Steel Column Welded Directly to Beam Bottom Flange 2

[GaStruct]

I have a residential project where the contractor and steel fabricator ignored my details requiring a column cap plate bolted to the bottom flange of the beam - the standard detail you always see. Instead, they field welded the column directly to the beam's bottom flange.

So, I have two questions:

1. Assuming the field welds are good, what issues do you see with this connection?

2. Assuming the field welds are no good, what repair options do you see? Just re-welding it is not ideal with all the adjacent wood framing already installed.

Photo is attached.

 

[DrZoidberWoop]

I can't exactly tell, but those look like some REALLY low-quality welds.... From a practical standpoint, it will probably satisfy serviceability requirements, but from an engineering standpoint, there is a lot to pick apart.

1) By welding directly to the beam above, the load path was changed. The new limit states should be checked w/ the shorter bearing value. I highly doubt the connection was prepared "fit to bear" and the welds will take compression load (perhaps a little tension in a severe situation). The base metal for the beam and post need to be evaluated.
2) If you want to avoid a field weld fix, maybe use a reinforcing plate assembly or small kicker w/ hollo bolts to HSS and std bolts to beam above (design each for half the load)? Field drilling might be adversely affected by subsequent construction tho.

 

[GaStruct]

The welds are indeed a mess and have been called out.

Can you elaborate on the change in load path? A cap plate larger than the column really just facilitates the installation of the bolts into the beam flange. The web of the beam then bears on the center of the plate and the plate distributes the load into all four walls of the column. Given how it was actually installed, how do you see the load path differing from this?

 

[human909]

It won't differ in a meaning full way. I don't see the big issue apart from seemingly poor welds.

 

[JoshPlumSE]

It won't differ in a meaning full way. I don't see the big issue apart from seemingly poor welds.

I tend to agree. But, who did those freaking welds?!

I haven't been in the field all that much, but those welds look awful. I feel like that's the type of weld someone does when they're first learning how to use the equipment.

 

[GaStruct]

That actually is one of the better welds done on this job believe it or not! It's a total mess out there. Contractor is getting sued. I didn't want to distract away from my main concern of direct welding a column to the beam bottom flange, which they did at 20+ locations.

 

[EngDM]

The beam has less of a bearing area now as well which leads to different failure modes than with cap plates. Maybe its just the photo, but the rightmost portion of the weld looks cracked and/or lifting from the fusion face? Maybe it's just poor shadows, but the quality of this weld could be put in a textbook of what not to do.

It's giving off the same vibes as those short videos where the shop owner asks the new welders what they think they should make $/hr, and some of them are just shitty.

 

[dauwerda]

Overhead weld in residential - That looks better than most I have seen.

 

[XR250]

I usually just have them weld directly to the beam. The loads are never high enough to matter and it saves them fab time and cost.
A lot of quys around here are fabbing on site with a torch and circular saw so I try to make it easy on them.

 

[Tomfh]

Yes the welds are a bit rough, but no worse than a lot of non critical welds you see about the place.

If you’re concerned and don’t want to reweld you could bolt on some brackets.

 

[gte447f]

This is what you get with steel framing in the residential market. These guys have no idea what they are doing.

There is only one reason for this... XR250 nailed it... "it saves them fab time and cost." However, I differ from XR250 in that, knowing this, I choose not to make it easier on them, but harder. I hate trying to work with the lowest common denominator, which is what these guys are. Owners need to be forced to realize that when you step up to a house requiring steel framing, you are stepping up to a different level of building, and the skill, quality, and especially cost, of everything has to go up, not stay the same.

GA Struct, who is your client? Owner, architect, or contractor? If owner or architect, I would not assist with this fix. Require the contractor to submit proof that the nonconforming construction is adequate. If your client is the contractor, well, I am sorry.

 

[Once20036]

To take a different perspective - i always consider these to be pinned joints (with a bolted plate) via the argument that the plate will bend and allow the beam to rotate relative to the column.
If you now have a fully welded connection, do you now have a fixed connection that will transfer moment to the column? It's a small column, is the bending strength sufficient?

 

[XR250]

However, I differ from XR250 in that, knowing this, I choose not to make it easier on them, but harder. I hate trying to work with the lowest common denominator, which is what these guys are.

@GTE,

Our best steel guys do it this way. (their welds look better though).
Why add unnecessary cost to the job?

 

[david339933]

Welds...

 

[gte447f]

Our best steel guys do it this way. (their welds look better though).
Why add unnecessary cost to the job?

XR250, I was originally trained designing large commercial projects in various market sectors. In general (there are always exceptions), we were taught that bolting is preferable to welding whenever possible because it is easier, faster, and cheaper to erect in the field. There were caveats of course, but we were generally taught to especially avoid field welding whenever possible. If field welding would be necessary, which of course it is sometimes, try to limit it to dead simple, 5/16 or smaller, single pass fillets.

Those early lessons, whether they were right or wrong, have stuck with me, and I think they are still good guidance most of the time, especially when dealing with low end markets like residential. Your mileage may vary, because you may have better skilled tradesmen working in the residential market where you are than I do where I am. Around here, they generally suck.

 

[GaStruct]

GA Struct, who is your client? Owner, architect, or contractor? If owner or architect, I would not assist with this fix. Require the contractor to submit proof that the nonconforming construction is adequate. If your client is the contractor, well, I am sorry.

My client was the arch for the design phase and now the owner for CA. The goal is to get them able to move into a safe house in a reasonable time period. I agree with you about field welding. I almost never allow field welding, especially in residential. There's a whole slew of requirements you're supposed to follow in Chapter N of AISC that are never done in the field. Definitely not done if fabricating on site.

If you now have a fully welded connection, do you now have a fixed connection that will transfer moment to the column? It's a small column, is the bending strength sufficient?

I ran two models, one with pinned connections and one with fixed. Columns are all OK in both scenarios. The beams are much stiffer than the 4x4 columns.

The beam has less of a bearing area now as well which leads to different failure modes than with cap plates.

Are there different failure modes? Or are they same? Just exacerbated without a cap plate? I see yielding of the bottom flange from bending (instead of the cap plate), shear lag in the column (still exists with a cap plate), and yielding/crippling of the column walls (also still exists with a cap plate). Am I missing anything?

 

[XR250]

For large commercial jobs that makes sense. With the introduction of metal cutting circular saws, partial fabrication on site can be easier for small jobs.

 

[phamENG]

If you now have a fully welded connection, do you now have a fixed connection that will transfer moment to the column?

I disagree with this line of thought. Even in the bolted connection, you had a fully welded connection. It was just between the plate and the column. But your bottom flange was effectively doubled by having the plate. The bolts will provide some amount of composite action. It's probably not definable, but it's there, and would in reality stiffen the connection a bit. So I'd argue that the plate with bolts is likely MORE of a moment connection than this is. But, given the proportions, not much will transfer through it regardless as GaStruct mentioned about the model results.

I don't treat steel on residential any different than I do on commercial. Just like I don't treat anything else differently. The idea that it's acceptable to have a lower quality of workmanship because "it's just a house" blows my mind. Especially lateral loading. On the east coast, we design for hurricanes. Where are you more likely to find somebody who didn't evacuate for a hurricane? At a store in the strip mall, or sitting on their couch? Around here, steel = special inspections. So unless you want to pay a testing lab to observe and test your field welds, it usually pays off to have it done in a certified shop and do as much bolting on site as possible.