built up column weld

[tmr4791]

Hello,

I am designing a built-up "cruciform" column as part of a rigid frame (using R=3.0) to resist wind and seimic loads. I am uncertain of how to design the weld of the WT webs to the web of the continuous wide flange column. I am thicking about doing a shear flow calculation for the bending forces in the column and sizing a fillet weld on either side to the WT webs. Is there any other forces I need to add in?

I noticed that the seismic design manual (section 2.3.2b) calls for a zone extending 12" above and 12" below the beam where the WT's are attached with a CJP weld AND a pair of reinforcing fillet welds. Is this recommendation a good idea (or just extra expense), if we are in a low seismic area and not using a special or intermediate moment frame? I appreciate any comments.

 

[InDepth]

You are on the right track. The end welding is definately necessary, because the local forces will be fairly large in the local zone of the connection. The 12" in my mind is a minimum to prevent fracture initiation and propogation due to an overstresss (tensile tugging on weld due to shear developed by moment).

One item: Why can't you use a weak axis moment frame for the second direction? From what I remember if you are using ordinary moment frames in a low sesimic design category then you could use weak axis moment frames (with some exceptions). Are you stiffness controlled?

As a side note, weak axis MF connections perform better than strong axis connections due to the force flow (emphasis on connection performance! not stiffness of system performance!). There is an AISC article on this.

 

[tmr4791]

Thanks. The frames are stiffness controlled, so a weak axis moment frame will not work for drift criteria. I am wondering if, instead of full pen welding, we could size fillet welds for 5/8 of the thickness of the WT web in the zone around the beam. This should develop the full strength of the web of the WT. To eliminate stress concentrations, we could add full depth stiffeners to close out the crucifrom at the flanges of the beams. Does this seem reasonable, given that we are in a low seismic area?

 

[InDepth]

Unfortunately, our codes don't directly address fracture toughness, and rather provide details that usually mitigate fracture initiation and propagation. The intent of the CJP with a reinforcing fillet is to eliminate any "crack" propagation. i.e. if you only use a fillet on either side of the web, you are essentially embedding a "crack" that may propagate through the weld during cycling. In essence, even though strength may work, fracture toughness may not. You may be able to argue it out using fracture equations and the actual stress levels (assuming you have low stress).

The full depth stiffeners are a good idea, and I always include them on any moment frame for load continuity, even if you can get away without them in low seismic zones (designed using the concentrated load section of AISC). I don't think you can use them to aid in web shear transfer forces.

I think you should CJP and provide a reinforcing fillet. The columns will probably be shop welded anyway.

Have you considered a box column? or arranging two columns with the column flange of one nested inside the web of the other (This could eliminate the cruciform)? ...or offsetting the two columns?