Is a composite spar under balsa is a good idea?

Although I am not a customer of yours, I hope you will help. I designed my first sailplane last winter, a HLG and am building a revised version this winter. The original wing was 1 lb. foam with balsa skins, and a spruce spar with ply shears/joiners like in the Chuperosa. I would like to trim the weight down by using composite spars of either CF or FG. But I don't have any reference designs to model. I have seen large triangles of CF/FG, thin strips of TOW and thinner/longer triangles of CF like in your designs. But these are all using FG skinned wings. Can you offer some advice on whether the composite spar under balsa is a good idea at all, and if so what would be adequate materials and sizing to use? Thanks.

From : Don Stackhouse

Scott,

So, not a customer, eh? Well, why not?? You might want to read our article about "Is scratch building cheaper than building from a kit?" ;-)

Seriously, though, this is another one of those simple question, complex answer situations (seems like just about any serious design question fits into that category!). From what you describe it sounds like your first wing is WAY overbuilt for a HLG. Depending on the airfoil you use, no spar at all (other than the balsa skins) might be sufficient. My old Brian Agnew Models "Vertigo" is built this way, 1/32" balsa skins over foam, 1/4" balsa strip leading edge, no extra reinforcement. Its Eppler 387 airfoil (9.06% thick) is thick enough for this to be adequate. It's possible to break one with an abusive launch, but that's true of almost any HLG worthy of the name. If they're strong enough that they can't break, then they might be good slope models, but probably can't thermal in average conditions anywhere east of the Sierra Nevadas or north of Dallas/Fort Worth (the thermals west and south of those areas are so strong that almost anything looks like a good sailplane, even Volkswagens and kitchen stoves).

If you use a thinner airfoil things get a bit more complicated. Bending stress goes up as the thickness goes down, proportional to the square of the thickness change. Let's say, for example, you decide to jump on an old bandwagon and use the S-4083 airfoil (8.0% thick). Note that I am in NO WAY endorsing that airfoil, you can read my opinions on it in another article in AJ&D. But assuming you wanted that one anyway, you now have a wing thickness that is 88.3% of the Eppler 387 wing. This means that the bending stresses have increased to 28.3% more than the original. That 1/32" balsa skin might not be sufficient anymore. You now have several options:

1. Use higher density balsa. This is very difficult to do consistently, even within one wing assembly. Your wing might not be as strong as you think. Obeche instead of balsa might be a good choice.
2. Use thicker balsa. Much easier to control than #1, but adds some weight, probably a bit more than #1.
3. Add fiberglass, Kevlar or carbon reinforcements to the inside of the wing, just under the skin. Another option is to add it to the outside of the skin, but I'll get into that one in more detail shortly.

At this point you should probably take a moment to read my article "What structural design considerations are needed to reinforce a fuselage? (the second short course in designing composite structures)". Ok, more than a moment, but read it anyway if you haven't already.

At first glance the table of data in that article suggests that Kevlar might be a good choice. Bad idea, for two reasons. One is that although most composites are actually quite good in compression, typically about 80% of their tensile strength assuming they are adequately supported against buckling failure (contrary to some of the misconceptions floating around in model design forums I might add), Kevlar is the exception. Its compressive strength is terrible, less than 40% of that of E-glass in compression. The Kevlar molecule actually has a kink in it, and fails in buckling at the molecular level at a very low load. The epoxy matrix crumbles when this happens, but the fibers stay intact with most of their original tensile strength, sort of like a loose rope. The compressive stresses in the upper wing skin can be very high during launch, and it's also possible to have moderately high compressive stresses in the lower wing skin during launch with certain types of launching styles. It will take a lot of Kevlar to handle the loads. On the other hand, the relatively large thickness of Kevlar required will help prevent buckling failures, so the news isn't all bad (just mostly bad!).

The other problem with Kevlar is its stiffness is a poor match for balsa. Carbon has a similar problem. This means that if you try to use either of those for reinforcements, the composite will try to carry all the load, and the balsa will be along for the ride, its strength largely wasted. This also means that there will probably be very high stress concentrations along the edges of the reinforcements, which could lead to cracking and splitting of the balsa in that area. If the wings had glass skins like the Monarch'D' and Wizard, the material match would be much better. If you are using balsa skins it's probably better to use fiberglass reinforcements to better match the stiffness characteristics of the balsa. This is the reason we used fiberglass for this purpose on the older style wood-skinned Monarch wings.

On the plus side, the balsa skins probably support the spar caps against buckling a little better than the fiberglass skins do, but I suspect that the difference is minor. There are a lot of other factors involved in that issue, such as the characteristics of the foam core and the shape of the spar caps in three dimensional space. If you use thinner material for the spar caps and put half on the outside of the skin and half on the inside, you could see some very meaningful improvements in this issue, but on HLG's it's usually not the determining factor, and this next item will probably dominate the tradeoffs:

Now for the other catch. Remember that the stresses are inversely proportional to the square of the thickness? Consider what happens when you put the composite reinforcements on the inside of a composite skin vs. the inside of a balsa skin. If we're using a thin airfoil like the S-4083, with a root chord of about 7 inches (a fairly typical number for HLG's), the wing will be about .56" thick at the root. With fiberglass skins the skin thickness is miniscule, so the distance between the spar caps will be approximately .56" . If you use 1/32" balsa skins and put the composite spar caps inside them, the distance between the spar caps will be only about .50" (.56" minus the total thickness of the wing skins), so the bending strength of the composite spar caps will be only about 80% of their strength in the glass skinned wing. This strength penalty gets increasingly worse as the wing gets thinner, because the thickness of the skins is an increasing percentage of the total thickness. For example, at 7% thickness the strength will be about 77% of the glass skinned wing, and at 6% thickness it will be down to 73%. You could use vacuum bagging techniques to put the spar caps on the outside of the balsa skins, but at that point you might as well use glass skins to begin with and skip the balsa. I suspect composite reinforcements on the outside of the balsa skins would also look rather ugly, and make finishing of the balsa more difficult.

The overall idea of glass spars under balsa is ok, carbon under balsa not quite so good, and both of those concepts get worse as the wing (or tail surface) gets thinner. As far as dimensions and configuration, that is influenced by a wide variety of factors, and is very dependent on the specific design. There are lots of ways to do it, and most of them can be made strong enough, but some will be excessively strong at the expense of weight, and some might make more efficient use of the material than others. There is no single "right" answer, but some answers are better than others. It really comes down to the question of whether you're just trying to design something that works, or if you're trying to come up with something better than the other available HLG's.

You can try copying other designs you see as a starting point, but it will be difficult to say whether their approach is appropriate for your particular design until you break a few prototypes.

If you have access to finite element stress analysis and the experience to use it effectively, you can get a good understanding of which reinforcement concepts are the most efficient, but the loads on the structure are so variable that you will probably have to do some lab testing plus get a lot of experience in the field before you can really optimize the design. We presently have thousands of our HLG kits flying all around the world, and I'm still learning new things about how they hold up in service.

Lots of decisions, lots of questions, no easy answers, lots of prototyping and testing before you can even begin to feel that you're getting close to optimum. Or you could just get one of our kits, we've already gone through all this for you!

I hope this helps, good luck and please keep in touch!

Don Stackhouse @ DJ Aerotech
djarotec@bright.net