Weak Axis SCBF Column Connections 1

[JLNJ]

I cannot find where weak-axis column connections for SCBFs are specifically disallowed by the manual. Am I correct to presume that this is permitted? All the connection design examples seem to be with the braces connecting to the column flange. Why?

What are the pitfalls to look for when connecting to the column web? I presume there are design requirements which make the web connections not a practical choice.

 

[sandman21]

The reason you dont see any examples is because they are not the typical connection. A weak axis SCBF connection is not prohibited. You will need to do the typical checks of the web for yielding, rupture, etc. Overall you will need to ensure that the columns are design to handle the moments associated with the drift of a SCBF. Elastically we know that loads will be transferred like a truss, tension and compression. However, post-elastic behavior we know that we will see frame action in the connection and performance of the SCBF, this should be considered when designing the columns.

 

[JLNJ]

My questions with (very prompt and concise) answers in italics from Larry Muir at AISC:

In a SCBF, are connections to the weak axis of the columns permitted? Yes.

I do not see anything in the Provisions which prohibit it, but the examples, papers, testing info, etc. all show braces with connections to the flanges of wide flange shapes. Presuming that such connections ARE allowed, what are the pitfalls to look for when designing them? There are no specific pitfalls. This issue - the misperception that web connections are not permitted because all of the examples are to the flange of the column – was discussed at a recent meeting of the Manual Committee. One reason examples are done for connections to column flanges is because they allow more considerations to be demonstrated, since eccentricities must be accounted for that result in normal forces at the interfaces. If you are looking for examples to column webs, there are some included in AISC Design Guide 29 (a free download for members from

http://www.aisc.org/dg),

though none of the web examples are designed to meet the Seismic Provisions.

From the dearth of examples, must be some pretty compelling, practical reasons NOT to connect to the column web. I'm guessing it has to do with the post-buckling drifts. I have a small-ish structure in SDC D, a portion of which has a single-bay penthouse. Again, examples are done for connections to column flanges because they allow more considerations to be demonstrated. When writing AISC Design Guide 29 (a free download for members from

http://www.aisc.org/dg)

Bill Thornton and I likely would have opted to include only connections to the column flange as well, for similar reasons. Some, fearing we would inadvertently cause engineers to avoid connections to column webs, pushed for us to specifically address the web conditions. There is no specific, or suspected, problem with brace connections to column webs.

Please let me know if you have any further questions.

Larry S. Muir, P.E.
Director of Technical Assistance
American Institute of Steel Construction

 

[BUGGAR]

Whatever happened to those two-direction moment frames with a shared cruciform column between them?

 

[sandman21]

Conxtech?

 

[BUGGAR]

These were fabbed up from a split WF welded to the web of another WF in a cross shape. I think the idea waned when test qualification of moment connections became popular??

 

[sandman21]

You can still utilize them AISC 358-10 2.2b(4). SidePlate has some connections plates which utilize cruciform column. For most buildings it is easier to layout frame to prevent sharing systems.

 

[JLNJ]

I had looked at the Conxtech system, but it's for a rather narrow range. They require concrete-filled 16" HSS columns.

The cruciform columns are certainly allowed as sandman points out. I am anticipating that a braced bay will yield the most bang for the buck for me.