Wide Flange Column to Boxed Column Using Plates

[marinaman]

I have a wide flange column that I want to turn into a closed shape. I'm going to plate across the flanges so that the column becomes a box, as I need increased strength and stiffness in the weak direction.

Can anyone direct me to a good design resource for determining the amount of weld needed to weld the plates to the column flanges? I know the plate sizes I need. I've calculated the combined section and reviewed for axial and biaxial bending....but now....I need to determine the amount of weld needed to fasten the plates to the column, and I do not have a good design resource.

 

[BAretired]

Can't go wrong by providing enough weld at each end to fully develop the added plates, then nominal welding in the middle section.

BA

 

[dhengr]

Marinaman:
Since you couldn’t be bothered to give any of the important engineering info. needed to really evaluate the problem, who knows? BA’s advice is always good and is about the best you’ll get with all the secrets you are keeping. The loads, moments and how they are applied would be interesting design info., as would be col. length, end conditions, member sizes and thicknesses, Fy of the parts, and the intended pl. to flg. tip detail. The problem involves more than just the new section properties. You might also want to check your new pls. for local buckling and your whole new col. for stability. Provide some weep holes down at the bot. of the col.

 

[KootK]

I don't know of a thorough design guide of the sort that you're likely seeking. I have this but I've deemed it too time consuming for practical application: Link. It is fascinating theoretically, however, in that it addresses the interesting question of how much weld do you need to prevent pure compression buckling when that buckling technically produces no weld demand until after movement is initiated. Neat. Anyhow, here's what I recall having done in the past:

1) You don't want the add on pieces buckling ahead of the composite member so you keep your kL/r on the new pieces less between welds less than the kL/r of the composite section with some margin of safety.

2) Often you need some of the axial load to be moved from the existing section into the added elements of the new composite section. This should happen as quickly as possible at both ends of the member. Lengths of continuous weld as needed to get the job done. Similar to BA's point.

3) Do the usual VQ/It business for the actual flexure that you need to design for. This will be your bulk stitch welding.

4) Address the weld demand created by compression buckling by throwing a lateral load of 2-5% of the axial load on at mid-span and then back to VQ/It for that load. It really doesn't take much to get this job done.

If your applied moments are anything at all, #3 will dwarf #4 and you can sleep easy on that.


I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[marinaman]

I appreciate the responses.

As much as I have looked around, in my AISC books and guides, my steel design library here in the office, online, and other places, I would think that there would be a design guide or procedure developed and written down by somebody, somewhere, about box plating a wide flange column.

I did find, in the 14th edition of the Manual of Steel Construction, in section 16.1-37 and 38, under built-up members, under section 2, some weld length requirements at the ends of plates and along the length of plates.....but neither of those are based upon loads....and really, are only minimums based upon geometries.

 

[SAIL3]

this may be a real stupid question...but here it goes.....what is the minimum conn required between the plates and the WF for it to act as a composite section?....I would use the applied axial load in the col and solve for the allowable moment in the interaction eguation...with this allowable moment I can backtrack and solve the VQ/It....to check if this moment is reasonable I would use KootK's #4 or similar.....