Steel Connection Design

[PG35]

In ontario & michigan, we specify that connection design is to be carried out by the steel fabricator and/or engineer retained by the steel fabricator. This is common practise here.

Is this common everywhere in the US? Why don't the design engineers always do the connection design?

 

[slickdeals]

This is interesting. As an EOR, I think you own the building and are responsible for everything. However, that is only my feeling. I don't know how the law sees it.

Say for example, you had all your connections to be provided by the fabricator and he submits signed and sealed calcs for you to review, and you skim through it and APPROVE it. God forbid a connection fails, does it absolve you? I think as a person who approved it, you are as guilty (or negligent) as the person who designed it. Like I said, I don't know how the law (or lawyer) will look at it.

I personally think simple shear connections take more time to review (and I mean thoroughly review) than to design yourself. I work in an East coast firm and our firm has a policy that we design all our connections.

 

[271828]

Slickdeals, I had no idea that anybody on the East Coast used this approach. 3 for 3 of the firms I worked for didn't do that and (I thought) had solid reasons for letting the fabricators do it.

Is your firm able to get a higher fee for doing the connections also?

Do you not ever run into a fabricator who is better set up for some other type of connection? (For example, you showed double angles all over the place, but they like shear tabs.)

How do you pick what type of shear connection? Angles versus shear tabs, for example?

 

[slickdeals]

I think our firm is probably an exception. The views in my previous post are from our "steel guru", who has done dozens of stadiums in his lifetime and is very well respected.

We have created standard tables with angles/ shear tabs for beam to girder connections and publish both in our drawings. The steel fab can choose one. Lately the choice has been single angles over shear tabs (all bolted single angles, with eccentricity accounted for in the girder).

Most of our heavier connections are double angles (girder to columns). But again, we discuss these early on in a project and come to a consensus within the office based on previous experience.

On a related note, we have always had problems with delegated engineers. Lately delegated engineers have been doing a very shoddy job of designing and detailing stairs. Most of these designs might work but are worrying. Like providing a single bolted connection, providing questionable weld sizes, providing plate stringers to which the pans are welded close the NA of the plate etc. The amount of time we spend in CA before getting it right is probably as much or even more than what it might have taken us to design it in the first place.

As for fee, I don't think we charge extra fees just because we do our connections, but I am sure we account for it.

 

[Lion06]

slickdeals-
I agree that the EOR owns the connections and if one fails it is on your plate. That is the reason for the submission of the signed and sealed calcs and shop drawings (probably not signed and sealed). You have to verify that the connection they are providing has adequate capacity per your reactions on the drawings (or a reaction table- which we use). I have sent many shop drawings back "Revise and Resubmit" for inadequate connection capacity.
Just because you are allowing them to pick the connections they want doesn't absolve you of checking the capacity. Additionally, you rarely have to check more than a couple of connections - they will most likely use one kind for all beam to girder connections, one for beam to column flange connections, and a third for beam to column web connections. Also, because most firms do this, the calcs have been reviewed by MANY engineers before it gets to you, so there have been many sets of eyes on it already. They don't redo their calcs for every job - these are, after all, standard shear connections, nothing more.
What do you do when the fabricators prefer double angle connections to the shear tabs that you designed and documented in your drawings? Do you just say, "No"? If so, I don't think that is the right approach. I have taken a pretty good amount of time to design and detail a thru-plate moment connection at a HSS tube beam to a HSS tube column and ended up getting a request for a different type of moment connection that would be easier for them to fabricate. I reviewed their calcs and allowed them to change the detail. That was a bit of a pain and a little more annoying - If that happened to every shear connection in the building.........

 

[slickdeals]

StructuralEIT,
We never reject any drawing/calculation because it deviated from what we intended. If a fabricator can substantiate it with necessary calcs and details, we approve it. After all it is in the best interest of the job.

However, I have been part of peer review teams for many stadium jobs (designed by notable firms) where the moment connections are provided in a table. That table is a standard table that has moment connections for all the W shapes available. Most of these connections are designed for the full capacity of the member (and remember these are in projects in the East coast with little or no seismic demands). No wonder the contractor wants a peer review team because the shoddy engineer just decided to save time and effort of designing the connections. It is just being lazy and irresponsible.

Again I am not debating as to why some firms design their connections and some don't. I am just trying to present the other side.

 

[PMR06]

Our firm designs all connections, shear, moment, baseplates, etc. We feel it is our job to. We've got spreadsheets and standard details that handle 99% of the connections. It's really not that hard, and some of them are actually enjoyable. And if the fabricator wants something different, he submits alternate designs AT HIS COST.

We were hired once to design connections for a fabricator. The EOR (from a large, nationwide firm) spec'd 100% shear and moment capacity for all connections (IMO, lazy & irresponsible behavior). The fabricator knew what this was going to mean, especially at the moment connections. We had full pen welds, web doubler plates, and stiffeners everywhere. The fabricator estimated tens of thousand of dollars in extra cost just by requiring 100% capacity. We argued with the EOR in meeting after meeting to change his design to the actual forces, not 100%. Finally it took a meeting that included the owner. The owner looked at us and said "Why are you even here, I thought I hired [EOR] to design the building". After we explained the cost savings to the owner, the EOR gave in.

Long story short, if it had been our job from day one, the connections would have been on the drawing, and the drawings would actually be constructible as is (imagine that!!).

 

[slickdeals]

Oh, and I have also seen a design where instead of thinking of alternative solutions, the EOR provided 3/4" by 24" deep plates welded on either side of a W24 to provide torsional capacity. This was also on a stadium which would have involved around 2000 linear feet of welding per concourse

 

[271828]

PMR06, are you in an East Coast firm?

 

[271828]

After reading all of this so far, I still like the way I used to do it (surprise, surprise, LOL--of course I'm biased!). All design forces were specified and the fabricator had to hire a connection design engineer who would provide calcs and seal their drawings.

Other engineers in the firm didn't require calcs and a seal and I do believe that is at least questionable. They would argue that the designs are in the AISC Manual and it's been done this way for decades.

The approach of providing a table of standard connections is what a connection design engineer usually does anyway. They have some standard conns in a big table and give those to the detailer and then go work on the harder ones. It seems to me that either way, a detailer is yanking standard designs from tables and some engineer is taking responsibility at some step in the process. The big difference IMO is that a connection designer does nothing but connections, so is a lot better at it. Like I typed before, most every EOR I've known has so little interest in connections that they really know only the bare minimum (what's that U factor for?!).

Obviously several different approaches that seem OK.

 

[PMR06]

271828, no midwest.

 

[Lion06]

I do see some benefit in specifying full-pen welds at all moment connections - but the stiffener/doubler plate requirements should be checked for the actual forces, not 100% capacity.
It is my understanding that if you provide something less than full moment capacity at the connection that it will be a partially restrained connection. That will allow greater rotation at smaller moments. Maybe I am overanalyzing, but that seems to me to translate into greater tip deflections, greater frame sway, higher second order effects.... am sure there are a couple others I am leaving out.
Am I wrong in my understanding?

 

[Lion06]

Also, if you are likely to get requests to change the connections, why waste your time designing and documenting them? Just leave it to the fabricator and check them when they come in - that is what you will do anyway if you get a request to change them?