Wing Design Day 15
Updated: Aug 19, 2020
Rib and aileron reinforcements
If you’ve been keeping up with our prototyping blog posts, you’ll know that our first rib test didn’t exactly go as planned. We thought the plywood cap strips would provide a lot more strength than they did and our ribs ended up failing at around 100lb, about half of what we had been anticipating. It’s just a good thing that we ended up testing the ribs before installing them. Phew.
In every instance, the leading spar tore out of the front of the rib, breaking the plywood while the aft spar showed no signs of damage or plastic deformation. We had originally designed the wing with the leading spar doubling as the leading edge because it would create a large surface to glue down the fabric to cover our wing. Both the Affordaplane and (a version of) the Blite Ultracub use this method seemingly without issues. However, the Affordaplane uses 26 ribs and the Ultracub uses aluminum, plywood, or carbon fiber ribs instead of XPS (eXtruded PolyStyrene) foam, so they aren’t the best comparisons.
We hadn’t run any tear out or bearing stress calculations prior to these initial tests, as it was almost impossible to find any strength data on the XPS foam. The only reliable source we found was Matweb, but it didn’t specify if it’s shear stress was ultimate or yield, and it was just generally incomplete. Because of this, we decided to undertake a series of tests where we used different sized bolts, and came up with an average ultimate shear strength that was only 3.5% off from Matweb’s shear data point! We then used approximations (ultimate/1.3 [this number was visually approximated during testing] = yield; shear *1.6= bearing) to complete our strength data, so we could calculate tearout and bearing stress. Note: we did our tests with a crane scale, a drill, a ruler, foam, and some screwdrivers, so needless to say: they’re not perfect. We’re going to build our own tensile strength tester soon to confirm/refute our initial data.
To reinforce the ribs we decided to make three main changes:
We moved the leading spar back a inch or so such that the distance from the top of the leading spar to the top of the rib would be equal to that of the aft spar. To mitigate the issue of fabric stretching, a 0.016” sheet of 2024T3 aluminum will now cover the leading edge at the cost of approximately 1.7lb per wing. With the leading spar moved back, the false ribs also needed to be redesigned:
Some other benefits of moving the leading spar back.
The lift distribution between the two spars will be closer to 50/50 (the aft spar had previously been carrying around 55% of the load based on our chordwise lift distribution calculations)
The drag and compression struts will be shorter making them lighter and stronger in compression (column buckling).
The plywood cap strips generally didn’t adhere very well to the foam. This is likely our fault for not developing a better clamping method but, regardless, it peeled up and still broke. We haven’t run any further tests to check the strength of the plywood but after our disappointing test results we all agree that we should try e-glass (fiberglass) next. We had done some messing around with the fiberglass on some test pieces of foam and it seemed to adhere really well and would break before delaminating. Fiberglass is also a lot cheaper than aerospace grade mahogany plywood. We do plan to repurpose the plywood for a mahogany interior though...
The next rib tests will now be reinforced with 0.016” aluminum “slings” to improve the bearing and tearout strength of the foam (pictured below). The interface between the spar and the rib in shear and the failures we observed were all because of tearout. These slings should theoretically increase the bearing strength of each attachment point around 2819.7lb (5639.4lb per rib) using a 1.5 standard safety factor and a 1.15 fitting factor. These numbers are totally theoretical based on the 75ksi bearing yield strength of 2024 T6 and are likely to be a lot lower in practice. If this works, the rib should break in the middle far before the spar holes tear out. We considered gluing thin aluminum washers to either side of the spar attachment but that would rely on the adhesion between the washer and the foam whereas the sling utilizes epoxy as well as its own strength.
We also decided to create a new “support rib” (pictured in yellow below) which will support the aileron torque tube at its mid point. We had tried to avoid using a support rib in the past in the hope that the two regular ribs adjacent to the aileron would be enough to support the torque tube but we now think otherwise. The support rib will also reduce bending in the torque tube which will help reduce the risk of it binding in flight.
Where did all the ribs go? Ollie admits that he got a little too obsessed with the parametric design capabilities of Onshape and he ended up making some design sacrifices just to preserve the variables. One of these sacrifices was deciding to use a uniform rib spacing. Since our spanwise lift distribution isn’t uniform and is actually closer to a quarter ellipse, rib spacing will increase towards the tip of the wing if an optimized placement is used. Until we know the strength of each kind of rib to optimize its placement, the wing will just have to look a little incomplete.