Beam moment connection through column 6

[mte12]

I wanted to know if this is an acceptable detail, for a beam to extend past a column, after which it acts as a cantilever.
Connection is through the web with thick plates.
Intention is for the moment passes through the beam only ... but the web will deform as well.

Is this an acceptable method and is there any literature on this?

Also how would you apply restraints so that moment is not induced in the column at this location?

 

[KootK]

2. If the cantilever isn't short, then use the end plate on both sides and add one extra row of bolts on one side for erection bolts.

Sweet, there's my professional development for the day. Thank you.

 

[271828]

LOL. Cool. I've learned a lot from your posts around here, so there's a little payback!

 

[human909]

I get what you're saying with that but, strictly speaking, I don't believe that the self-limiting nature of the problem alone guarantees that column will not feel any adverse affects from the moment coming into the joint. I believe that it is the stiffness of the beams that shield the column from joint moments.

Agreed. In absence of the full picture I the statements from BAretired and Tomfh are generalisations that can be unconservative.

As I said in my first post, it should be established that the moment transferred is minimal. Judging by the relative sizes of the column and beam this might be the case but that also depends on the load and the back span neither of which is described.

If you really don't want to have significant moment transferring to the column, you could try something like:

Yeah that is what I meant when I mentioned to get "creative with cutouts and spacing shims". But again I'd expect something like that to be a last resort choice. Go bigger beam or strengthen the column would probably be simpler and not cause every future engineer passing that detail to go "WTF!?"

 

[Tomfh]

I the statements from ... Tomfh are generalisations that can be unconservative

How come?

 

[mte12]

Thanks for additional comments.

For information backspan is larger compared to cantilever.

Perhaps interface with column needs to be looked at with FEA. My hope was that a paper was available to provide some confidence in the additional stresses.

Also, I presume it's not possible to input retraints such that moment at this particular location is either zero (or less than full section capacity), from the connection in question.

 

[human909]

How come?

For all the reasons previously highlighted. The load WILL impart some moment on the column so ignoring this completely could be quite unconservative. How much moment the load will impart on the column depends of the load and the relative stiffnesses.

Perhaps interface with column needs to be looked at with FEA. My hope was that a paper was available to provide some confidence in the additional stresses.

As mentioned unless your plates are significantly narrower than the flange width of the column then I believe rotation of the column will be very similar to the rotation of the connection. AKA the moment transferred will be in accordance with the relative stiffnesses. FEA won't assist, a standard node and member analysis program should be sufficient.

Also, I presume it's not possible to input retraints such that moment at this particular location is either zero (or less than full section capacity), from the connection in question.

It it should be possible. Though naturally it will depend on your software. I go my software to remove all moment transfer but not shear simply but making multiple nodes and then constraining the nodes together in the vertical plane. (Essentially making the column only connected to the beam in the vertical direction, so no moment transfer.)

A slightly more nuanced approach and you could do the same with a spring and only transfer some moment.

But this really is software specific so I can't advise further.

My quick model with zero moment transfer and only shear.

Same model again but with moment transfer as your would expect.

 

[mte12]

Thanks for explaining further.
When I first looked at it my first impression was that yes there would be some rotation and bending in the column web, agree that the wider the column the less pronounced. But how to quantify? For this reason FEA was mentioned.

 

[Tomfh]

For all the reasons previously highlighted. The load WILL impart some moment on the column so ignoring this completely could be quite unconservative.

Yes I get that, which is why I said to design the column for the fully rigid scenario, as you've shown in the second model.

 

[dik]

Why the end plate connection, as opposed to clip angles, and the bolted splice at the backspan?

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[human909]

Yes I get that, which is why I said to design the column for the fully rigid scenario, as you've shown in the second model.

Sorry Tomfh. I misunderstood. I takeback what I said there. Your approach seems suitable and conservative.