Thanks for taking the time to answer my question so extensively. I completely follow your explanation, with two question marks:

Step 1 is where I keep hearing alarm bells. And I have problems finding the words to describe it, please bear with me. Maybe the analogy with a false spar would work. A false spar is not clamped at its root, it's probably not even pin jointed. Yet in a beam analysis, running from the tip to the root, that false spar is assumed to be continuous, and that makes it in effect a clamped beam. You cannot simply leave it out of the analysis, it's there, and it adds to E*I, and it carries a small part of the lift load. I just checked the index of my 10+ books on aircraft structural analysis, and none mentions 'false spar', so that did not help me to understand or explain it better.

Step 5 assumes a main spar weak in torsion, correct? For a box spar, it would be a statically indeterminate problem. That still can be solved, but it's a a lot more work.

Rob

OK, let's remember that structurally, the outer panel can be isolated for purposes of structural analysis. Once we know the loads into the outer panel and how it is constrained you can get to your answers.

Starting with how airloads are distributed, we can work the lift, shear, bending moment, and torsional moment. The outer panel is a big beam with several important pieces: Skin, Main spar, Aux spars. At any spanwise section, you can work up the stiffnesses in shear, bending, and torsion. In sheet metal, you might recognize them as G*webi, E*Ixx & E*Iyy, and G*J. In wood and composites or with mixed metals and using plate theory, they are A66, D11, D22, and D66. If you know the configuration and loads at each spanwise cross section, you can work up the local deflections and work to the total deformations in shear, bending, and twist.

Sure, all of the wing skins and spar pieces count in doing this math. For my airplane I have run all of these numbers. In wing tip bent up bending, the main spar is 72% of my bending stiffness, the skin is 27%, and the drag spar is 1%. OK, how about bending the tips aft or forward? Main spar 4%, skin 71%, and drag spar 24%. And anti-drag is what, a quarter of lift loads? Then we have torsional stiffness, 11% main spar, 22% drag spar, and 67% wing skins.

Now let's contrast that with assumptions. Size the main spar for all bending, size the skins for torsion, and size the drag spar to react torsion and drag/anti-drag to the next structure... Hmm. Use the simple rule and the main spar gets over spec'ed, everything else is no problemo on stress. We do have to bump cap thickness and shear web until it survives, and we can seek out a min weight combo of web and caps. Look at local strains in each direction and sum them up to get deflection starting at constraints, and we see how much our wing bends and twists.

Then we can constrain our outer panel as we think the real world does and see what happens. The assumption we have mostly been working under in this thread is main spar constrained against bending and drag spar pinned to center section drag spar.

While all this fancy stuff is neat, it can let you make sure your outer wing panel stays together, let's you forecast changes to wash-out, what does it do for the drag spar discussion? Whether you do this using composite plate theory and a lot of Excel or by making up FEA models, if you constrain your wing with the main spar connections against movement in all three axes and constrain the drag spar against up-down and left-right movement, one of the things that comes out is loads in all three directions at each of those bolts. The subject of this thread...

The main spar is torsionally soft. In my bird, as cited above, the outer panel at connection to the center section has only 11% of its torsional stiffness from the main spar... The skin is 2/3 of the outer panel torsional stiffness, and the very conservative airplane design professors tell us to just assume the wing skins carry all of our torsion. We may over spec the skins this way, or just keep twisting of the main spar to small levels. Do the full up composite plate analysis and/or FEA, and get all of the loads and deflection considered. Go a little further and look at constraints again, and torque from pitching moment is tough to get into the center section unless you stick in a connection at the drag spar. Yes, you can make a big torsion box main spar for both center section and outer panels, bolt them together. Look at Long EZ and derivatives.

But when you use an aux spar or two to anchor the outer panel against torsion, it can become a straightforward exercise of calculating loads as I summarized in earlier posts. Or do the full up anaylsis that puts in all of the stiffnesses, distributes loads accordingly, and outputs loads at the constraints. Then throw a FOS of 4 on the pin, check stresses in the fasteners, bushings, webs for adequate FOS, and move on to your build...

Billski