Foam/Plywood wings

Help Support HomeBuiltAirplanes.com:

wsimpso1

Super Moderator
Staff member
Log Member
Bart hits it well. There is negative pay value for the work. Other people have done the math before you.

Analysis is straightforward but big. Spar carries shear and bending - do initial sizing by standard beam theory. Wing skin plus spars carries torsion - You could size the skin just to carry torsion, but everybody uses 2 UNI at +/-45 deg plus 1 UNI lengthwise or something close to that for the outside and 2 BID or 2 UNI at +/- 45 deg for the inside of the panel, so start there because you will probably end up there just to have a wing you can build and handle. Skin panels carry airloads where the skin is broken into panels by main spar/ ribs/ auxilary spars/ other stiffeners. The skin ends up having several things happen simultaneously: Whole wing bending deflection (compression on one skin, tension on the other); Whole wing torsion deflection (shear on the skin), and; Panel deflection form air loads. Once you have the various system stiffnesses together (composities mechanics is a lot of matrix math) and the oddball skin buckling fuss, you need to determine each of the load cases, turn the crank to determine the system deflections, lamina deflections and check against failure criteria. The initial design will fail the shear web, and maybe some other stuff too. You have several paths to pursue to make it strong enough and then to minumize weight, but it will be a blend of beefing up the main spar caps, main spar shear webs, and decreasing the rib spacing. A seriously multivariate problem.

The composite mechanics is a graduate class in mechanical engineering with prequisites in solid mechanics and matrix mathematics. The optimization is a designed experiments problem. I have all of this in my background, and I understand it. If you do too, you can work your way through the composites texts and understand what you are doing and maybe even have some fun with it. Yeah, I am a geek. Maybe you are too. Grin.

For little airplanes (Even the Rutan Voyager has the non-fuel tank sections built this way), the surfaces are lighter with just a solid core skinned on the outside. The weight savings from removing some of the foam is offset by the weight of the fiberglass skin on the inside, the fiberglass/foam ribs, and the adhesives and tapes to couple them together.

Billski

Bart

Well-Known Member
Bart hits it well. There is negative pay value for the work. Other people have done the math before you...

...The weight savings from removing some of the foam is offset by the weight of the fiberglass skin on the inside, the fiberglass/foam ribs, and the adhesives and tapes to couple them together.
Billski
Not only that, but in addition to all the extra work and expense you went to, perhaps for no or negative weight savings, now you've got stressed parts in places you cannot see or inspect. This is where Gremlins camp out, like hobos under a bridge, and they're not housebroken, either.

Best bet? Read and heed Strojnik.

The whole idea of a foam core wing is to avoid stress concentrations in the first place, like a great big pillow.

Even if I could save a bit of weight by hollowing a foam wing, I wouldn't since the savings would be minimal, if any. It would be better to pay the slight weight penalty, if any.

aerogant

Well-Known Member
Thank you very much for the insight on analyzing sandwich core wing skins. It made sense. I can already envision a monster spreadsheet to work out the details on such a thing.
However, in my application, I'm really chasing weight with an extreme mindset. I did a few rough calculations using full foam core and the weight is more than just a slight penalty.
I was inspired by these photos of a Europa wing being built. They hollow out the foam to a pretty big extent. I was thinking that my projected weight would drop greatly if I could do something similar.
http://www.kaon.co.nz/europa/272h.html
They don't use fiberglass on the interior of the holes either.
I remember Bruhn talking about multiple cell torsional tubes. Could this hollowed out foam be thought of in that fashion, or would the foam be too weak to appreciably carry the load? Bruhn's diagrams seem a little cryptic and lead me to believe that he is talking about things more substantial than blue foam when it comes to transferring torque. This would seem to be a neat and fast method of construction if the numbers work out. Thanks.

wsimpso1

Super Moderator
Staff member
Log Member
OK, I am sure that some things were missed in the thinking here.

First and foremost, if there is no glass on the "other" side of the foam, the skin bending stiffness is really low. How low? The difference between the single sided glass on 1" foam and double sided glass on 0.38" foam is 20:1 in bending stiffness. So, the skin will be buckling limited in how much torsion and bending load it will carry. A big change...

Now making the main spar carry torsion costs a bunch of wieght. Usually, a vestigal drag spar and/or leading edge spar is built into the root end of the wing to transfer torsion loads from the skin to the vestigal spars and into the fuselage. So, either you beef up the skins so that they can carry more load to the vestigal drag spar, or you carry a full length drag spar so that the skin only has to carry torsion to the drag spar. Either way, there will be a weight gain in this area because the foam was removed.

Then there are wing skin stresses due to the direct airload. It bulges the skin between ribs/spars/longerons. With the skin bending stiffness way down, you might need to add skin thickness in the glass. Nominal Rutan style wing outer skins are 3 UNI. And the Europa's wing skins? The slower the airplane, the smaller this effect.

