Joint Analysis 1

[Stezza]

Hi All,

I need to sharpen the pencil regarding a spar to skin joint analysis.

I have a global fem with CQUAD elms for the spar webs and skins and rod elms for the spar caps and stringers. There is one skin element between stringers.

I already have a freebody force extraction in Patran for the spar elms and one skin elm that gives me P,V,M for the spar and skin (i.e. half of the skin to the first stringer is included in this section cut).

Currently the spar cap to skin joint is stress based on q=V/h (h, section height) - easy and conservative.

I am considering using q=VQ/I to get the running load at the spar cap to skin interface, my questions are:

1. What skin area to put in the Q and I calculations, (a) half the distance to the first stringer (b) half the distance to the rear spar.

2. As the box is loaded in bending and torsion is the VQ/I unconservative for the portion of the vertical shear V due to torsion.

3. If so how could I calculate V_bending and V_torsion from the global FEM.

There are other ways I could stress this joint, i.e. using freebody forces from the skin but I'd like to understand if I have options to use the P,V,M section forces already extracted as this will cut down on the rework.

Thanks

 

[rb1957]

this is a wing, right ?

normal bending is the primary loading, usually. Q and I would be whole wing sections (not local to the spar). Now you can say the rear spar reacts so much shear (50% ?) and so much bending (<50% ... the upper and lower skin/stringers react most of the bending, the front and rear spars only a little). so if you try to make a free body of just the spar cap you'll balance most of the moment with a shear force in the upper and lower skins (a couple, right?). Because the lower skin/stringer panel reacts a lot of the moment, the shear on the spar fasteners is much higher than if you look at just the local section.

another day in paradise, or is paradise one day closer ?

 

[MOHR1951]

Think the best thing for you to do,and I think rb1957 might be implying the same is to get out paper and pencil,draw a FBD,and solve for the unknowns,then apply this high dollar software,after you intimately know the problem.Even if it is an unsolvable problem using statics,strength of materials and perhaps some harmonics/dynamics ,at least it lets one focus on the problems,clearly,you arrive at the unknowns and what's given.Sometimes this is 90% of the problem,then precede with FEA,it gives one a much clearer view of the situation.

 

[Stezza]

Thanks for the feedback Rb, I can see now that as would need to do the VQ/I for the complete wing section, then separate the vertical shear between the front and back spars. I was initially thinking that given I knew the front spar vertical shear I could treat the front spar in isolation and make an assumption about skin effective width. Most straightforward path will be to stress the joint using the freebody forces or Nxy in the skin elements would also be an option.

Mohr, thanks for taking the time to respond but I was looking for an answer for a specific issue that I tried my best to clearly define.

As an aside I never come across a good textbook or company manual dealing with “Detail stressing from a coarse grid FEM”

 

[rb1957]

I never come across a good textbook or company manual dealing with “Detail stressing from a coarse grid FEM” ... and I'd be surprised if you did ! most companies have their own way of dealing with this issue; most will treat their way as right and everyone else is "wrong". Most will tell you don't use the stresses, extract the loads (like nodal loads) and hand calc from there.

you can treat the spars in isolation, so long as you remember the shear into the skins ... and I suspect this should be big enough to scare you !

another day in paradise, or is paradise one day closer ?