know this is a gross oversimplification but that's the general idea.

-Dana

The number of elected federal officials is limited to congress, the president and the vice president. That's only 537 people. The federal bureaucracy numbers in the millions.....

No disagreement there for sure. However to put the thickest part of the foil with a reasonable distance from the LE such that the resulting D cell isn't 2 feet wide you need to use a foil section with the maximum camber well forward and comparatively short wing chords.

My design has an 8 foot root chord. In the airfoils I'm considering, the maximum camber thickness is at least 2 feet and more frequently 3 feet back from the LE when I plot the coordinates.. That makes for one hell of a wide D cell which is fine in a fixed wing but useless for me in a design that packs for transport.

I do like the idea of a triangular spar. Does anybody know if orientation makes any difference? In the design I posted the base is vertical to act like an I beam, but I seen no reason why the base couldn't sit horizontally and the two sides act as inclined I beams?

Also are there an good ideas for eliminating the point of the triangle gaps and eliminating the need for internal ribs. I don't have a router that would cut the kind of angle in thin ply to miter join the tips. I was thinking square graphlite CF "rod" could fill the gap and complete the tips - I could cover the whole thing then in one layer of S glass and epoxy.

But it's sooo much easier just to get aluminium box section and slide in some plugs and end caps. Problem is the only stuff I can find is building material of unknown alloy.

[wsimpso1]

Spar at the thickest part of the foil, and D-tube formed of everything forward of the spar is the way that it is done. The length from the thickest part of the wing forward is what gives you torsional stiffness. You want length here to make it light... Add ribs for stability and to support the skin against airloads and that finishes the forward part of the wing. This sort of a structure is fixed. The rear half ribs and skin could then be removable...

You have a heck of a challenge ahead of you with this "packs for transport" idea. Usually, you fly an airplane for transport... If you plan to pull the skin and slide the ribs off the spar to ship this thing, you can either use a D-tube, and ship it with the aft portions removed or skip the D-tube, add a substantial drag spar with some sort of way to couple the two spars together. Dr. MacCready's ideas and every fabric only airplane you can look at will also be a help to you. But taking the wing apart to compactly ship it...

Now, this stuff about a triangular spar... they are not efficient. Achieving lightest beams that are stiff enough occurs by putting as much of the beams weight as far from the neutral axis as you can. Wide flange beams, box beams, I-beams, etc. A triangular beam will add aquite a bit of mass at the same stiffness and strength. As to the beam carrying torsion, well, it is lighter to just carry torsion using a drag spar and/or stressed skin. But in a collapsable wing, the skin is what? Just fabric? No torsion carrying capacity there, so it would all be in the spacing between a main spar and a drag spar and the connections between them.

Tough problem here...

Billski

[BBerson]

This motorglider was at Oshkosh 2008. The molded D-cell was able to fold away from the aft part of the wing. The joint closeup is the second photo.

The third photo is an ultralight with a triangle spar and no D-tube.

[Michealvalentinsmith]

Thanks Guys,

The motor glider in the pic looks like the Aeros design that arose as a modification to the swept Stalker and later Phantom rigid HGs.

A crucial point in my design I didn't mention is internal cable truss bracing.
I plan to use square sections that connect the upper and lower ribs with a hole to allow sliding to the center along the spar - then the spar is removed.

From the square bulkheads I run internal cables from each corner. With 4 bulkheads along the span I'm getting some impressive strength to weight ratios and the cable add torsional as with as bending moment resistance. With an 18% thickness ratio I get a 4 g loading resistance in a 20 lb spar. If I go to 25% it goes up to 6 gs.

