Tricky CHS to CHS Connection

[SydStructEng]

Hi everyone,

I'm working with a colleague of mine and trying to rationalise the design for a CHS to CHS connection. His description is included below. Please view the drawing attached.

The design of connections in hollow sections is tricky. This is an example.

We have a small CHS welded under a larger CHS, no internal stiffeners. The small CHS has bending applied radially. It is easy enough to design the small CHS. But the local bending imparted into the wall of the large CHS may fail the wall locally. So we add a stiffener each side, but what design criteria apply, and how far should we carry the stiffener around the large CHS?

AISC Vol 1 for Hollow Sections doesn’t cover the case.

Does anyone have experience and can offer assistance?

Thanks,

 

[hokie66]

I think the CIDECT manuals may have something on this. But I would tend just to stiffen the wall by providing a "saddle" plate wrapping the bottom half of the big CHS. The stiffeners projecting as in your sketch don't appeal to me.

 

[SAIL3]

A saddle pl would certainly help but the distribution of stresses in the large pipe would still be complicated....it would be somewhat similar to saddle supports on pipe/tanks but with a much larger conc moment...
ASME has alot of data on nozzles to tanks with applied mom on nozzles but your D/t's may not fall in the range of their test data...
Rather than try to chase down applicable test data I personally would use a complete ring on the large pipe and slot the small pipe....this would reduce the problem to analyzing a ring with a conc mom which you could find in Roark and then just check for typical shear, bending and torsion in the pipes