Welded fin connection

[JDCengr]

I have a simple shear fin connection where the fin is securely embedded in a concrete column. I am fastening a W21 beam (with a modest load) to the fin with a single row of bolts, which it easily handles. Contractor got something wrong and can't align bolt holes. He now wants to have a welded fin-to-beam web connection. I don't like this much. What are my biggest concerns?
Thanks

 

[dik]

Can you weld the holes up and redrill them where they should be? I usually use the moment capacity created with the bolts to minimise the eccentricity of the shear load to the column.

Dik

 

[Ron]

Field welding is not an issue if you use the right process and can have enough weld area to provide the proper strength. SMAW is the correct welding process to use for such procedures. If they propose GMAW or FCAW say no. Welders must be certified to AWS requirements and you should check their certifications.

 

[human909]

Why are you so insistent on a bolted connection? A good field weld should perform just as well if not better. Moment transfer might be slightly higher due to a stiffer connection, but a quick run through on those calcs probably will show that it isn't a concern. That is what I'd be checking first up and if it passes then all good.

 

[JDCengr]

My biggest concern is the very rigid fin and the web/weld having to take all of the moment and rotation in the relatively small connection. I have no way of attaching the flanges, so the web has to absorb the moment and rotation.

 

[human909]

The welded connection won't be fully rigid. Just like a bolted connection isn't fully pinned. Thus the connection won't take all the moment and all the rotation.

I'd be less worried about the strength of the connection which based on your description should excessively strong. The only issue that I foresee is the additional moment on the column. Which may or may not be a concern.

Though I haven't pulled out my calculator yet... I suggest you do so you might be surprised on the moment capacity of the welded connection. Spend 1 hour doing an FEA or similiar analysis and you may also be surprised how flexible a welded fin plate is.

(I'd happily be proven wrong. Currently I'm just guessing because I really don't have the full picture of the beam and the connection.)

 

[BridgeSmith]

I think there could be cause for concern, depending on the answers to a couple of questions:
1) Are you expecting significant rotation at the end of the beam, i.e. is it a long span where deflection relatively large?

If the answer to 1) is yes, then the second question that needs to be asked is:

2) How securely is the fin plate anchored into the column? Enough to develop the moment capacity of the fin plate without shearing the anchorage or breaking out of the concrete?

If the beam span is relatively short, you may not have enough rotation at the end of the beam to apply significant stress to the fin plate, but if you do, and the weak point is the anchorage into the column, I foresee spalled concrete in your future.

How far off are the holes? In which direction? As long as there is adequate bearing area for the bolts on the plate and the web (so the bolts aren't going to break through into an adjacent hole) and the washers are of adequate size and thickness so they don't warp because of the oversize holes (if there is concern, plate washers are an option to consider), drilling would seem to be a viable option.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[Agent666]

For a competent fabricator, filling in the holes and redrilling shouldn't be an issue like dik noted in the first reply. It's done all the time in these parts and no one bats an eyelid at it, and it maintains your original nominally pinned design intent. Coming from a country where seismic forces dominate, going from pinned to fixed creates a number of issues that I'd rather avoid. Depending on the structural system you could attract significant flexural demand for example.

 

[KootK]

What are my biggest concerns?

1) Lack of rotational ductility in the joint. For most applications, I'd not loose much sleep over this.

2) Incidental moment being transferred to the connection inducing a concrete breakout failure there.

In my opinion, however, these concerns are nearly identical even if you had the bolted connection. With bolt plowing, hole tolerance utilization, and all the rest, a bolted connection is usually pretty crappy pin in reality.

There are a couple of strategies that you might employ as ways to justify the welded option:

3) My first stop will always be calculate the anticipated slope of the beam at the connected end assuming a pin there. It's a night and day difference between and L/600 beam and an L/120 beam. Use that calculated slope to see if it might be reasonable to absorb that rotation in embed flexibility which will depend very much on your stud layout and embed plate thickness.

4) Use AISC's procedure for extended end plate connections which assumes that the knife plate is moment connected to the beam. Take the plate moment capacity that you calculate there and carry them forward in the design of your embed and column.

 

[BridgeSmith]

An excellent response, KootK. I think you're right on the money, as usual.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[JDCengr]

Thanks to all. That helps.
JDC