Web Panel Zone Shear 4

[BadgerPE]

I have a situation where I am reviewing a set of PEMB drawings for a building that was purchased and delivered over 10 years ago. The owner at the time elected to not erect the building, but new owners would like to put the building up. The original PEMB supplier went out of business so we are in the process of verifying that the drawings conform to current code requirements. This is actually a pretty simple building because there are no web tapered members so it makes the verification process much less rigorous.

The structure is a simple frame that is an addition to a larger building and is in SDC A. The roof is a single slope and we have designed the existing building to connect to the addition so that there is no differential deflections between the two frames. All members check out structurally, however the columns do not have sufficient web strength to resist panel zone shear effects because there is only moment acting on one side of the column. The connection of the rafter is a flush end plate on top with an unstiffened extension on the bottom. Moments induced by gravity are significantly higher than moments induced by lateral loads. The provided web thickness is 3/16" and the required thickness is 5/16" so I am designing diagonal web stiffeners to strengthen the web zone. This is about the only option as transverse stiffeners are already in place at the beam flange locations, so a web doubler is probably not a viable option.

I find it hard to believe that this connection is as under-designed as it is because loading of the structure went down under new code requirements (old code 30 psf SL; new code 23 psf SL). I am trying to figure out how this connection could have been designed originally (I know that most PEMB designers use some form of black magic to make their buildings work so I can determine if I am making an incorrect assumption in the capacity of the connection.

While reviewing AISC J.10-6 there is a statement that "When frame stability, including plastic panel-zone deformation, is considered in the analysis;. Using this statement increases the capacity of the connection. What does this statement actually mean? Would this be a situation where for seismic purposes a frame was designed so that the beam would have a plastic hinge type failure near the connection? If that is not the case, how would one go about determining if a plastic panel-zone deformation was considered in the design? For this case, assuming that plastic panel-zone deformation had been included only gets me a 15% increase so column web stiffening would still be required. I would just like to better understand the intent of AISC requirements.

 

[JoshPlumSE]

The commentary does state "Equations J10-9 and J10-10 limit panel-zone behavior to the elastic range. While such connection panels posses large reserve capacity beyond initial general shear yielding, the corresponding inelastic joint deformations may adversely affect the strength and stability of the frame (Fielding and Huang, 1971; Fielding and Chen, 1973). Panel-zone shear yielding affects the overall frame stiffness and, therefore, the resulting second-order effects may be significant.”

In my opinion, a good amount of inelastic behavior can be relied upon for the design of a typical column panel zone. As the panel zone yields, the axial force of the column gets re-distributes to the flanges (which is essentially what eqns J10-11 and J10-12 are doing). Therefore, these equations are actually reasonable to use from a capacity standpoint. The question then becomes whether or not the increased deflections (associated with panel zone yielding) has been accounted for in the stability analysis of the frame.

Different engineers will answer this question in different ways. Some engineers will say that if you use centerline modeling in your analysis, then it adequately addresses panel zone deformation. And, there is some commentary in the NEHRP documents to back that up. However, I'm not entirely sure whether the NEHRP commentary is referring to elastic panel zone behavior or not.

If you believe that you need to account for some inelastic softening of the panel zone then you would want soften up that joint in your analysis model in some way to account for this. Though I would think this would still be an elastic analysis model... just where the stiffness of the joint has been artificially reduced to a level that better represents its stiffness after it has experienced some yielding.

Personally, I don't really see people use this joint modification technique in their analysis models. But, I do see people use the NEHRP commentary to justify using J10-11 and J10-12. Though I cannot conclusively state whether it is fully appropriate or not.

My biggest concern with using this for your project is that typically these equations (J10-11 and J10-12) have been used for IMF or SMF frames where the joint / connection has been designed for the maximum probably strength of the members... accounting for actual yield strength and strain hardening and such. Force levels that only exist during a major seismic event. If you rely on these equations for regular load conditions, my fear is that you will get too much inelastic behavior at service level loading.