Existing steel beam supporting new moment- lever arm assumption and any uplift?

[jsharkbait]

Hi everyone,

I'm dealing with an interesting design problem that I'm trying to resolve. We have an existing building with flat arch terracotta slabs on steel beams at 4'-0" OC. At the top floor, the architect wants to put up a new cornice over the top of the existing brick walls. We'll use reinforced CMU to support the cornice, but don't want to try and tie the rebar into the existing brick. I want to tie the CMU into the beams instead, so I have added a tube underneath the CMU and will connect that to a pair of angles at each steel beam, because the location of the cornice and end of the steel beam don't align (and the architect won't want to move the CMU inside, leaving him with less terrace space).

The overhang of the cornice and wind load will produce a moment in the CMU, put the tube into torsion and apply a moment to the angles, which I'm treating as being fixed to the steel beams (please see the attached sketch). Will this moment that is then applied to the beam have the potential to cause uplift in the beam? I am unsure if I can assume that the steel beam is stiff enough to distribute the moment that is applied at its support so that I end up with a lever arm that is the span of the beam (13') by which time the T/C forces are small and will be resisted by the dead load.

I hope this makes sense. Cheers.

 

[jayrod12]

Don't you essentially have this? but with M2=0?

If so, you just need to analyze this and check the beam for the resulting moments. keeping in mind your unbraced lengths for positive and negative bending appropriately (critical flange bracing depending on your local terminology).

 

[KootK]

You got this:

1) There is the potential for uplift.

2) In most practical situations, I'd expect the uplift to be easy to deal with.

 

[KootK]

I am unsure if I can assume that the steel beam is stiff enough to distribute the moment that is applied at its support so that I end up with a lever arm that is the span of the beam (13') by which time the T/C forces are small and will be resisted by the dead load.

1) I'd think any reasonable beam stiffness would do for that purpose and;

2) I'm not sure that you need a particular stiffness to get this done. It's statically determinate after all so basic equilibrium should be enough unless your trying to prevent beam end rotation over top of the connection to the masonry.

 

[r13]

Another load path. Note the tension T tends to lift the angle up, you should ensure the roof load and beam weight can weigh it down.

 

[jsharkbait]

I get a bit confused by joining these responses together. I want to use the entire beam length to minimise the T/C forces, which is in line with jayrod and KootK said. But I also thought about the load path retired13 mentioned. And it seems to me that I would need to follow that load path before I could look at using the entire beam as my lever arm. The issue is, at that point, my Tension force is higher than the dead load I can work with. But at that point, if it tries to lift up, does the stiffness of the beam distribute the tension along the beam?

On a similar thinking point: At what point does the lever arm become so small that the T/C forces cancel each other out? While I look at retired13's markup and the potential for uplift looks obvious from the T, if I sum vertical forces in the beam are they not doing that? First principles lapse here for me.

Thanks for your responses.

 

[r13]

If Beam end reaction less than the lift force, you need something to weigh the beam down, or anchor it to wall below. Note other than the cornice, will there be wind load effect?

Also, did you include the weights above the HSS? It would negate the lift force.

 

[hoshang]

Hi
retired13's path can be changed if you used stiff tube and angle so the distribution won't be triangular. Instead it'll be uniform.