Design of Beams supporting Hollow core slabs for torsion and bending 2

[Engr. Masood]

Dear Members,

I am designing a building which has hollow core slabs as seen in the attached screen shot, i am using Etabs 9.7.4 for the design. I have to design the beams for the torsional affect due to the eccentricity of the hollow core panels. Please guide me how I can abstract data from Etabs to calculate the torsion in the beam and also design the beam. Any references of manual design are also highly appreciated.

 

[MotorCity]

Can't help you with Etabs, but depending on the connection detail between the hollowcore and the beam, you may not have torsion. In fact, I'd be surprised if you have torsion if the ends of the hollowcore are simply supported and not transferring end moment.

 

[milkshakelake]

In ETABS, using one way slab will not produce torsional effects. Maybe try Assign -> Shell -> Insertion Point, though I've never tried it.

Another way to do it is to model each HC slab as a pinned beam with Assign -> Frame -> End length offset. This is janky and might cause other stability issues you have to go through one by one. The diaphragm will be perpendicular to the real slab direction in this case.

If that doesn't work, there's a very hacky and janky way to do it, and highly not recommended. Make moment connected beams coming out perpendicular to your beam with the length of your torsional moment arm (say 75mm) at each break point of the hollow core slab. Create another pinned beam parallel to your main beam which is moment connected. This will transfer the load to a "dummy" beam, which will transfer it into a real moment connection, which will transfer into your main beam as torsion. This is incredibly hard to model and check, you will lose about 75mm of floor load everywhere you do it unless you use some more modeling tricks, and the diaphragm won't work so you'll need a separate model for gravity and lateral.

If these methods don't work, I'd just calculate the torsion by hand and get other things like moment and shear from ETABS to Excel tables. You can manually add torsion to the tables and run beam calcs in Excel. This would be the absolute last step, after everything else is modeled and set in stone.

Also agree with MotorCity. Without seeing your detail, I can't really tell where the torsion would come from.

 

[mtu1972]

I had a project a long time ago with two equal spans of hollow core slabs. I designed the center beam for the total LL and DL. The contractor placed the hollow core slabs the full length on the rear span. Full bearing on the edge beam and half bearing on the middle beam. The center beam failed in torsion with only half of the DL in place. The planks had embedded plates that had not been welded to the beams yet as we had detailed.

The good news was that no one was injured as there was a mandatory safety meeting that morning.

My design was reviewed by my supervisor and found to be OK for the total anticipated loads. For the rebuild we added mid-span beams to provide some torsional stiffness, and notes stating to place the planks in both spans consecutively for the length of the beams.

That became new notes on all later projects.

gjc

 

[cliff234]

Do not use D-beams on the perimeter. Use W-shapes on the perimeter, and the torsion will be eliminated. (In our office, we use W-shapes at perimeter edges (where plank frames perpendicular to the slab edges) on all Girder-Slab projects.) D-beams should only be used where you have plank on both sides of the beam. When you have plank on both sides of the beam there will be no torsion - even when you have different plank spans on each side. (The torsion is self-limiting.)