Structural steel - slotted pipe (round HSS) end tab

[ATSE]

Consider a structural steel pipe 3" diameter to 8" diameter, with slotted pipe end tab connections. Also termed knife-plate connection, where the end of the pipe is slotted. For instance, say 9/16" wide x 6" deep slot in the end of a 5.563" pipe to accept a 1/2" x 7" wide plate, which is welded with (4) fillet lines.
The new AISC specification classifies this as Case 5 in Table D3.1.
The code infers, but does not explicitly require the length of the slot to be at least as long as the diameter of the pipe. That is, Lw >= D.
Is it a code violation to use Lw = slot length = weld length = (0.85)D? Is it a bad idea? For the connection in question, the axial demand on the connection is a little less than 25% of the axial strength (phi-Pn) in tension.
I would just use a shear lag factor = 0.6 to penalize the connection since the member slenderness controls the design (not the connection strength).

 

[connectegr]

The recommended slot length is primarily a shear lag issue. For a HSS Round using weld length equal the diameter of the pipe, will give a U = 1.0. However, this is only a recommendation. If you are considering the appropriate reduction in net capacity, you are fine.

http://www.FerrellEngineering.com

 

[nutte]

To get to U=1.0, you need a weld length equal to 1.3 times the diameter. At weld length equal to the diameter, you get U=0.682.

The original post asks about using weld lengths less than the diameter. I wouldn't have a problem doing this, as long as the U (equal to xbar/L) is calculated accordingly.

 

[connectegr]

Nutte is correct, the recommendation is 1.3D in 13th Edition. Table D3.1 (p16.1-29) I mis-spoke in my response earlier. The 1.3 was based on research which was published in the Engineering Journal.

But, you can calculated U as shown in the table.

http://www.FerrellEngineering.com

 

[ATSE]

Connectegr,
Thanks for the pdf posting. However, it looks like page 138 - which probably includes the Conclusion - is missing. If you could re-send with the Conclusion page, that would be sweet.

 

[connectegr]

attached

http://www.FerrellEngineering.com