Super-T torsional stiffness values in Appendix of AS 5100.5

[kww2008]

Does anyone know the source of the torsional stiffness values tabulated for the standard Super-T sections in AS5100.5-2004 ? I wonder which software application was used to determine these values as I used two software applications (both using membrane-analogy method) and the values I got from using them differ by about 10%.

 

[IDS]

I don't know the source of the figures, but I have done my own calculation using Strand7 with the following results:

The last line is the appropriate comparison because the note to Clause E4(ii) states that the contribution from the cast in place deck slab is reduced to one half of the full amount, because "continuity of the top slab removes the effect of the vertical shear stresses that would otherwise be present at the free ends of the slab".

My comments are:

1) The Strand7 values for area and Ixx were within 1% of the tabulated values. I presume the difference in the J values is because the code values are based on a theoretical approach ignoring distortion, whereas the Strand7 values are based on an FEA using an automatically generated fine mesh.

2) The differences are not hugely significant, given that the code requires 20% of the calculated uncracked value to be used for the ULS, but the figures are easily calculated, and the best available values should be used.

3) It is not immediately obvious to me why continuity of the top slab would reduce its contribution to the torsional stiffness, but presumably the theory was reviewed at the time the code was written. In any case a reduction in the contribution of the top slab seems appropriate since it is not prestressed and will crack before the pretensioned girder.

4) The next version of the code is currently out for public comment (closing 17 March), and I will suggest that the torsion values should be reviewed.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/