Non-Linear Analysis - Compression Resistance

[JeremiahVanderlaan]

I am having an issue with a non-linear analysis regarding compression resistance.

I am performing a non-linear buckling analysis for a steel truss, with a wind uplift which will result in the bottom chord of the truss in compression. Intuitively, the bottom chord will be the member to buckle as it is long and slender and in compression.

I have placed Interacting P-M2-M3 hinges at three places along the center bottom chord (Relative locations of 0, 0.5 and 1). The hinge should activate whenever the compression force, Cf, is greater than the compression resistance, Cr, in the bottom chord and attempt to redistribute the stresses elsewhere.

But what I am noticing is that the hinge is not doing anything until the stress in the bottom chord reaches the yield stress. Which is correct if the hinge was looking for tension, but since compression resistance takes into account the unbraced length of the member, Cr < Tr, meaning that the stress at a compression failure can be much less than the yield stress.

So, I realize that a compression failure is due to Euler Buckling, and it is not plastic deformation, and that is most likely why the hinge is not doing what I want it to do (Redistribute forces when Cf > Cr). Does anyone have experience with something like this and can suggest a way to get what I want?

Thanks!

 

[stressed]

Jeremiah, did you run NL P-Delta or P-Delta with large displacements which accounts for changes in element stiffness? Hinges account for plastic deformation. Buckling case (Eigen) is a linear analysis although you can base the stiffness on a NL P-Delta case. NL analysis with P-Delta large displacement is another approach for analysis of buckling. Attach your .sdb model if you're still having problems.

 

[JeremiahVanderlaan]

Stressed,
Thanks for the quick respsonse.
So the analysis is NL P-Delta with large displacements.
I tried a very simple analysis where I reduced the section size of the bottom chord, and I made it a pin-pin member (before it was fixed). I did this to encourage a compression failure before overall buckling of the truss.

When I ran the analysis I noticed that this member exceeded the Cr (128 kN), it went up to about 158 kN. The deflection at the center of the member was definitely non-linear, however the analysis still let this member get fairly over stressed.

I originally thought that maybe it would allow the member to get up Cr = phi*A*Fu(1+lamda²ⁿ)^-1/n instead of the regular Cr = phi*A*Fy(1+lamda²ⁿ)^-1/n but this isn't the case as I tried a few different sections with varying material properties.

Do you, or anyone else, know how the Non Linear Analysis in SAP determines when a compression member cannot take anymore force?

Thanks.