1. the applied load for each object in bending is the average compressive load... so if the entire object is in compression, it will be just the average load. If part of the object is in compression and part in tension, the tension part is thrown away and not included in the summation. Then the applied load is the average of the compressive part.
2. I'm not sure about bending in the M2 direction. I actually don't think it is taken into account. Most of our efforts have gone into stiffened panels where the stiffeners do not bend in the 2 direction.
3. This is sort of related to the last question. The upper flange is actually considered to be one long object for crippling. The right and left sides are lumped together into one. The length that is put into the crippling equation to calculate Fcc will be 1/2 the total width. As long as there is no bending in the 2 direction, this would be equivalent to handling left and right as separate entities, however in 2-direction bending I can see this as causing an inaccuracy. If this is a critical item, it could possibly be changed.
As a side note, when HyperSizer does STRENGTH analyses for the flanges (such as von Mises), it does take 2-direction bending into account, so it evaluates extreme stresses at the tips for example.
4. We do not include separate equations for formed vs extruded. that is true, but we could incorporate. see below.
5. Yes, I think we could incorporate other methods. Would these be items that we could expose to all users or are they proprietary to your company?
If proprietary, we do have a facility for allowing user-defined crippling. It is not something that most engineers can do because it requires a pretty knowledgable programmer and you have to be able to create a Dynamic Link Library. See this post: http://www.hypersizer.com/dload/wp01_using_plug_ins.htm