The effective section properties change with loading, so that isn't it.

AISI give specific criteria for how to evaluate the properties of a section (using line integrals). These formulas have changed with the latest code provisions, I believe. So, that is probably the reason for the differences.

Josh-
Is it incorrect to input the properties for a custom Z shape from RISA Section into RISA?
In other words, does AISI require that line integrals are used to determine the section properties?

The reason I ask is the section properties from RISA Section can be significantly different than those calculated directly by RISA.

I am not sure which is the proper way to proceed.

Thanks Josh!

Toad -

I just ran through some quick tests by creating a custom cold formed section in RISA-3D and the same custom cold formed section in RISASection. I got basically the same results.... within about 0.6%. This minor difference is likely due to the changes to the AISI line integral method. The newest version of this method is in RISA-3D. But, I don't think we've released RISASection since we implemented those changes.

Some thoughts:
1) RISASection uses different algorithms for calculating the properties depending on the shape type.

2) Use use the AISI line integral methods when you select Cold Formed from the Basic Steel Shapes than it uses for other shapes.

3) If you use a DXF imported shape or a general shape then we are using a more generalized formula.

For AISI cold formed shapes, the most appropriate method is to use the AISI method. Personally, I'm not positive that this is truly the most "accurate" method in all cases. However, I imagine there is a reason why AISI requires it. Perhaps some of the traditional methods fall apart for very thin sections.

I take it back.... The AISI code does NOT require the line integral method in the code. This actually comes from their Design Manual and it is just a suggestion for simplifying the calculation. In the manual, it is acknowledged that some error is present with this method. But, that error shoudl be insignificant for the thin walled sections given in the AISI manual. If your section is considerably thicker than the ones shown in the SSMA or AISI databases then that could explain where there error is coming from. Reading through that section of the manual it seems that the error in the line integral method will tend to under estimate the section properties.

RISA uses this method because that way, the program can exactly match the calculations and tables given in the AISI design manual.

I'll have to look again Josh, but I added a custom stiffened Z-shape (ZS) by adding a custom AISI shape in RISA under the cold formed tab and had it calculate the section properties.
I then made the identical shape in RISA section by importing a DXF from CAD and the section properties were quite different than those given by the custom AISI feature.