Braced frame connections - modeling issue

[formi]

Hello everyone!

I'd like to find out how should be modeled braced frames members - girders and diagonals in high-rise buildings? As a truss bars - only axial forces acts or they should be fixed?

This is something that is really bothering me. As I look at these pictures I think it should be treated as fixed...

 

[hokie66]

Those bracing frames seem to have some rigid connections and some pinned connections. When the flanges are connected, they are rigid. Where only the webs are connected, then pinned.

 

[mike20793]

I treat them as pinned and prefer to attach them to the webs of the the column with a gusset plate. They are typically only welded to the gusset plates and not the columns/beams. I haven't seen them directly connected like that, but I am relatively green still. I'm not sure your going to get enough rotational restraint to treat those specific connections as rigid.

 

[hokie66]

Be careful about mixing terminology. "Fixed" as in the OP's post means it doesn't rotate, but "rigid" connections can rotate, as controlled by the stiffness of the connected members. "Pinned" connections rotate because the connections are flexible.

mike20793, there are several types of connections in the photos. They are certainly not all pinned.

 

[mike20793]

Thanks hokie; I did mess up my terminology. I meant to say fixed as the OP stated. I agree they are not all pinned. I should elaborate. I should have said the only braced frames I have designed connected to the column web with a gusset plate and I treated them as pinned. Sorry for the mix-up on my part.

 

[hokie66]

The problem is in modern terminology, as used in computer modelling. None of the connections in the photos are "fixed", as the members can rotate. They are either rigid or pinned. The horizontals to column connections appear to be pinned, and the others are rigid.

 

[Lion06]

It makes almost no difference. Any rotational stiffness of the joints can't compete with the axial stiffness of the members in the braced frame.

It's similar to modeling a truss with fixed or pinned connections; the difference is usually negligible.

 

[KootK]

Those braces are gorgeous.

All of these connections will behave rigidly in plane to a pretty high load level.

If we're talking about eccentrically braced frames in seismic country, the moment connection to the brace at the top of the frame may be beneficial / necessary in order to ensure that the structure beyond the plastic hinge remains elastic. Splitting the moment at the joint between the beam and the brace helps in that regard.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[Agent666]

Definitely consider as rigid in this scenario, design for the combination of the moment, shear and axial loads that results from this assumption.

 

[BUGGAR]

With the ease of fixing or releasing joints in computer programs, If I'm not sure of the fixity, I run them both fixed and pinned to see if there are any potential failures in either. Then I make a judgement on the forces to use.

You can use partial fixity, but at what percent? I like to range my solutions.

 

[slickdeals]

Pinned. It makes the designing the connections simpler and may not make any difference in the end results. No difference in braced frames or trusses.

 

[formi]

So many men so many minds...