Basically the ribs slide to the center with the internal cables. The ribs remain fixed inside the wing. Once accordioned to the center the spar pops off and breaks down to 2 pieces. So far a package as small as 6 feet long that sets up in 5 minutes appears plausable - giving me quite a portability and convenience advantage over current hang gliders and a performance advantage over paragliderss

[wsimpso1]

This is an unusual construction, using cables (tensioned?) to form a large part of the spar, etc. I assume that the fabric skins are tensioned too. I am suspicious that the cables will put the beam into bending in the same direction as the aeroloads will place on it, using up some of its capability in the process. Additionally, the aeroloads will tend to unload the cables. I am skeptical that you can simultaneously make this design scheme as strong and as light as you think you can.

This is an unconventional design. You really need to convince yourself that it is as strong and as a light as you think it will be. Before building a full scale wing, I urge you to design a scale model, analyze it, test it to destruction with loads distributed per the aeroloads, and compare it to your analysis.

BIllski

[Michealvalentinsmith]

No so unconventional as you might think.

I have a simple method of for tensioning the last rib on the spar with an eccentric cam. The fabric is tensioned to a degree as well. Fabric tension is not a new idea in aviation Henry Mignet and a great many others since have successfully used what is called tensioned membrane wings. None have used them with an internal cable truss though.

However the cable truss has a lot of precedent in various applications including aviation. The truss is really a series of short kingposts to overcome the limitations of one large one.

You can see them with small spaced kingposts on sailboat masts which take considerably higher loads than I'm considering. But in aviation a HG company called Bautek used internal truss bracing on a traditional HG aluminum cross tube to brace for negative load and make a topless wing spar with the same strength to weight as a built up CF spar. They used only 2 internal kingposts and a kingpost hight of about 3 inches - but got a 6 g negative load tolerance - admittedly they had the additional support of the LE tube as well as the X tube.

The bending moments you describe can occur - but only between the kingposts - if widely spaced. I have found with one mid span kingpost/bulkhead (one in the center one at the tip - for 5 total on a 30 foot span) the spar will bend between the kingposts if the loads are high enough - as you describe.


For the wing to bend at a kingposted point, the spar must bend, as the spar bends the relative distance between the cable ends increases so the cable is placed under higher tension - not lower tension with aeroload. The higher tension acts at the kingposts and pulls the kingpost back down. Essentially bending moments are converted to compression moments by cable tensioning.

Obviously the more kingposts you have and the higher the kingpost, the greater the strength. There appears to be an ideal number of about 4 to 6 per side on a 15 foot span - so 9 to 11 total to get strength to weight ratios well within current HG requirements.

[Michealvalentinsmith]

Here's a pic of the Bautec internal cable truss. The pic doesn't show a second short kingpost at the mid span of the X tube. The kingpost is about 2 inches high. With a deeper foil section I plan to get effective king post heights of 9 to 12 inches. I can also just use the truss bracing for positive load and traditional cable bracing off the A frame for positive load.

This way I get an effective kingpost for negative load of 18 to 24 inches.

[Michealvalentinsmith]

As an aside I've been using old scrap hang glider tube for testing. In the final version I'd like to use box or triangular section to prevent rib rotation on the tube and transfer torsional loads to the spar and cables.

For testing I've been placing weights - cement bags and water bottles at spaced intervals and have loaded a truss with 4 kingposts per side to 850lbs without failure or detectable bending - I ran out of weight to test to failure. The tube is old and corroded and is thin walled LE tube, not thicker X tube section.

To get an idea of the strength, I weigh 180 lbs and my friend weighs 175lbs. We can sit on the tips and bounce without the spar flexing or breaking.

Unlike a HG, the strength runs the entire wing span. A HG braces to the X tube junction only the outer tip half is completely unbraced.

[wsimpso1]

So it is internally braced for negative g's and externally braced for positive g's? The positive g part was missing... I am very familiar with sailboat mast systems.

This sort of scheme is essentially spar caps that have the size and strength of the cables being used, which means you have big spar caps fullspan. When they are external, no doubt about it, you have a lighter rig. When the braces are internal, you probably can save some weight by making the spar full depth and sizing the caps to allow skipping the bracing...

