Web doublers
Web doublers
(OP)
Recently I was asked to look at this paper on the design of moment connections. In particular, I was asked to look at example 2 (page 3) and more specifically the calculations given for that example (page 5). In the example they are connecting a 2-W24x162 beams to a W14x120 column utilizing a OMF seismic requirements. They calculate the panel zone shear using the full capacity of the 2-W24x162 and end up with a demand on the panel zone of 2266kips. Now I contend that the panel zone shear would be limited by the column capacity of the W14x120 giving a demand of roughly 1026kips vs the 2266 kips they show in their example. Which method is correct when check the design of the panel zone?
http://engineering.tufts.edu/cee/people/hines/docu...
Now, this doesn't take away from the intent of the paper, but I am just not sure if the initial calculation was done properly..... and in turn, making sure I am doing my calculations properly when faced with similar situations.
http://engineering.tufts.edu/cee/people/hines/docu...
Now, this doesn't take away from the intent of the paper, but I am just not sure if the initial calculation was done properly..... and in turn, making sure I am doing my calculations properly when faced with similar situations.






RE: Web doublers
The 2266 kips demand on the panel zone is for plastic moment design of the connection where the connection is designed for the capacity of the beam. I think they were trying to highlight that the SER was telling the connection engineer that the connection needed to be designed for the full moment capacity of the beam and not the overall capacity of the connection. That is why incredibly thick doubler plates were required. The italicized portion on sheet 4 indicates exactly what you are asking in that the connection has to be limited to what the column capacity is.
I don't think they performed the wrong calculation, rather they were trying to show what the original design was calling for based upon the SER's interpretation of AISC Chapter 11 requirements. Once the decision was made to design the connection based on the required demand (and not the total beam capacity) the doubler plate requirements were significantly reduced.
I do agree with your statement that the connection capacity is limited by the panel zone shear of the column in this situation and would design accordingly in the future.
RE: Web doublers
RE: Web doublers
It is my contention that the column limits the forces in the connection as you can not drive anymore load into the system w/o the column failing. So the column would have been the member that limited the moment that could be developed by the system. The calculation should have been prepared with the column capacity used for the calculation and not the beam capacity. The would have yielded a single plate that was 2.75" thick or two plates that were 1.5" thick. Bad, but it probably could have been fabricated.
RE: Web doublers
I am limiting the capacity of the connection to the capacity of the column. Using their equations:
W14x120 Zx = 212in^3
1.1RyMp = 1.1*1.1*212/12=1068 ft-kips/column (they even forgot the .9 here)
Vu=(1068/25 + 1068/25)*12 =1026kips (they didn't deduct for the thickness of flange here)
1026 kips is the demand on the panel zone not the capacity.
RE: Web doublers
I am not contending what you are saying but rather, I am agreeing with you. In a design setting, the connection has to be limited to the capacity of the controlling member, whether it is the column, beam, etc. However, in this case, I believe the authors are trying to illustrate the incorrectness of what the original SER was requesting.
"The calculation should have been prepared with the column capacity used for the calculation and not the beam capacity." This is the point the authors of the paper are trying to make.
This quote is from page 4: By now the engineer-of-record was realizing that the Chapter 11 seismic provisions allow for ordinary moment frames to be designed for “the maximum moment that can be developed by the system”, and the wording in Section 11.2.a. commentary:
“It is reasonable to limit the requirements to the maximum moment that can be
developed by the system, because the size of the beam or girder may have been
determined to meet demands greater than the seismic demands. Factors that may
limit the maximum moment that can be developed in the beam include the
following:
(1) The strength of the columns;
(2) The strength of the foundations to resist uplift;
(3) The limiting earthquake force determined using R = 1. ”
Previous to this, the SER was stating that the connection needed to be designed for the moment capacity of the beam and had neglected the fact that the column would be the controlling member in this case. In the authors comparison calculation, I think they are trying to show what was originally asked for not what their interpretation of the connection capacity should have been in the original design. I think this is the case because they state that in the original design, web doublers exceeding 2" were required. In the calc, it shows that the required total plate thickness is 5.2" or about 2.6" per side.
