Prying Action on Round HSS Endplate Connections

[MGaMart]

I'm curious how to approach analyzing this connection scenario. The connection I see most often (and the focus of this discussion) is a square end plate with bolts at each corner. In other literature I've read, the development of the hogging plastic hinge is dependent on the connection geometry. For T connections, the hinge forms at the face of the stem; for rectangular HSS, the hinge forms on the interior face of the HSS chord wall (at least for design purposes). In terms of Round HSS, the location doesn't appear to be addressed, or even design procedures for Round HSS in general. Any thoughts on this?

 

[DCBII]

Look at AISC Design Guide 1 for base plates. They assume for round column base plates subject to uplift that the yield line forms at about 80% of the depth of the column.

 

[MGaMart]

Well that certainly is the closest approximation I've seen yet to what I am looking for. Thanks DCBII. I wanted to try and relate back to the formula for prying action (t_min=sqrt[4.44*r_ut*b'/(p*Fy)], specifically the definition for b'. Using 80% column depth in my mind seems quite conservative (but who am I to pick a fight with the AISC). The formulae derived in AISC DG1 was intended for W columns, and for only 2 bolts as shown in Figure 6, hence my hesitation to use it for a 4 bolt Round HSS connection (despite the fact that DG1 says it's acceptable). In the case that the Round HSS is subject to bending and only half the plate/ bolts are subject to tension/prying effects, is it reasonable to assume the plastic hinge that develops is one straight line (parallel to the bolt line), or perhaps multiple hinge lines which more closely follow the critical section of the HSS-plate contact region (similar to a yield line analysis)?

 

[DCBII]

Here's an interesting study I dug up on this several years ago:

Link

Take a look at Figure 7.12 and then at some of the yield line patterns in Figure 7.13. Notice the hot spot in the corner of the column on some of them. What's happening to the weld at that location? Most of the yield line patterns assumed by engineers for base plate design are inaccurate, but everyone does it. In reality the load will take the stiffest path to a stiff element. The yield line will form perpendicular to this path. As engineers we like square bolt patterns because they're hard for the contractor to screw up, but if you keep the bolts within the flange width (which usually requires a rectangular pattern) you ensure your yield line assumptions are valid.

Next time you're at Chili's see if you can find a "Chili's To-Go" sign that someone bent with their bumper. I've seen a few of these where the bolts were spaced so far away from the post that they looked like some of the plates in Figure 7.13.

 

[MGaMart]

Thanks for the link DCBII. I've since found the literature I've been looking for. There is a design procedure created and verified by Igarashi (1985) which specifically looks at circular HSS Flange plates. The article you provided is an interesting read. Use of yield line analysis is always a last resort for me, namely for the reason you stated; choosing the correct path is subjective and subject to error. Upper bound theorems like YLA may lead to a false sense of security in selecting adequately thick plates for the connection. Thanks again for your input on the matter.