AISC Section J10.2 and J10.3 1

[SteelPE]

I am looking over AISC Sections J10.2 and J10.3 of the 13th edition manual. I would be using these sections to check the requirements of a column used in a beam/column moment connection. At the end of each section the AISC recommends a pair of transverse stiffeners or a doubler plate be used to increase the strength of the column web. How do you determine which is better to use, stiffeners or doubler plate?

From talking to fabricators I realize that stiffeners a a much more cost effective solution. Based upon this information, I assume that the AISC allows a doubler in the odd case that a connection does not need stiffeners but needs a doubler (so why use both). Is this correct?

 

[dhengr]

SteelPE:
When you guys pose your questions the way you have, without attaching the few pages of AISC, Ed.13, which you are looking at, you stand a good chance of eliminating answer/comments from some of the people you might most like to hear from. It seems more and more common here on E-Tips that people ask about formula (J10-3.a.7, sub. 3.g) without even mentioning the basic subject matter, not exactly your case here. And, many people with some knowledge of the matter, may not have that Ed. 13 in hand, so they will tend not to participate. If you don’t have the time to ask your question in a complete way (un-lazy way?), why should we go out and buy the book to participate? I think there have been several other thread on this general topic in the past.

In any case, it seems that you’ve answered your own question by talking with some fabricators. You want to get those concentrated forces into the entire member, actually distributed to both flanges through the web, without buckling its web, acting as a flat plate. And, you want to do this in the most economical way for the fabricator, as long as your analysis justifies that particular approach. There may be instances when one or the other approach is the better detail, although not the least expensive. Obviously, increasing the web thickness will reduce the likelihood of web buckling while transmitting the flange forces from the beam to the column. But, the web doubler pl. is tougher to fit and weld properly.

 

[SAIL3]

I typically use stiffeners as it results in a more integrated conn...the only reason that I can think of also using a doubler pl in conjunction with stiffeners is if the web is overstressed locally in shear from the tension/compression force in the stiffeners...and if so, I would use a diagonal stiff instead of a doubler pl....get a copy of Omar Blodgett's design of weldments books...best investment I ever made..they are available from Lincoln Electric web site...

 

[SteelPE]

dhengr:

I worded the problem the way that it was to try and eliminate the people who are unfamiliar with the AISC 13th edition manual. I also don't want to photocopy/scan 5 pages out of my AISC manual as I am not a large shop and that would take some time to complete. The material also has a copyright.

Sail,

We are pretty much thinking the same thing. Stiffeners whenever possible and doubler when they are required for shear. I have thought about using a diagonal stiffener as you have suggested however, I though that would make a mess out of the weak axis shear connection so I have never bothered investigating the possibility further.

 

[Once20036]

While not directly applicable to your question, I've always heard that fabricators hate doubler plates, and that the cost to install a doubler plate is roughly equivalent to 500 lb of steel. So if you're seeing doubler plates required, it's probably more economical to use a heavier column instead.

 

[SteelPE]

Structural20036

This is more of a bit of research for a potential project. I would need to design some connections for a fabricator. The EOR is requiring that the connection be designed for the full moment that the connection could see (column or beam strength). Changing column sizes around would not work since the column being used are W8x67's (plus this would mess around with frame stiffness). This would required a section size change which might not be architecturally acceptable.

 

[cliff234]

??? Is this structure being designed for R>3? Is it a seismic connection? If not, then why is the engineer requiring that the connection be designed for the full flexural strength of the beam or column? The biggest moment that the connection will see is the moment determined from the analysis - not the full flexural strength of the beam or column - unless the connection is part of a R>3 seismic load resisting system. Unless the connection is part of an R>3 seismic load resisting system, there is the possibility that the engineer was taking the easy way out in requiring that the connections be designed for the full strength of the beams or columns. Your fabricator should ask the EOR for the actual moments and shears for which the connections must be designed. That one RFI might save the fabricator a lot of money - provided that the engineer is reasonable and permits the connections to be designed for the strength required by analysis. It's interesting how engineers often give little thought to doubling or even tripling the required strength of connections above what is required by analysis - yet they don't do that when designing beams. If analysis requires that a beam be a W14x22, do engineers arbitrarily increase the size of the beam to a W18x40 because it "feels better". Of course not. Then why do they not give a second thought to doing that with connections? Many engineers seem to give little thought to the cost of connections and to ramifications of specifying that connections be designed for reactions and moments far in excess of their actual required strength as determined by analysis. It is true that most structural failures are connection failures, however properly designed connections will support the applied loads with a sufficient factor of safety. If an engineer feels the need to double or triple the safety factor on connections above what is required by the building code, they should do so only after informing the owner of the project as to what they are doing, why they are doing it, and the corresponding cost implications. That's just my opinion.

 

[SteelPE]

cliff,

Actually the system was designed using R=3.5. I don't think there is much difference in this instance between designing with R=3.5 of R=3 but I am not the EOR and I have begun to send the information up the proper channels to see if something could be done (not likely with R=3.5). FYI, this is something that is done all the time and causes many problems for fabricators. On the same job I am to design the simple beam connections for 2x the reaction I would get using the uniform load tables.

When I told the fabricator about the doublers he asked if there was any way for him to know if they were needed prior to bidding the project. I said sure, I have a spreadsheet that performs the column strength calculations for me..... however I'm not sure it is wise to include the cost of the doublers in your proposal unless you have the ear of the client or are an incredible salesman as others may not be including doublers in their proposals.

Looking back at my spreadsheet I was just wondering when to use a doubler or a stiffener. It appears as the following (as I see it anyway)

Flange Local Bending :stiffener
Web Local Yielding, Web Crippling, Web Compression Buckling: stiffener or doubler
Web Panel Zone Shear: doubler

 

[Gumpmaster]

I think your above summary sounds good. Whenever possible use stiffeners instead of doublers. Not just for cost reasons, but doublers can be sketchy for ductility reasons as well (welding in the root area).

It looks like you have Ordinary Steel Moment Frames with R=3.5. Make sure you meet the detailing requirements of AISC 349 as well as the AISC 360. AISC 349 requires the connection to not just be designed for the full moment strength of the beam but 1.1*Ry*Mp, which may be significantly more (see clause 11.2a).

I'd also recommend that you check out AISC Design Guide 13 which has some examples of stiffener design.

The use of some of the other requirements you mentioned (2 times the uniform load capaicity for shear connections) indicate that the EOR may not be super familiar with current steel design practices. This may be an opportunity for you to suggest some ways to increase efficiency and save everyone some money.

 

[cliff234]

Actually there is a huge difference between R=3 and R=3.5. For R=3 you can detail the moment connections per AISC 360 to resist the moments computed per the lateral load analysis. For R=3.5 you must detail the moment connections in accordance with AISC 341 to develop the plastic capacity of the sections. "Ordinary Steel Moment Frames" are not so "ordinary". That's a common misconception on the east coast (where I live). "Oh, I don't have big seismic forces, I'll just use R=3.5 "ordinary steel moment frames". We use R=3 moment frames (and braced frames) when permitted by the code to minimize the cost of the connections.

 

[SteelPE]

cliff,

I understand the detailing difference..... but the difference in base shear would only be 16% increase (3.5/3-1). Again, I'm not sure if the structure could qualify using R=3 but if it was switched around I don't think there would be that much difference.

 

[UcfSE]

To your original question, I would do exactly what you did: talk to fabricators in your area. Stiffeners are cheaper from what I've experienced also. As to why doubler plates are allowed, I would suspect it is because they provide a good solution, even if more expensive. It just provides us options. There may be detailing specific reasons as to why one would be better than the other in a specific case.