So I did a little model of a wing 16' long with a 3 foot chord. I assumed that the wing aft of 75%C is the same (ailerons, flaps, hardware), and left the foam 1" thick on the skin and in five longerons. Hollowing out the wing saved 4.6 pounds. If the changes to the spar system and skins added 4.6 pounds, you are even. If the changes to the spar system and skins added more than 4.6 pounds, you have a net gain in weight. Fat chance on getting ahead this way. My example wing gained 4.7 pounds by the addition of one extra ply of Uni cloth over the wing, but it only increased wing skin bending stiffness by about 5%.

Believe what you want. Reality is that, in little airplanes, your airplane gets heavier when you hollow out the wings, tail, and control surfaces because the structure you have to add is heavier than the foam was.

If you really want lighter than Rutan style, you have to tolerate being fragile. Sometimes that is OK, but only in very special purpose airplanes.

Billski

Bart

Well-Known Member
Bingo, bingo, and more bingo.

All that extra work hollowing out the wing foam would better be spent on making good, light skins, per Strojnik method. What's more, if you screw up a panel per Strojnik method, it's not an expensive boo boo.

And what is the Strojnik method of wing skin fabrication? Layup of ~4' wide panels of fiberglass on a glass or plexiglass flat table, with vacuum bag, if desired. Then, after ~45 minutes and the resin has started to gel, peel the skin off the glass-smooth surface, which leaves the layup panel also glass-smooth. Drape it over the foam wing foam and vacuum down to the surface so it conforms properly, until the resin reaches final cure. Very little or no sanding this way, with glass-smooth surface. Also, as there is little or no sanding, there is no danger of sanding through or compromising the structure of the skin, which is a stressed structural member of the wing.

Jeremy

Well-Known Member
I am pretty sure that the Europa wing design is using the foam as a substitute for ribs only. The wing still has a substantial, and fairly conventional, pair of spars, so the hollowed foam sections are acting pretty much as a continuous rib.

The Europa wing isn't particularly light, even compared with Rutan-type construction, but it is fairly easy to fabricate to a consistent overall strength, as much of this relies on the spars, rather than the skin.

As this design was intended to be home built in the UK, where our certification requirements are pretty tough, my guess is that Ivan Shaw deliberately chose this form of construction to minimise some of the hassle builders would have had when trying prove that critical layups were all completed to an adequate standard.

I've now built a test section, using a tubular alloy spar, foam ribs and a thin balsa/glass leading edge skin and am very pleased with the ease of build and finish. Foam has the great advantage of being really quick and easy to cut to shape with a hot wire for ribs (I've used 30kg/m^3 extruded polystyrene, 1" thick). The ribs have thin epoxy/glass cloth on either side.

The balsa leading edge is even simpler, just layup 200g/m^2 BID on a smooth sheet, lay the 1/16th balsa sheet on top, vacuum it until cured and you're left with a sheet of material that readily curves in one direction (glass on the outside). This can easily be laid over the ribs and bonded on. I opted not to glass the inside of the leading edge section, just coat it in thin epoxy. It is both very fair and smooth, plus seems stiffer than thin ply. I've made rib caps for the rear 3/4 of the wing from epoxy/glass/balsa as well, but with the grain running chordwise.

Jeremy

wsimpso1

Super Moderator
Staff member
Log Member
Ah, so the Europa does have full length drag spar... You will only find one full length spar in a solid foam wing, and the skins can carry all of the torsion down to the root, where it is reacted into the fuselage or strakes pretty easily. Same airplane, the solid core wing design will be lighter.

With the cert requirements in the UK so stiff, has anyone been able to register Long EZ's, Cozy's things like that?

Now Jeremy, it sounds like you use the same foam and glass that we use or very close to it. Why a tube spar? Does it have an internal web or other reinforcement? Unless we are talking a low speed low wing loading bird, you will really have a bunch of ribs. What is the wing loading and max airspeed that you are designing for? How is the wing covered from the spar aft? Fabric? More balsa/glass?

Billski

Jeremy

Well-Known Member
Bill,

The Rutan designs are OK here, at least some of them, but only because they had a proven track record of having been built and flown in the US for a long time before the first UK example got approved. Ivan Shaw was starting from scratch over here with the Europa design, which may be the reason for his conservatism. We do have to put up with a lot of paperwork here, as we have no experimental class; every homebuilt is supposed to meet the same design codes etc as a certificated aircraft, even ultralights.................

I'm building to meet the new deregulated microlight category that has just been made legal here. This calls for an empty weight not exceeding 115kg and an empty weight wing loading not exceeding 10kg/m^2. I've opted to go for the lightest empty weight I can to increase the MTOW wing loading to a reasonable value. The design is also being entered in a competition here, and I've opted to enter the "simple design" category.