One thing you have to remember about cable braced columns, they are already in compression before the air loads get to them, and then the compression increases, as well as bending loads from the airloads. If you cover all of these, it is fine.

Now here is the point where I ask which is more important? Light weight or low drag? If low drag is important, you would put all of the bracing internal at a small penalty in weight. If low weight is important, just use external bracing and save some weight.

One thing not covered. Your slender spar will not have a lot of torsional stiffness, and you wing may twist off quite a bit at the tip. A full depth spar will give a lot more torsional stiffness. Something to check out.

Billski

[Michealvalentinsmith]

So it is internally braced for negative g's and externally braced for positive g's? The positive g part was missing... I am very familiar with sailboat mast systems.

This sort of scheme is essentially spar caps that have the size and strength of the cables being used, which means you have big spar caps fullspan. When they are external, no doubt about it, you have a lighter rig. When the braces are internal, you probably can save some weight by making the spar full depth and sizing the caps to allow skipping the bracing...

One thing you have to remember about cable braced columns, they are already in compression before the air loads get to them, and then the compression increases, as well as bending loads from the airloads. If you cover all of these, it is fine.

Now here is the point where I ask which is more important? Light weight or low drag? If low drag is important, you would put all of the bracing internal at a small penalty in weight. If low weight is important, just use external bracing and save some weight.

One thing not covered. Your slender spar will not have a lot of torsional stiffness, and you wing may twist off quite a bit at the tip. A full depth spar will give a lot more torsional stiffness. Something to check out.

Billski

Thanks Bill,

As usual you bring up excellent points. I've only made a few full scale wings - none met my expectations - and a lot of RC test models. The main reason I joined here was to get as many devil's advocates and cover as many caveats as possible from informed sources before wasting a lot of time, money and energy on a bad idea . And this is much appreciated

I'm no engineer but read all I can find. I think your are spot on and torsional loads are the big question mark. With such an unusual design there not a lot of precedent to learn from and information is scarce.

Topless hang gliders place quite a bit of torsional load on the LE tube with the sprogs - BUT the LE tube is braced by the X tube connection giving a lot of torsional resistance.

I'm hoping a combination of strategies will cover torsional loads without the X tube. Firstly I simply plan to use a thicker section - HGs use a X tube and a LE tube so a single tube can be beefed up quite a bit before hitting comparable weight. By placing the tube further back at the AC and using a plank or low sweep design - say 8 degrees, then using spoilerons rather than ailerons - I hope to reduce torsional loads further.

It also seems possible to space the cables considerably wider than the spar. To avoid necessitating an excessively heavy spar cap I can either make a wider cap and cut out all the non load bearing sections, or use vertical members at the cable attachments between top and bottom ribs(diagonals between to resist distortion) , and a separate one to house the spar.

One technique I developed that works very well is to simply use double side wires off the lower A frame to the mid span rib fore and aft of the spar. This virtually eliminates all rotation at the mid span but adds an additional lower side wire (net drag wise I'm still well ahead of a traditional HG with upper kingpost - 4 main wires and 8 smaller wires to hold the TE up).

My whole concept is to make a structure that falls between external cable and kingpost bracing (with lower net drag than external cable bracing) without the expense and complexity of cantilever design and with portability as a premium. The deeper foil almost certainly will have more drag - but it appears a deeper foil with internal bracing has less total drag than a thinner foil with external bracing.

Since the A frame under the wing needs bracing anyway, I figure I may as well use it for positive and torsional loads . Most older style designs using a single spar, also used a set of double cables off the A frame for positive load - 4 per side. I think by moving the internal spar closer to the bottom of the foil I can internal and external cable bracing for positive loads - with only a single set of external lower cables.

I'm also considering using seat belt webbing rather than cable for the internal bracing. The reported strengths are as good or better, it's kinder to the wing fabric and easier to use - but I can find no precedent for it's use in aviation. I suspect stretch may be a problem.

