Is this the right way to go from linear to nonlinear buckling analysis? [COMSOL] 1

[tugni925]

I ran a linear buckling study for a structure, and now want to do similar study but make it a nonlinear buckling analysis. What I want to do is described here:

I did the original buckling analysis according to this:

Would it be correct to just include geometric nonlinearity at both steps and run it?

 

[FEA way]

With those settings you would only account for geometric nonlinearity in preload step but the buckling analysis itself would remain linear. If you want to perform truly nonlinear buckling analysis, you need just the stationary study step (with geometric nonlinearity turned on of course).

 

[tugni925]

So I just delete "Step 2: Linear Buckling", and leave the stationary step as is, since geometric nonlinearity has already been enabled?

 

[FEA way]

Yes, you can do it this way to solve static analysis with geometric nonlinearity.

 

[tugni925]

"With those settings you would only account for geometric nonlinearity in preload step but the buckling analysis itself would remain linear."

Could I ask why the buckling analysis itself remains linear when I have specifically enabled geometric nonlinearity in the linear buckling step?

 

[FEA way]

Because it's a linear buckling step and nonlinearities can't be included directly. This geometric nonlinearity option is used for compatibility with nonlinear preload analysis. Details (along with equations) are explained on Comsol's website. Check the documentation as well.

 

[tugni925]

Thank you FEA way, if I may ask a last question to close off this thread: when I run the study it finishes but I dont get the mode shape. Do you know why this might be?

 

[FEA way]

You won’t get mode shapes from this type of analysis, they are reserved for linear buckling simulations. So you can delete this plot and create new ones for stresses and displacements.