This has shaped the design to use the absolute minimum number of mechanical fastenings, to be as easy to build as possible, not use overly esoteric materials, whilst keeping the empty weight down to about 65kg, including the little 20hp engine. As you can guess, this has been tough, but the tube spar/foam rib option does make the wing very easy to build, relatively light and keeps the structure simple. It's a high wing, single strut braced, design, with the lift-off wing in one piece (the span is only 6 metres, it uses a standard 5 metre length spar tube with short tip extensions). Wing weight should be around 17kg if I can build it carefully!

Jeremy

wsimpso1

Super Moderator
Staff member
Log Member
That is some tough design spec that jeremy is trying to live within. With power loading as low as he is running, wing loadings must be very low. At these levels, airpseeds are so low that laminar flow is almost a given, and foil accuracy that we might pursue in a Long EZ is unnecessary.

This is a classic example of designing his airplane for his flight regime, and doing what is right for that.

Cool stuff Jeremy!

Billski

Jeremy

Well-Known Member
It is indeed a tough spec! In addition to the wing loading and empty weight restriction, we have to have a MTOW not exceeding 300kg (661 lbs) and a stall speed not greater than 35 kts CAS. The latter two restrictions are academic really, given the empty weight and wing loading limits.

The design competition rules are simple, stay within the legal restrictions! There are two competition classes, "simple" and "state of the art", and I'm entering the simple class.

The approach I've opted for is to go for minimum drag, simple construction and a light empty weight to allow the use of a smaller wing area, so gaining a higher MTOW wing loading, primarily for handling and performance reasons. I've selected a low pitching moment aerofoil section, to reduce the induced drag from the horizontal stab. This has allowed me to save a bit of weight too, as the stab area can be a bit less.

At the moment I'm working on a design MTOW of 170kg, but may increase this in the light of the final analysis results. The useful load of 105kg (~231 lbs) is OK for me, as I weigh 84kg, which leaves room for 29 litres (~7 US gals) of fuel, plenty for an engine that only uses perhaps 4 or 5 litres per hour.

I may increase the MTOW in the light of the final stress analysis to allow for heavier pilots and a reasonable fuel load. I'm working to +4g, -2g, with a normal SF of 1.5, plus special factors; fitting factor of 1.15, cable factor of 2 and bearing factor of 2. Even the modest 170 kg MTOW gives me a ratio of useful load to empty weight of 1.6:1, not really up in the Rutan Voyager league, but still respectable.

Overall, I have to say that having tough goals makes the design process far more interesting. The only problem is that it takes hours to get each element optimised for strength and weight - I spend forever looking at the drawings trying to see how I can gain better structural economy, whilst staying inside the "simple" theme.

Jeremy

Last edited:

david brown

Member
hiya

Sandwhich construction when correctly engineered is a sound solution.
The problem is allowing enthusiasts with little expertiese or knowlege to get involved.
Since this is an affordaplane site & it is infered from your comments that you know about engineering matters. Im supprised that you have not advised prospective builders that the box section fuselarge is unecesarily heavy. that a significant weight saving is easily obtained by substituteing the box section for something lighter. ive done the calcs

so many construct

Jeremy

Well-Known Member
hiya

Sandwhich construction when correctly engineered is a sound solution.
The problem is allowing enthusiasts with little expertiese or knowlege to get involved.
Since this is an affordaplane site & it is infered from your comments that you know about engineering matters. Im supprised that you have not advised prospective builders that the box section fuselarge is unecesarily heavy. that a significant weight saving is easily obtained by substituteing the box section for something lighter. ive done the calcs

so many construct

This old thread you've dredged up has nothing whatsoever to do with the Affordaplane and the implication that I didn't know what I was doing with the Mayfly design several years ago is just plain wrong. The wing sample turned out fine and exceeded BCAR Section S when tested, although I never went on to finish it because I was medically grounded.

IMHO the Affordaplane has so many design flaws and uses material in such a structurally inefficient way that I think it should never be built, but that is, as I've said, just my opinion.

chapmanite

Member
The only exception to this rule has resulted in the use of continuous core in leading edges and main skin panels where the skin sandwich lies over the ribs. This can be seen in several aircraft including Glasairs, the Express and even the Lancair. But in each of those cases the two skins are joined (core removed) over the main spar.

One of the potentially problematic areas of the sandwich over spar arrangement (or similar application) is the mode of failure, which is a function of bond strength and localized crippling, and the resulting panel instability (this was the failure of the KR's skins). Even with today's tools this cannot be predicted to a high degree of accuracy. Given the use of the relatively light cores in small plane construction, skin adhesion may not be sufficient to provide problem free service over the anticipated life of the structure. As such, this mode of construction has been pretty much ruled out by most in the industry.
I realize I'm a Johnny-come-quite-lately to this thread, but this sounds like the sort of problem that in the automotive composites world is addressed with hardpoints. Would it be possible/practical to have a core-thickness solid strip over the spar and/or ribs for the sake of transferring the skin load into the spar without excessive stress on the core material itself?