The only remaining problem is I would have like to use a moving A frame to activate the controls like the current range of rigid HGs. Using the fixed A frame as a kingpost pretty much rules this out so I'm stuck with slider type controls to activate the spoilerons. The advantage if I place them close enough to the CofG the pitch effect is minimal and I can activate both surfaces for descent control.

[Michealvalentinsmith]

Incidentally I'm trying to decide on a zero sweep, zero twist design or using the available 8 degrees of sweep the concept can accommodate without bending the spar at the center junction.

So far the guys in the know say I'm better off not using sweep and twist with such a low sweep angle.

Seems to me though I can go for a Culver type twist with most of the twist inboard - get a near elliptical distribution - eliminate a lot of the outboard torsion on the spar - get excellent spin characteristics and use about 3 degrees total twist with positive Cm foils. Spoilerons should overcome the adverse yaw problems.

[wsimpso1]

Taking a page from yachting, you don't want to use nylon, it is springy and stretches. Polyester was preferred for many years as it is stiffer and does not stretch once the load is established. Then there are the newer fibers. Check out any of the mail order sailboat racing suppliers for info on synthetic lines. Long term UV will weaken them.

Billski

[GESchwarz]

BBerson,

Thank you for sharing the nice pics of a triangular spar. I've never seen anything like it.

Because I'm using the Grumman folding wing geometry, my main spar is leaned back 26 degrees. Like the Grumman Avenger, my center section spar will probably have a double web, the one at 26 degrees from vertical and the other at 0 degrees, joining at the top. This is the section that will pass through the fuselage and will detour (one web fwd, the other aft)around the pilot's seat to allow for seat suspension travel (a high-G safety feature).

[Michealvalentinsmith]

Thanks guys.

Yeah I figured stretch would be a problem. UV exposure isn't though as the straps would be inside the wing.

I was considering dyneema or spectra type line - you can get lightning rope fairly cheap, but I can't seem to find any tape - which would be much better.

I need at least 4 lines to replace one cable and I'd need to space them quite closely - so that gets expensive. I was also considering a solid ply D section (with aft aluminum tube) and going around the D circumference with spaced line as I figure this would support the entire LE better.

Which brings me to the main problem with the internal spar at the AC - tringular or otherwise. You really would only use it alone if portability was a premium - most designs add an extra LE former. Just as the wing scallops between the ribs - it will do so at the LE as well if there's no solid support.

I have seen a sleeve in the LE with PU foam inserts work well but I'm concerned with repeated folding the foam will crinkle badly. Bob Rousse use inserted shingles - but that's as bad as inserting ribs in a HG. Mignet just used a fatter rib and allowed the LE to scallop - but that's inefficient IMHO.

So I'm thinking a combination of tensioned cable, partial LE ribs and ram pressure like a PG will do the job. Small holes in the LE should allow enough internal pressure to fill out the LE section if I get it right.

[Michealvalentinsmith]

BBerson,

Thank you for sharing the nice pics of a triangular spar. I've never seen anything like it.

Because I'm using the Grumman folding wing geometry, my main spar is leaned back 26 degrees. Like the Grumman Avenger, my center section spar will probably have a double web, the one at 26 degrees from vertical and the other at 0 degrees, joining at the top. This is the section that will pass through the fuselage and will detour (one web fwd, the other aft)around the pilot's seat to allow for seat suspension travel (a high-G safety feature).

I have a pie in the sky design I'd like to one day make - that is similar in some respects where the spar must accommodate a central hole - or box - to house the pilot. I have a real soft spot for pilot in wing designs like the Hortens even though there has never been a commercial success.

I do much the same but with the cable junctions inclined fore and aft on a central box. Unlike your design where the spars are continuous, I must must attach the single spar to the side of the box walls and I fear high compression loads may cause wall failure.

I've attached a very rough sketch and a pic of the Mitchell wing using the same concept with standard D cell spar.