Single Story X-Braces 1

[seemant]

1) Could we really reduce the unbraced length by half if we were to provide a connection in the middle (of a single-story X-brace) and have all the braces break at the gusset plate? The reasoning that goes along with this design is that the tension brace would prestress the connection so much that the compression brace can not buckle along its full length. Do you agree?

2) Please suggest all possible design considerations for the gusset plate in the middle. Its an atypical connections as we must break ALL 4 braces because its an unusually long bay.

 

[SperlingPE]

Can you add a column and use two x braces?
Does the seismic load control? Is the lateral load large?

 

[DaveAtkins]

Why do you even care about unbraced length? Aren't the diagonals designed for tension only?

DaveAtkins

 

[UcfSE]

I would use the full length from corner node to corner node and modify your "K" as appropriate. The tension force may add some stiffening effects but unless you can show those effects quantitatively I don't think you should use it. Saying the tension helps and proving it are very different things.

 

[seemant]

I have a vague sort of a proof from a powerpoint someone gave me. Its not nearly convincing enough though. Therefore, seeking opinions.

Its an essential facility and it has to be an SCBF therefore, unbraced length is a concern. Also, although the seimsic loads are not great, its a pretty long bay about 60' wide and 30' high. We have to split the braces otherwise it couldnt even be transported.

My concern for the gusset plate is, I feel it might kind of warp in this situation.

 

[JAE]

If you are using the X as a compression-tension brace instead of a tension only brace (per DaveAtkins above) then the compression part of the X does indeed get braced by the tension member and reduces the unbraced length.

The gussets you refer to MUST be designed to not buckle under compression. We did an X in South San Francisco once where we used steel tubes for the X and at the crossing had a pretty substantial set of splice plates that provided equal or greater capacity across the connection. Basically, on tube went through the crossing and the other was butt-terminated at the through-tube. Then we laid almost-full width splice plates on either side to lap over the continuous tube. Longitudinal fillet welds were used to connect plate to tubes. Also - butt welds were used at the tube-to-tube interface but were not fully counted on.

Remember to consider Chapter B and the effective areas for tension members.

 

[UcfSE]

If it's a tension only brace what difference does it make what the unbraced length is for the member that would be in compression? Even so, aren't there cases in seismic design where you are not permitted to have tension-only bracing, such as a special concentric braced frame? I seem to recall that but can't say for sure.

 

[seemant]

Its not a tension-only brace. As i mentioned earlier its an SCBF (Special Concentric Braced Frame). In an SCBF the compression brace is allowed to buckle.

 

[AlohaBob]

Suppose the total load is designed to be properly carried by the tension brace. And it is designed accordingly. Now, the code prescribes that this brace under a reversal of load must be designed to carry 10 - 15% of it's tension load in the other load direction under compression. So the member is rechecked to see, how does it's design under 100% tension load compare to it's design under 15% compression load. If the 15% compression load requires a greater section, then increase it. Then when the X-braces work together in tension and compression, the actual design factor of safety is increased because of incidental load carried serviceably by the compression load path.

Now if the design is to be leaned out to it's greatest efficiency, a first iteration reduction in section would be proportional, and then a more rigorous analysis of compatibility would be applied.

I investigated a tower once where the X-bracing was buckled outward at the center gussets. The owners were concerned the tower might be inserviceable. The tower however was still serviceable and would perform as it was designed under designed loads. We did reinforce the joint to bring the braces in-plane again. The tower was overloaded, but it performed as it was designed. Once compression braces buckled, the tension braces were still sufficient to carry even the overload.

If bracing members are to work only as tension-only braces, the connections must not allow the member to be loaded in compression when total tension deformations are accounted for.

It's the repeated stress reversals that become a problem where a member buckles, then becomes strain hardened, then actually breaks. Of course the shear capacity of joints must be sufficient to allow this. That initial design buckling load design would attempt to keep the member within an elastic buckling range.

 

[ronster]

What code now requires tension members to be designed for 15% reversal in load? Is that for all tension members or only certain types of loading or bracing?

 

[AlohaBob]

That would be incidental loading of x-brace members designed for tension but function also in compression. I can't cite the reference from memory right now. I'd need to check my codebooks.