Beam moment connection through column 6

mte12 said:

Is this an acceptable method

Click to expand...

Yes

mte12 said:

and is there any literature on this?

Click to expand...

Probably somewhere. But it is an uncomplicated connection and checks for such moment connection apply. Expect shear to be transferred to the column.

mte12 said:

but the web will deform as well.

Click to expand...

That depends on the degree of rotation and the relative stiffness's of the beam and column. To avoid this use big beams. Based on the presume relative scale in your drawing and presuming the back span isn't excessive very little moment will be transferred to the column. And vaguely modern structural framing software should be able to provide you with the amount of moment transferred.

mte12 said:

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

Click to expand...

Short answer is you don't. Long answer, if you are creative with cutouts and spacing shims you could achieve this but that is likely to be a crazy amount of work for little to no benefit.

Thanks for responding.
Agree on the logic of using relative beam/column stiffness. The issue that I have though is that the column is working hard with all other loads, and even though the beam is stiffer, there is still some moment transferred to the column.
With attachment to the web, was aiming to somehow reduce the moment at the column even further, by modifying restraints.

Is this existing or a new design. If it is a new design the I'd just upsize the beam/column and make it work.

mte12 said:

With attachment to the web, was aiming to somehow reduce the moment at the column even further, by modifying restraints.

Click to expand...

What do you mean by this? You mean making changes in your model to get the answer you are after? Yeah you can do that. But you will need to dig deep to ensure you are not being unconservative.

What you might be implying but not saying explicitly, is that because the attachment is through the web of the column the web will deform and not the entire column including flange. This would likely be true if the plates have a 'significant' space between the end of the plate and the radius of the web. Thus you will get less moment transfer than what is represented by your calculations or structural framing software. But again, suitably quantifying this in and applying it in a conservative fashion would be quite a challenge.

I'd either upsize the members or strengthen the members as appropriate.


For what it is worth the I've done something similar before with a heavy load being applied eccentrically to a support beam. (Multiple parties in a project meant that their heavy 20m gantry landed significantly eccentric to MY beam that was provided and already errected) I provided a bolt through moment connection similar to picture with a back span. The beam took the shear load and the back span took the moment induced by the eccentricity.

Thanks for responding again.
Yes, this is an existing column.
Agree it is a challenge, perhaps it may be that the column is strengthened.
Thanks for the comments.

See AISC Design Guide 4... it has lots of information.

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

-Dik

OK thanks, I didn't quite see this type of connection though.

Extend the bolted end plates downward, and add some gussets on both sides of the beam.

mte12 said:

OK thanks, I didn't quite see this type of connection though.

Click to expand...

What you've shown is basically a moment end plate. AISC's DG4 and DG16 cover these.

In my experience colleagues and I have typically ignored the moment that would transfer to the column in such a situation. This seems like one of those self-limiting rotation types of problems.

If you think the weak-axis moment in the column is critical, you could build a shell model of the column and beam and see how much moment ends up in the column. The column web out-of-plane bending stiffness seems small compared to the beam flexural stiffness, so my guess is little moment would go into the column.

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

The only time I use that type of connection is to transfer moments. I may have a couple of these a week. I usually use clip angles, if no moment. See the next reply.

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

-Dik

This is what I do... a common connection.



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

-Dik

I think we're typing about different purposes.

The double angle is intended to transfer insignificant beam end moment to the column or beam on the other side of the column, right?

The connection I sketched is intended to transfer insignificant moment to the column but transfer lots of moment to the beam on the other side of the column. I've never had to resort to this connection, thankfully. LOL

You said that if you had insignificant moment... the connection shown may be more common... I misunderstood your connection. I don't think I would ever try to avoid transfer of a moment to a column with a beam on the opposite side... I'd simply transfer the moment through the column. If you really want to do that, use a flexible connection at the beam on the opposite side of the cantilever. You also have to be careful with shear transfer.

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

-Dik

These are the connection types for stiffened end plate connections I've written SMath programs for. I have about 30 connection types and I add them to a *.pdf conn file that's on the desktop to I can pick the right program (This is a snapshot of part of the conn file) I have over 800 SMath programs for design so it's sometimes a chore in finding them).



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

-Dik

Moment in the column can be safely ignored because, as previously stated, it is self limiting by the beam rotation.

The original detail is fine. Design the beam assuming zero column stiffness and design the column assuming full column stiffness.

Thanks to all for further input.

BAretired said:

Moment in the column can be safely ignored because, as previously stated, it is self limiting by the beam rotation.

Click to expand...

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.

mte12 said:

The issue that I have though is that the column is working hard with all other loads, and even though the beam is stiffer, there is still some moment transferred to the column.

Click to expand...

I might be inclined to take it in the other direction and make the moment connection to the column as stiff as possible such that the beams can be relied upon to provide meaningful rotational restraint to the column at the joint. This likely increases the effort of column analysis a bit but might be a robust way to squeeze more mileage out of the column.

Given that the column exists, I worry a bit about the practicality of erecting your proposed connection:

a) Will the backspan beam be able to be threaded in between the column flanges without excessive shimming or fitup issues? This will depend on the framing condition at the other end of the beam.

b) With bolts common to two end plate connections, will that complicate erection? Given that it's meant to be a beam moment splice, you'll likely not want to clip the end plates and have "erection only" bolts.

These concerns might also tilt me towards a connection that promotes moment transfer to the columns. With some flange plates welded to the column flanges, you might have more erectable options available to you. Of course, you'd also then have to give some consideration to the impact of welding to an existing column under stress.

I know that we have some specialty steel connection designers in the room here. I'll defer to their judgement on the erectability front.

KootK said:

b) With bolts common to two end plate connections, will that complicate erection? Given that it's meant to be a beam moment splice, you'll likely not want to clip the end plates and have "erection only" bolts.

Click to expand...

I've seen the following at similar situations going through girder webs.

1. If the cantilever is short like shown in the OP's sketches, then omit the end plate on the cantilever and shop weld the cantilever to the column web. Try fillet welds on both sides of each flange and both sides of the web. If that doesn't work, then CJP at the flanges. Keep the end plate on the left span. One of my fabricator pushed pretty hard to use this approach on a lot of connections.

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.