Welded-Welded Single Plate Shear Connections 1

[DaveAtkins]

I am working on a project where some of the floor beam connections have been installed incorrectly--single plate shear tab welded to BOTH the supporting girder and the supporting beam.

Of course, the contractor is asking me if this is acceptable--my first response was no, since there has been no testing on this type of connection (as far as I know), and it doesn't have enough ductility to be considered a "pinned" connection. I have been asked to look at the problem one more time.

Any ideas?

DaveAtkins

 

[MotorCity]

Due to the lack of ductility and rotation capability of the connection, I would probably consider the beam ends to now be fixed (this may also require some stiffening or reinforcement of the supporting beam web). If the shear tab, welds, etc are ok considering fixed supports, I'd call it good. Essentially they have turned the pinned connection into a moment connection (again, assuming that the supporting beam can also resist this moment.)

 

[TehMightyEngineer]

I've run into this before. If I recall the conclusion I came to was it results in a poor moment connection; probably more in line with a partial moment connection. I enveloped my design as both pinned end or fixed end. For the connection I verified it could transfer the full fixed moment, definitely a tedious check. Don't forget the torsional load this will put into the supporting member; for me in was a column, as you're going into a girder the torsional strength you have is probably much more critical. For mine the loads were light and the connection has plenty more strength than I required so I let it pass; if you're close to full capacity then I'd probably just say "no, fix it" and be done. The fix isn't impossible; field drill bolt holes and install bolts, then grind off the weld.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[DaveAtkins]

MotorCity,

Thanks for the suggestion.

TehMightyEngineer,

Thanks also for your suggestions--in fact, they already installed bolts, but have not ground off the welds.

I went to a seminar a LONG time ago, where the presenter explained a single plate shear connection cannot transfer torsion into the supporting girder, because the end of the supported beam can only rotate a certain amount, resulting in a very small rotation of the supporting girder, and therefore little to no torsion. I think I will look into this.

DaveAtkins

 

[TehMightyEngineer]

I went to a seminar a LONG time ago, where the presenter explained a single plate shear connection cannot transfer torsion into the supporting girder, because the end of the supported beam can only rotate a certain amount, resulting in a very small rotation of the supporting girder, and therefore little to no torsion. I think I will look into this.

Interesting, that would make some sense. Perhaps that's where I got partial moment connection in my head regarding my previous welded-welded single plate cnxn.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[canwesteng]

Agree with the presenter - go through the exercise of how much rotation would occur in the girder and compare to the end rotation of the beam if it's pinned and I suspect you'll find it behaves as a pinned connection. However, if the connections is on both ends of the girder you now have moment continuity through the beams.

 

[nutte]

There used not to be any research on this connection... But now there is. What I take from the paper is that despite the objections listed above, the welded/welded shear tab does provide adequate ductility after all.

Behaviour of Single and Double Row Bolted Shear Tab Connections and Weld Retrofits

http://digitool.library.mcgill.ca/webclient/StreamGate?folder_id=0&dvs=1475776929859~209

 

[cliff234]

If the web of the girder is fairly thin, you could probably justify the connection by assuming that deformation of the girder web would allow the slight amount of rotational ductility required. I would also oversize the weld to the girder web so that the weld is not the weak link (due to the increased rotational restraint of the connection). For a 3/8" plate, I would use a 5/16" fillet weld, both sides.

 

[TehMightyEngineer]

Nutte, fantastic reference; that's exceedingly good information and actually somewhat surprising.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[sandman21]

Welding the web to the shear plate does not create a fixed connection. AISC design manual table 10-3 has an all welded connection in the simple shear connection section. After all you can make the two connection as stiff as you want the shear plate will still yield between the beam and girder.

 

[DaveAtkins]

I don't think the all-welded double-angle connection is analogous to an all-welded single plate connection, because the double angles can flex, which creates the pinned condition.

DaveAtkins

 

[TehMightyEngineer]

Agreed with Dave; while the reference nutte shows that single plates with bolt holes do appear to have flexibility enough to be pinned connections simply assuming they will act identical to a welded-welded double or single angle connection seems inappropriate given that only the toes of the angles are welded to the supporting member allowing flexiblity of the connection.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[sandman21]

And you have a single shear plate similar to Table 10-9? The double angle has twice the area of attachment to the girder. Plus shear yielding of the plate is ductile and the desirable failure mode, which you have not precluded.