Steel Beam to Concrete Wall Connection

[etsbahamas]

I have searched but no real answer. We traditionally use 3/4" embedded plates with 1"x7" studs embedded into the 8" concrete to attach steel beams to the concrete walls. The problem is the connection from the steel beam to the plate. The concrete walls were moved in by 2 inches so we want to weld the two 4x4x3/8 angles to the beam web and fully weld it to the plate. I have heard some concerns about moment transfer and pull out of the embedded plate studs. I have seen this fully welded connection on several projects. Can someone advise if there is a reference and is this a major concern. The beam is a W30x90 and it will extend no more than 1/4" from the face of concrete wall and the top flange will not be welded.

 

[hokie66]

Both csd72 and I suggested a single plate shear connection, welded to the embed and bolted to the beam. No one commented. Is that because you guys don't like that type connection?

 

[connectegr]

Single plate is not flexible. The embed must account for the shear and moment.

http://www.FerrellEngineering.com

 

[hokie66]

I suppose this is primarily related to earthquake design. We do most of our connections this way in Australia. Never seen a double angle connection here.

 

[BAretired]

connectengr,

Okay, I was wondering about that. The part I still don't get is why AISC wants a return of two times weld size on the top. Zero return would be my preference. Is there something wrong with that?

hokie,

I have no particular objection to a single plate shear connection welded to the embedded plate and bolted to the beam web, provided the wall can sustain the moment of R*e where R is the reaction and e is the eccentricity.

BA

 

[hokie66]

BA,

Yes, the wall has to take that moment locally, but usually the studs (or drilled in anchors) are controlling. Our code actually classifies this type connection as flexible.

 

[Lion06]

BA-

I remember one of my professors from school talking about the return helping to keep the stress concentration away from the weld that we're actually relying on for strength. I can't say I've actually looked into that, but I do remember that from class.

 

[amecENG]

BA -

The Canadian code also calls for this same return. See CSA S16-01 page 3-65 "Welded Double Angle Beam Connections".

 

[Lion06]

I just checked out the commentary in AISC and it says that the static strength of the connection is not affected by the presence or absence of the return and that it's optional(it doesn't appear optional per the spec), but if it is used the length is restricted to 4x the weld size (as noted in an earlier post).

 

[etsbahamas]

Thank you everyone...I'm planning on requiring a bolted connection for the W30, for all the other beams I will allow welded but with no more than 2x the weld size return as per AISC. I think the return will help with the stress at the ed of the weld. Any comments on whether the shear plate is better than the double angle in the future.

 

[Lion06]

This is just a personal preference, but for your condition I prefer the double angle because I believe the shear tab will want to dump a moment (equal to the beam reaction times the distance from the face of wall to first line of bolts) into the embed plate.

 

[etsbahamas]

StructuralEIT....I agree

 

[connectegr]

You can use double angles with a bolted/welded knife angle connections. We frequently using knife angles to embeds, with long slots in the back-to-back legs (requires cover plates) or in the beam web (no cover plate). Due to the increased eccentricity of the long slots, the bolted connection must consider this eccentricity. The angles can be erected one at a time, or the bottom flange of the beam can be coped and the web slid between the two angles field welded to the embed. I do not recommend shop welding the angles to the embed. The embed plates are rarely located precisely.

http://www.FerrellEngineering.com

 

[OHIOMatt]

I have used bolted connections for this type application in the past with limited success. The plate embedded into the concrete often is not flush with the wall, and the walls are typically not cast to tight enough tolerances to allow for easy fit up.

Since you the shear plates or angles will need to be field welded to the plate, field welding to the beam web is not that big a deal.

My preference is angles with a return weld to the embedded and welds with a nearly full return to the beam web.

 

[larsacious]

Why not specify horizontal, short slotted holes if you are concerned about transfering end moments to the embed plate? You still need to design for the eccentricity from the centerline of the bolts to the face of the wall but at least there will not be a mechanism for restraining the ends of the beam.