Moment Connection to Beam Web 2

[jreit]

I have a condition where I need to add a cantilever connection to an existing girder.
I'm unable to find a lot of literature on the type of moment connection I will need and the best way to analyze it.
I am using a welded end plate connection member which will bolted to the existing girder.
Existing steel is historic so using Fy = 30 ksi but the new beam will be 50 ksi. Also rules out a welded connection as an option unless we do some testing and analysis.
The existing steel will be retrofitted with cover plates to ensure it can handle the moment demand (~1600 kip-ft).
The main issue I see is the connection will have a lot more bolts in the web than the flanges and I wanted to see if I could include those web bolts to resist the moment demand.
Typically the flanges are assumed to carry the moment but it would be difficult to get enough bolts in the flanges to resist that demand.

 

[DaveAtkins]

First question - can the girder take the torsion?

DaveAtkins

 

[EngDM]

Are you coping the top and bottom flanges? It appears based on your sketch that you would have to. You'll need to check the beam web for local buckling.

 

[XR250]

I'd be more concerned with the torsion as DA stated.
Also, it might be easier to visualize things if you used a CAD system other than PowerPoint

 

[BridgeSmith]

Beyond the obvious question of how torsional restraint for the existing girder is provided, I'm confused by the web splice. I don't see how it's connected to the new cantilever girder. I'd be hesitant to count on the web connection for more than vertical shear resistance, anyway.

I'd be fabricating the new beam so that the end the flanges butt up against the flanges of the existing girder, an extending the web of the new beam to the web of the existing girder. (The new beam will have to be a plate girder, right?) Then bolt angles to connect the webs of the girder and beam, and plates for the flange splices.

However, if the moment on the connection beam is substantial enough to require the splices you're showing, the existing girder is going to need substantial torsional restraint, which begs the question, why not attach the connection beam directly to whatever is providing the torsional restraint to the girder?

 

[BAretired]

Sounds like the cantilever is a bad idea.

 

[AskTooMuch]

I've beem wondering about I torsional capacity. If you will just based on "books", it seems it has little torsional capacity. But I've seen several existing gangway cantilever in our plant supported by I beams, the end of beams are just clip angle connections to columns. And it's doing "fine".

 

[driftLimiter]

Its doing fine is far from "I ensure that this design meets the requirements of the building code"

A 12ft cantilever off the side of a wide flange beam sounds like madness to me.
If you were able to get some kickers or another beam on the opposite side of the cantilever to resolve the moment instead of torsion on the WF then you'd be starting on a more reasonable path imho.

 

[BridgeSmith]

I've seen several existing gangway cantilever in our plant supported by I beams, the end of beams are just clip angle connections to columns. And it's doing "fine".

For small loads and small moments, it doesn't take much, but the OP stated "moment demand (~1600 kip-ft)". Granted, that appears to be for the existing girder, so possibly that's the vertical bending moment and not the torsional moment, but a small torsional moment wouldn't require the massive splice in the OP's sketch.

 

[JLNJ]

Making the moment connection is the easy part. Focus on the remainder of the load path. You might not have one.

Even IF (and it's a huge, unlikely if) the existing 72" girder can "take" (not really a structural engineering word) the torsion, the torsion needs to be resolved at the ends. The horizontal reaction would be something like 150 kips.

Sounds like you need non-internet-based engineering help.

 

[Tomfh]

Do not assume that web bolts carry moment. Due to their proximity to the neutral axis, web bolts take too long to engage.

As a side note - A similar issue arises with web bolts carrying shear. Everyone “assumes” the web bolts carry the shear, however in practice, if there are flange plates, the shear is transferred through the flanges and flange bolts, with the web bolts rarely engaging until the flanges and bolt have deformed significantly.


Most importantly - how is the torsion resolved?

 

[TLHS]

Beyond torsional strength issues, torsional deflection is huge on wide flanges. Pure cantilevers off of wide flanges tend to have completely unacceptable tip deflection even before there is a stress issue.

If people are seeing platforms cantilevered off of wide flanges I really hope there are back spans or they are very specific limited purpose things that maybe take one guy. I can tell you that this is really not typical in the many industrial facilities I have worked in.

If you've evaluated this by computer be very very careful. There are still a lot of programs that don't evaluate all the torsional related stuff. Like, I think STAAD still just ignores torsion forces completely in some design codes.

 

[canwesteng]

I've seen light cantilevers twisting the support beam with only cable tray and piping on them noticeably twist, and nearly drop the load. The only way a clip angle is holding up a cantilever off a beam web is if there is simply no load on it.

 

[hokie66]

Well, the girder is 72" deep. We don't know much about it. And 'simply supported' can mean different things to different people.

 

[AskTooMuch]

Canwesteng, I work inside a plant and I don't believe you. Maintenance do this all the time.

 

[bones206]

The fact that the word “torsion” wasn’t mentioned at all in the OP is concerning.

 

[BAretired]

The girder can't be "simply supported" with a 12' long cantilever.