STIFFNESS mODIFIER TO BE USED FOR CALCULATION OF DEFLECTION? 2

[dhg]

Dear friends,

If I take the stiffness modifier as follows in my analysis model as prescribed by IS:1893:2016 (Indian Code)
Column: 0.7 Ig
Shearwall: 0.7Ig ( not specified in Code)
Beam: 0.35 Ig
Slab( thin shell in ETAB model) : 0.35Ig ( not specified in Code)

I will design all the structural members for the forces, I get using above analysis model.

Then if I have to check the deflections of slabs and beams to verify that whether they are within the limit specified by IS 456: 2000, Can I use the same model ( as above) or I need to have a different model with other stiffness modifier values?

Also when we define the slabs as a thin shell, the forces in beams and columns are quite different than defining slab as a membrane, what should be the correct approach for slab modeling?

Hiren G. Desai

 

[MrHershey]

Those appear to be almost exactly the ULTIMATE stiffness factors given in ACI 318 (US concrete code) that should be used for forces/stability. The main differences are shearwalls should be 0.35Ig if they are 'cracked' (axial-flexural tensile stress exceeds concrete modulus of rupture) and that the slabs are given as 0.25Ig instead of 0.35Ig.

Using these same stiffness factors for deflection determination will be conservative, but generally not appropriate. While ACI 318 does not address SERVICE stiffness factors to be used for deflection determination, ACI 318 commentary indicates that the provided ULTIMATE factors can be converted to SERVICE if divided by 0.7.

Based on this, your SERVICE stiffness factors would be:
Column: 1.0 Ig
Shearwall: 1.0 Ig (0.5 Ig for cracked walls)
Beam: 0.5 Ig
Slab: 0.5 Ig (0.35 Ig per ACI 318)

Would note again that these numbers are based on commentary to the US code. Various firms will use various values and I don't know what the IS may stipulate.

As far as slab modeling goes, it depends. If I'm in a high seismic zone where my forces have been reduced by a reduction factor (R in US codes) to account for inherent ductility of whatever system I've chose, I tend to model either as membrane or as thin-shell with an artificially low stiffness factor (say 0.1 Ig or lower). This is to keep the forces in the primary force resisting elements rather than getting some benefit from the slabs acting as frame members when slabs typically are not provided with the same level of detailing that merits the reduction in seismic force we're using. For wind applications where your force level is not dependent on your level of detailing, I don't have as much problem keeping the slab stiffness in the model as long as the moments and forces the slab attracts are actually designed for.