Plastic Moment - Failing Flanges

[iponom]

Anyone know a good way to design Shear + Moment connection with directly welded flanges with flanges failing in tension at Full Pen welds?

Not sure how to distribute the rest of the moment in the beam web - and not too excited about doing it at all.. Constraints: No flange plates, no extra field welding.

 

[nutte]

What are the test results?

See Huang, et al. (1973), Krawinkler and Popov (1982), and Beedle, et al. (1973).

If an EOR would not approve this standard connection, used successfully for a long period of time, approved and recommended by AISC, and supported by testing and documentation, I would say that EOR was uninformed. If he wouldn't approve it after the above evidence was presented, I would say he was unreasonable.

 

[iponom]

nutte,

I am a recent graduate and I was wondering what do you think about strain hardening as a design limitation? In what situation would it/not be appropriate...

Another interesting journal on that topic:

"Truss Analogy for Steel Moment Connections"
Subhash C. Goel, Bozidar Stojadinovic, and Kyoung-Hyeog Lee

 

[ToadJones]

iponom-

The beam would begin to yield at the yield moment, My=FySx.
Beyond that, the stress distribution on the cross-section is no longer linear. At some point the entire cross-section would in theory reach Fy, not just the outer fibers or just the flanges--> plastic moment Mp=FyZx.
Your calculation of the force in the flanges is more of an analogy to a force couple than the actual summation of force in the flanges. The flanges will not carry the entire moment. In reality, there will be some interrupted stress flow from the flanges into the web.
In-other-words, you are doing a very rough calculation of the force in the flanges and expecting exact results.
Furthermore, it is okay to follow the code in this situation as you have already predicted loads and are dividing by a nice cushy factor of 1.67.

 

[frv]

iponom..

As others have stated, the code has approved this connection. They have done it for a reason.

I think you are looking for a physical description of what is actually occurring, so what follows is a summary:

In a fixed fixed connection, the ends of the beam begin to yield first. If both your flanges and your web were welded to the supporting member, then you would develop (theoretically) the plastic moment capacity at the joint.

In this case, however, only the flanges are welded to the supporting member so you get somewhat less than the plastic moment at the joint. However, the maximum moment is still at the support, so your beam has more capacity slightly away from the support. the joint will begin to rotate and the plastic moment will spread away from the joint.

This is a fairly long way of saying that there is ductility in the connection which allows the true plastic moment to form close to, but not exactly at, the joint. If you pick up a steel design textbook and look at plastic design, you'll see what I'm referring to (not to be confused with plastic moment capacity of a cross-section).

 

[abusementpark]

I know people have referenced previous editions of the manual in earlier posts, but can anyone tell where this issue is discussed in the 13th edition?

Also, to add to this discussion: I believe the reason you can rely on strain hardening in this connection is because it only over a small distance (i.e. the distance that the beam flange overhangs the web to make the weld).

Consider a tension member with bolt holes; it is for this same reason that you check rupture (which relies on strain hardening)on the net section and yielding on the gross section. Yielding occurs at the net section before rupture occurs, but since the deformation is so small because it is over such a small length (think PL/AE) that it is not significant enough to be considered a limit state. You can think of this overhanging flange as a tension member.

 

[nutte]

They didn't put it in the 13th edition. That doesn't change all the testing that has been done showing that this connection develops the plastic moment.