Punch. Shear at Perimeter Col with Spandrel Beam

[DEL2000]

I am relatively inexperienced with concrete slab design, and had a few punching shear issues to get worked out that are outside of the cookie cutter examples found in the ACI. Note that this is my first Slab design and I am working under a supervising engineer who has a lot of experience in concrete design, but who doesn't always explain things where I can understand.

Anyways, I have a 12" slab framing into 16"x20" perimeter concrete columns and there is a concrete spandrel beam that is 22" deep and 32" wide (6" each side of concrete column). If I were to check the column for punching shear with moment transfer, the critical section is not a standard ACI shape since it has two different d's (ie. 10" on one edge and 19" on other two edges), and this calculation seems very difficult on this shape. My supervisor had some ideas for approximating what the punching shear capacity, looking at a 12" slab shape and a constant thickness 22" thick slab shape, and interpolating between the two.

However, I am not sure even punching shear is an applicable check in this instance. Section 11.12.6 of the "Notes on the ACI 318-99" (pg 18-11) states that moment transfer isn't critical when beams frame into the columns, as those beams provide alternate ways for the moment to get transfered into the column. With this, I am thinking that the spandrel beams provide torsional restraint of the concrete slab, and all of the moment transfer is spread uniformly across the entire bay. With this method of transfering the load, I would think that you could simply design the shear reinforcement in the spandrel beam for combined shear and torsion. Am I looking at this correct? Are there referenced standards that I should read to help me with punching shear with moment transfer problems?

Thanks for any guidance provided. Note that the example described above is the simple case that I have to work out, and there are other cases that I need to review which are more difficult to explain here, but any guidance provided in regards to torsion/shear in the spandrel beam will help me out with those issues.

 

[JAE]

Generally if you have edge beams (spandrel beams) framing into a column, you don't have a punching shear problem with the adjoining slab. The vertical shear forces are sucked into the stiffer elements (the beams) and they should be checked for full shear load.

 

[MALKOBANI]

Loads will transfer from the slab to the beam, and then to the column only if the beam is stiff enough, minimum 2 times the stiffness of the slab. Otherwise, some of the load will go directly to the column. Think of the beam as same thickness as the slab, the slab is only a flat slab right?. I ignor punching shear if the beam has twice the stiffness of the slab. Otherwise, I check punching shear considering the column could punch through the slab and the beam.

 

[dik]

Just to muddy things a tad... you might also want to check the spandrel beams for torsion... Often a problem with spandrels.

You should talk to your supervising engineer and explain that the stiffness of the edge beams is such that punching shear is not a consideration but that there is torsion in the beams... approximately 1/2 the moment going into the columns. You can ask him how he proportions the torsion... this affects the design of columns also. Some engineers use 1/2 the column stiffness, permitting it to crack as well as the compression load has an effect on Ec that is normally not considered. Ec is not linear.

You might consider the system as elastic using a live load that is realistic to determine torsion and moments into columns. There's a whole pile of things to consider and since this is one of your first projects, it's best to have some guidance at the start...

Dik

 

[ron9876]

The beams should be designed for shear and redistributed torsion. Even if the slab cracked at the face of the column (theoretically speaking) there couldn't be a shear failure because the beams have been designed to resist the total shear. Technically you should check the slab for beam shear but this likely won't govern.

 

[DEL2000]

Thanks for the responses, and I agree with all of them. Just a quick follow up question, when you determine how much stiffer the beam is than the concrete slab, do you compare the beam stiffness (moment of inertia) with a similar width of slab, or would you compare the beam stiffness with the stiffness of a column strip spanning perpendicular to the spandrel beam.

 

[ron9876]

What method are you using to calculate moments--if you are using a program based on the equivalent frame method you should model the system as it is so you will know what the slab moments are. There won't be a punching shear consideration for the column because you have spandrel beams. You should consider designing the beams for redistributed torsion. It sounds like you are making this more complicated than it is. Whatever method you are using (Direct Design, Equivalent Frame or Finite Elements) will address spandrel beam stiffness. Remember you are designing in each direction seperately.

 

[swivel63]

check ACI 13.5 or 13.6. if Vu < .75*phi*Vc, then you can transfer all of that unbalanced moment through flexure. and yes, take into account the beam you have. normally, punching shear is never a problem is you have an edge beam.