Hi,
I have also noticed the same weirdness : very different bending/torsion stress results in branch connections (and sometimes in headers too), if the type of the branch is filled, or if I only manually implement exactly the same values of SIFi/SIFo.
Considering that roughly Sb = iM/Z, and that with or without the type of intersection filled, the SIFs ii/io and the bending moments Mi/Mo are the same, the fact that the results are sometimes very different lead me to question the section modulus used in branches.
After a some tests and manual calculations, I came up with the hypothesis that, surprisingly, CAESAR may not use the same branch section modulus for stress calculation, depending on whether or not the user specifies the type of intersection…???
I couldn’t confirm nor invalid this theory because is it impossible to know what section modulus the stress calculations are based on. I don’t think its displayed anywhere in the software… ?