extended end plate moment connection, field welded to column, not bolted

[samdamon]

Two questions- First, is anyone aware of research and/or a design method for an extended or flush end plate conn that is welded to the supporting column rather than bolted? The AISC design guides 4 and 16 address only field bolted end plates. I have not found anything further on the web for a field welded end plate. The connection I want to design is not in a frame, its for a renovation, helping hold up several cantilever beams that carry a canopy attached to the exterior of existing building columns. The end plate connection will carry dead load, snow and wind. Field bolting may be very difficult due to existing conditions.

I would like to use fillet welds to attach the end plate to its supports. Before erection the cantilever could be shop welded to the end pl with cjp welds as needed. Doing some quick and approximate calcs using elastic theory and assuming the end pl is perfectly rigid, code minimum forces on the proposed welds seem small enough to do this, if I use an extended plate, but I would like to hear how someone else might have approached this condition.

This leads to my second question. Would it be best to develop the full bending strength of the cantilever with this end connection or should I use the code design forces, which are far less (deflection governs the cantilever beam size). I also am not sure if the supporting existing columns should be checked for just the code loads or instead for the forces resulting from developing the full bending capacity of the cantilever. Any feedback appreciated.

 

[ajh1]

If you are not going to field bolt the splice to the column, what is the point of the extended end plate? You would be better off simply welding the beam itself to the column face. Be sure to check the existing column for localized forces from the beam flanges. Stiffeners and potentially a doubler plate may be necessary in the column depending on relative member sizes, loads, etc.

 

[BadgerPE]

I also am not seeing the point of an end plate. Are you trying to eliminate CJP welding in the field by "...use fillet welds to attach end plate to its supports"? Maybe we are a little confused, at least I am, and posting a sketch might clear up the confusion.

 

[KootK]

I wouldn't lean too heavily on the provisions for bolted extended end plate connections. They are predicated upon the ability of the plate to undergo significant plastic load redistribution. Welds likely aren't compatible with that. You could design the weld assuming that there will be weld only on the top and bottom of the plate. That reduces your problem to just two plate cantilevers and two line welds. I'd also multiply your load by an additional 1.25 to account for non-uniform weld stress distribution.

I see no need to design anything for the bending capacity of the cantilever beams. Actual loads ought to do here.

I've provided end plates in theses situations in the past to make it easier to square up the pieces prior to welding. In retrospect, I'm not sure if that helped or hurt. With an end plate, making field adjustments to the beam angle is tough.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[samdamon]

Thanks for your replies. Yes I am trying to eliminate field CJP. Using an extended end plate (The plate runs several inches above and below the cantilever beam top and bottom flanges) gives me a greater "d" as in "M/d" and eliminates need to stiffen existg column flange. Kootk I thought an end plate would help erection but perhaps am totally wrong.

 

[nutte]

Bad idea. The end plate is not perfectly rigid, so it will rotate, inducing a torsional stress in the fillet weld across the top of the plate. Torsional stresses along the axis of a fillet weld are to be avoided.

 

[KootK]

Why does omitting the plate result in a field CPJ? Couldn't one just fillet weld the unplated section directly to the column?

I thought that the plate would aid erection too. Clamp it in place and weld away. My concern is that there's no way to adjust things in the field if the column isn't plumb or the plate isn't square. Maybe I'm overthinking it.

I take Nutte's point but the torsional weld stress thing doesn't bother me. The vertical welds on every welded double angle beam connection have the same problem and they seem to do okay.



I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[RFreund]

Correct me if I'm wrong but I think these are two different situations. I don't think either are ideal but maybe under small stress are ok? Maybe someone can help clarify?

Torsional stresses along the axis of a fillet weld are to be avoided

This I imagine

The vertical welds on every welded double angle beam connection have the same problem and they seem to do okay

See attached sketch. The more I think about it the more I may be confusing myself. I'm thinking of a weld unzipping as opposed to the torque on the weld. When is "unzipping" a problem?


EIT
www.HowToEngineer.com

 

[KootK]

@RF:

It's always risky to speak on another's behalf but I'll give it a go anyhow. I believe that Nutte's torsional stresses are more aptly referred to as weld bending stresses. It's this phenomenon and it's real: Link.

A version of this occurs on pretty much every base plate weld when columns have tension or OT moment. It also happens at the upper end of welded clip angles when supported beams rotate. It looks just like your upper sketch.

In a clip angle connection, I believe that the return weld is about avoiding low quality start/stop weld at a location where stress concentrations may arise.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[KootK]

I'm curious now: did others recommend penetration welds above because they are less prone to weld bending issues?

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[nutte]

KootK, that is the phenomenon I'm referring to. That is a nice article - thank you for posting it.

When AISC says they recommend not loading welds this way, I take it seriously. The stresses in a clip angle weld due to small end rotation are one thing. These are secondary to the main load-carrying elements, the vertical welds taking the shear. In this moment end plate case, the primary stress is the one causing tension on the root of the fillet weld. I would avoid it at all costs.

 

[KootK]

Do you know where the AISC recommendation is located nutte? I believe you; I'd just like to check it out myself.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[nutte]

August 2011 Modern Steel Construction, Steel Interchange section.

http://msc.aisc.org/globalassets/modern-steel/archives/2011/08/2011v08_interchange.pdf

That sketch leaves much to be desired, so I posed the question to the Steel Solution Center, asking about in particular clip angles welded on the outstanding leg. Can we take axial load through connections like this? They recommended not to, due to the torsional loading in the welds along the length of the angle outstanding legs.

 

[RFreund]

So rotation about the root of fillet weld shall not occur. Is what I seem to be taking away from this.

Thanks for the replies and thanks for the article KootK

EIT

http://www.HowToEngineer.com