AISC Design Guide 4 - Extended End-Plate Moment Connection Design - Beam Web Capacity Calculation

[Illbay]

Performing this design calculation, I was struck by the strange "murkiness" of the requirements for shear capacity of the beam web. The pertinent explanatory paragraph of the Design Guide 4 (Second Edition, published March 2015) is as follows:

"Only the web to end plate weld between the mid-depth of the beam and the inside face of the beam compression flange may be used to resist the beam shear. This assumption is based on engineering judgment; literature is not available to substantiate or contradict this assumption." [Paragraph 2.1.8]

I was struck by the incongruity of this phrase in comparison to most modern design literature. "Back in the day," when I was getting started in the profession - and so much of research had yet to be done - one frequently encountered such statements. However, in the decades since (yes, I'm old), the burgeoning amount of research not to mention the availability of sophisticated computer modeling techniques and products, has reduced much of the uncertainty one used to encounter in recommended design practice. I actually studied as an undergraduate with the late Prof. Krishnamurthy, who originated the end-plate moment connection research, and I recall that these types of "unsubstantiated assumptions" existed then - but that was in the late 1970s! Surely we have dissipated some of that fog by now!

Does anyone know of any research-based recommendations on this topic? As it is, I'm having a very tough time with this design, and it simply makes no sense that there is a shear failure at this point given the practical sizes of the members I'm coming up with based on the other criteria. And failing that, I am minded to add a doubler plate to the web in the "compression zone" of the web, to deal with the ASSUMED required shear capacity. Any opinion on that?

"No one is completely useless. He can always serve as a bad example." --My Dad ca. 1975

 

[JoshPlumSE]

I don't know of any research on this topic. But, I'll say this: It's a conservative assumption based on engineering judgment. Reasons why I agree with this judgment:

1) Webs are stiffer to resist shear loads than flanges. Alot stiffer.
2) Flange welds certainly can take some of that shear resistance. However, if the flange is in tension, then we generally don't want this to occur. So, might as well discard any resistance this provides.
3) There has been a lot of testing of these connections under seismic level deformation / ductility demand (see Virginia Tech papers used as references for that design guide). And, this testing demonstrated the adequacy of these connections.

Let me understand your issue better. Please confirm that:
a) Your beam webs are adequate to resist the demand shear.
b) That demand shear is based on what design force. The shear in the beam based on the beam developing plastic hinges at both ends?
c) You cannot make the weld between the beam web work to resist this force? Using fillet welds on both sides? Certainly, you could make it work if you were to use a full pen weld for the web. Right?

Can I guess that you're using RISAConnection and that there is a "base material proration factor" that is decreasing the strength of your weld by a significant amount?

 

[271828]

The same recommendation is in Design Guide 16, Page 12, Note 10.

Note 9 is probably providing a glimpse at the reason: "Beam web to end-plate welds in the vicinity of the tension bolts are to be designed to develop the yield strength of the beam web unless the full design strength of the beam is not required. ..." The web is very highly stressed in the vicinity of the tension bolts. I think the authors are just trying to make sure that region of the web isn't being counted on to provide shear strength also.

Soon, the new moment end plate design guide will be available. I've heard that it's enormous and very detailed. Maybe there will be a more detailed explanation.

 

[Agent666]

Usually, you'd work back from the full tension capacity from a shear/moment interaction relationship to determine the shear that can be supported at the same time as the web reaching yield in tension. Most codes have some form of shear/flexure interaction.

Typically there is no interaction with full Web capacity in tension until you're over 75% of the sections shear capacity for an I-SECTION. Exact percentage obviously depends on code formulation. But I've never heard of only limiting the web shear capacity and force to being carried by the welds closest to the compression side of the connection.

I'd feel pretty comfortable working out the resultant weld shear and tension and sizing the welds to suit for this resultant force. If uneconomic for a fillet, then a complete penetration butt weld obviously give you the same capacity as the web. We don't follow the aisc design model where I'm from, we mostly follow the more recent design guidance out of the UK by SCI for moment end plate design which is codified in Eurocode 3 (search for SCI398 for a freely available copy). Though I don't think it specifically covers beam shear, likely because the assumption is the transfer of shear to the end plate is over the full beam depth like any section design, and not concentrated over part of the beam deoth as per what you've noted from DG4.

https://engineervsheep.com