Site Welded Connection - Flanges Only, No Web Connection 2

[ImposterSyndrome]

Hello

This is my first post here so my apologies if my etiquette is not correct.

Location: Canada
Beam Sizes: W8x31
Connection location: T-intersection of two steel beams. Both beams at same elevation.
Proposed Connection: Single Angle bolted connection through the web of the beam. Angle welded to web of supporting beam. Supported beam would be coped. Pretty standard connection

Problem:
Contractor didn't cope the beam, and just welded the flanges of the beams together. No stiffener plates in either member. See sketch attached.

My assumption:
Shear is dominant in the web (fact).
1. If the flanges are the only items connected, I have more of a rotation (moment) connection, than a shear connection. Heck, theoretically I don't have any shear transfer since the webs are not connected at all, but we all know that isn't true.
2. Shear forces will re-distribute into the flanges and eventually make its way to the web of the supporting beam.

Question: Does anyone have documentation available on the strength calculations for this type of connection? I have never seen it before. It seems unconventional and rather sketchy.

 

[Settingsun]

Doubtful you'd find documentation. This is an error, not something someone would recommend and therefore prepare a design method for.

It should be rectified properly. However, as a thought exercise:
- How much flange weld length is needed to transfer the shear force? The supported beam flanges bend about their weak axis for at least that length.
- Supporting beam flanges are bending about their weak axis for their full outstand length. Need to be stiffened so you're into major work territory - might as well fix it properly rather than a bandaid.
- How stiff is the supporting beam in torsion? It may be effectively pinned so not much moment at the connection but rotation of the supporting beam occurs. What problems does that cause? Or what do you have to do to provide torsion support near to the supported beam?
- If there is significant moment due to torsion stiffness of the supporting beam, you then need to look at how that interacts with the shear since they're both going entirely through the flanges.

Just reject it and get them to propose the fix.

 

[ImposterSyndrome]

@Steveh49

Ya, that is what I am leaning towards. Not worth the risk on my end.

Something like 40 connections done this way throughout the project. Not ideal. Would have saved them lots of money to just ask me ahead of time if it was ok.

Thanks!

 

[SWComposites]

Reject it. Make them add proper shear clips on the webs.

 

[dik]

I'd likely fix it with a web-web plate welded in place... first time I've seen that connection.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[XR250]

I have fixed one of these before during a renovation by installing a 16" long C6 channel in the web of the right member with the end welded to the left members web. This gives an easy way to resolve the shear.

 

[CRQUT]

Looks to me like they dont know what they are doing. A decent steel erector/fabricator would know thats not standard. It would be a red flag for other parts of the project they have worked on in my book.

 

[KootK]

1) This article contains provisions that could be used to check the beam flange with some adaptation: Link

2) I'd still have this fixed. The welds seem critical to me even if the girder flanges work. The only way that I might let it go is if it were inspected, full penetration welding, and all of the shear was assumed to be delivered roughly between the fillet toes of the supported beams. That laundry list should be more than enough to deter your contractor.

3) My fix would be the same as XR250's. I'd probably be willing to consider bolted connections all around if desired although welding may well be cheaper.

 

[ImposterSyndrome]

Thanks everyone for the help. Will certainly be looking to remediation options. I believe XR250 may have the most straight forward approach at this time.