New Steel Beam to Existing Concrete Column connection Design

[ranmoo]

Please refer to the attachment. It shows a new steel beam being connected to an existing concrete column; however the connection does not rely on post-installed anchors into the concrete column, but rather a steel plate on either side of the concrete column and clamped to the concrete column via high strength bolts holding the steel beam in place.

I am requesting assistance in how to design such a connection. It will have a downward (gravity dead and live load) downward. I have seen similar connection post applied to precast concrete structures for jib crane and bridge crane supports, but have never designed one myself, or even know if it has a common name with which to search for a design / analysis methodology.

Thank you!!

 

[msquared48]

I've seen this for temporary connections, and it does work, but for a permanent connection? I'd be very leary of that.

Mike McCann
MMC Engineering

 

[ThePenguin]

The only thing holding up the beams is the clamping pressure? I have seen a similar connection where the rods were drilled right through the column, in which case you would conceivably have to shear the rods to drop the beam. I have also seen the same thing done with wedge anchors in blind holes. For a long term connection I would be concerned with creep in the concrete relaxing the clamping load. I suppose the length of the rods would determine to what extent this may be an issue. What about seismic considerations?

 

[connectegr]

I am not aware of any design spec that would allow this detail.

http://www.FerrellEngineering.com

 

[hokie66]

A big "no" vote from me. There is no reliable shear transfer mechanism.

 

[csd72]

I really do not think this is a good idea for permanent connections.

 

[ishvaaag]

There being a concrete column already in place, maybe better use a steel corbel, even one from where to hang or a pair from which to hang your beam.

 

[alihoop]

Run the same threaded rods through the concrete column instead at say 6" gage at center spacings min edge cover 4" preferably 6". If you hit a rebar, do not drill through just patch and relocate. Much safer. If temporary then just remove and patch and refinish surface.

You can also use Hilti HIT 150 and dowel into concrete on one side only.

Your detail is not worth the risk. You are relying on someone in the field to determine how much clamping to do on those plates. There is no built in FS against human errors or slippage.

 

[271828]

Playing devil's advocate. I wouldn't use it either, but I have a question.

What, specifically in a code would preclude this connection? (Might be something, but I can't think of it.)

 

[connectegr]

Off the top of my head...

AISC/RCSC all fillers for structural bolts must be steel. The slip-coefficient of faying surfaces must be the same.

http://www.FerrellEngineering.com

 

[271828]

Thanks.

I'm trying to figure out why this is different from shear friction. For example, in Design Guide 1, shear friction is given as one method for resisting column base shear. Any ideas?

 

[connectegr]

I agree that there are similarities, but I would never recommend this configuration.

http://www.FerrellEngineering.com

 

[hokie66]

Shear friction has ties passing through both faying surfaces. A clamped connection as shown by the OP does not. At any rate, I avoid using shear friction, as I think it is only a desperation fallback when nothing else works.

 

[msquared48]

First vibration and this assembly will be on the ground.

Mike McCann
MMC Engineering

 

[ranmoo]

Thanks to all that have responded. Couple of responses to the comments:
• I too thought perhaps an analogous design comparison could be a shear friction design for column bases.
• Seismic loading is minimal. Wind loading would control, but minimal at this location as well (compared to dead and gravity, which would be consistent for this application). Vibration concerns minimal, but possible.
• Concern about creep in the high strength bolts resulting in lessening clamping tension is a huge concern. If designed for more than a temporary time frame, I had considered tensioning the bolts to impart bearing pressure between the plates to the concrete surface to then calculate the normal friction force to overcome this bearing force ….. and have this resultant friction force be many many times higher than the calculated loading.
• I have most often noted “permanent” connections similar to this where jib cranes have been connected to precast columns (industrial plant settings). In these applications, the jib crane has a relatively light lifting capacity (couple of tons), but here again, as had been mentioned above, dynamic and impact loading would be a concern.

Thanks again, and welcome any more thoughts.

 

[VTPE]

I would argue that friction connection does not provide a positive load path as the friction force is dependant on the clamping force that may relax over time. Both IBC and AISC require positive load paths.