Steel beam to Concrete Column

[deemtocomply]

Hi gents,

I have a UB steel beam connected to a reinforced concrete (RC) column, either positioned on top of the column or attached to its side. Based on my analysis, which assumes pin connections at the ends, I am obtaining the following design actions.

However, I believe both connection configurations, as shown below, will induce moments at the connection. When I model it as a rigid connection in my frame analysis, with the RC column having dimensions of 600x600, I am getting a large moment (about 80 kNm). I find it challenging to make the connection work with such high moments.

My questions are:

1. Is modeling the steel beam–RC column connection as rigid in the frame analysis the correct approach to determine the end moment?
2. Are there alternative connection methods you recommend that could make this design feasible?

Cheers!

 

[W460x68]

Most engineers I know (myself included) would not consider any moments in the case of the beam sitting on top of the column. I don't expect the connection to be stiff enough to induce any considerable moments.

For the beam into the side of the column, the offset from the beams end reaction is likely to induce a moment in the column. That being said, I see this regularly go unchecked and I wouldn't normally check this unless the end reaction is high. In my books 130 kN is not large enough to warrant considering this moment.

To answer your questions:
1. I don't think modelling this as a rigid connection is the correct approach.
2. In my opinion your first option is the best option. I don't see any alternatives being an improvement over this.

 

[phamENG]

I think you're right to consider that moment. If the connection configuration can attract the moment, it will. So it becomes of question of how your connection handles that deformation and whether or not it fails. A simple shear connection transfers some moment at service level, but as loads approach the connection's ultimate capacity it fails in a way that allows rotation (bolt plowing) but doesn't fail in shear. Here in the US, most 'standard' steel detailing has this consideration baked in, so we don't often think about it. When you get into some of the more unique configurations you have to be cognizant of all of the little things that allow us to make the assumptions we do in typical connection design.

So, your connection will need to either be stiff enough to reliably transfer that moment or be capable of deforming in a way that allows the connection to rotate at the design load without a collapse.

 

[deemtocomply]

Most engineers I know (myself included) would not consider any moments in the case of the beam sitting on top of the column. I don't expect the connection to be stiff enough to induce any considerable moments.

For the beam into the side of the column, the offset from the beams end reaction is likely to induce a moment in the column. That being said, I see this regularly go unchecked and I wouldn't normally check this unless the end reaction is high. In my books 130 kN is not large enough to warrant considering this moment.

To answer your questions:
1. I don't think modelling this as a rigid connection is the correct approach.
2. In my opinion your first option is the best option. I don't see any alternatives being an improvement over this.

Thank you very much for your reply.
I was actually thinking the other way around but I am not sure. The first one is more likely to induce moment as the anchor at far end restraint the beam from rotating, and thus creating a rigid connection (until the anchor being pull out by a bit (not sure) then the moment will be released by a fair amount). The second one, on the other hand, since the anchor bolts are bit away from the beam flanges, allowing the beam to rotate and relase the moment. so I guess we are on the same page on thi. (I just don't like the idea of using anchors to transfer shear in gravity like this)

 

[deemtocomply]

I think you're right to consider that moment. If the connection configuration can attract the moment, it will. So it becomes of question of how your connection handles that deformation and whether or not it fails. A simple shear connection transfers some moment at service level, but as loads approach the connection's ultimate capacity it fails in a way that allows rotation (bolt plowing) but doesn't fail in shear. Here in the US, most 'standard' steel detailing has this consideration baked in, so we don't often think about it. When you get into some of the more unique configurations you have to be cognizant of all of the little things that allow us to make the assumptions we do in typical connection design.

So, your connection will need to either be stiff enough to reliably transfer that moment or be capable of deforming in a way that allows the connection to rotate at the design load without a collapse.

Thank you.

I was considering the same for the first connection. The moment would likely fall between a pinned and fully rigid connection, with the worst-case scenario being similar to the one I described earlier, where the anchor pulls out. I suspect that as the anchor begins to deform and before it completely pulls out, the moment would significantly reduce. However, I'm not entirely confident about this and would prefer to avoid using FEA. I'm also unsure if there's any dedicated text box for this.

Regarding the shear, in the worst-case scenario where the far-end anchor fails and pulls out, embedding a shear key into the column might be a better solution for the first option. That said, I might be overanalyzing this.

 

[JAE]

For the side connection type - we've simply embedded a steel plate in the face of the column using headed studs or deformed bar anchors.
Then we use a single plate shear connection, welded perpendicular to the embedded plate with simple shear bolts (bearing). There might be some secondary moment transferred into the embed plate but for steel to steel connections usuing the single shear plate we've never assumed any fixity.

For the "on-top" connection, we've used embedded anchor bolts and set the beam on an elastomeric pad which allows some rotation without inducing much moment.