RE: Web doublers
RE: Web doublers
The way I read it, the EOR only requested the connections be in compliance with the OMF of the seismic provisions:
During the early stages of the detailing, the fabricator submitted Typical Details and
calculations for the standard connections, including the moment connections. The engineer
rejected the submittal and requested calculations for the moment connections which conformed
to the Chapter 11 AISC Seismic Provisions for Ordinary Moment Frames. The detailer’s
engineer designed the connections, and the resulting web doubler plates were
in excess of 2". In addition, the oversized thickness of continuity plate could not be
developed in welding to the web of the W14 without using a full penetration weld.
I didn't see anything where the EOR written where the full capacity of the beam be developed in the connection. If I were to be asked to design the connection shown, I would have use the column capacity to calculate the demand on the column web, not the capacity of the beam.
The authors are actually illustrating another method to limit the capacity of the connections..... in effect switching from R=3.5 to R=1 and amplifying the seismic loads accordingly. Using this method the authors arrived at a moment demand in the doubler of 660 ft-kips which is much less than their initial starting point. However, the initial starting point should have been utilizing the column capacity not the beam capacity.
RE: Web doublers
Not sure how the column failing before the connection allows for adequate overall frame ductility. So, that is not a design path that I would personally follow. It seems to me that this is exactly what we're trying to avoid.
RE: Web doublers
I'm not saying that the AISC is wrong with their analysis using R=1 (I actually didn't even know this was possible before I read the paper), just the authors of the report are wrong when they begin their calculation stating that a 5.2" thick plate is needed. Either they didn't think that the capacity of the column would control the design of the connection or they wanted to show the mistakes that are made when the EOR wants to be hard headed.
I'm not saying which path is or is not correct. If you have 2-w24x162 framing into the sides of a W14x120 and you are trying to design the panel zone then the maximum moment that can possibly be seen by the connection is the moment capacity of the column. If you go over the capacity of the column then the column will fail in bending.... and thus the frame will as well. This falls under "or the maximum moment that can be developed by the system" approach absent the EOR modifying his framing reactions and increasing his seismic loads by a factor of 3.5. You can't get anymore moment capacity out of the connection, in regards to designing the web doublers, than the moment capacity of the column.
Had the authors taken the approach with the fuse being the column instead of the beam then the doublers would have only needed to be 2" thick overall which would equal 2-1" plates either side of the column. This certainly could have been fabricated and wouldn't have made such a great case study.
RE: Web doublers
To me, this is a simple moment connection and the use of "seismic provisions" in the paper just confuses the matter. "Seismic provisions" without provisions for strong column/weak beam or ductile failure mode is not designing to seismic provisions. It's an everyday moment frame and nothing more. In that case, I agree that the moment capacity would be limited by the moment capacity of the column. I just think this should be framed as a moment frame in an everyday application and not confuse the matter with the addition of the "seismic provisions".
RE: Web doublers
Lion06 - I had an old mentor who would get very upset when other engineers designed their connections for wind forces only and ignored seismic detailing. Yes, the wind moments might be somewhat larger than the seismic moments. But, the seismic moments were factored down by the R value assuming a certain level of ductility. And, that level of ductility is only assumed to be achievable if you are following the seismic provisions. Therefore, the seismic detailing provisions will still often controls the design of the connections.
Thankfully, I
RE: Web doublers
That requires a lot of detailing and additional connection checks for a minimal benefit with R and can actually end up requiring greater connection capacity than if you use the lower R value.
RE: Web doublers
When I design buildings I do the exact same thing you do. I use R=3 and opt out of the seismic provisions. I don't see any benefit in selecting R=3.5, pay the 17% penalty up front and be done with it. However, I do a bunch of work in MA and they required you to amplify the loads in the connections if you select R=3. I imagine this has something to do with the requirements in the back of the seismic manual.
Still, the frames typically sized for drift and stiffness and the moments in the connections are no where near what they need to be when the EOR says to design the connections for the full moment of the members being connected.