You can either
- use a linear buckling analysis. It's a very quick assessment of the critical load that will trigger buckling of your structure. However, it usually overestimates the load and in real life, buckling may occur at lower load. This analysis will only give you the shape of the structure at the beginning of buckling (no post buckling) and an idea of the critical load.
- Use a non linear static analysis and perform a actual buckling case. It is much longer but more accurate. You will get disp and stress results for each step of the buckling process (including the collapse of the structure if needed).
Linear buckling is part of Simulation Pro, Nonlinear is part of Simulation Premium.
Thanks for your reply.
The output of linear buckling analysis is critical axial load and it doesn't give you buckling capacity due to bending. What I am actually looking for is lateral buckling due to bending moment, not axial compression load.
Do you think in non-linear static I can get any result for torsional-buckling?
Thanks for your post. In buckling analysis, can we have a combination of moment and axial loading?
Second question, can I run the same design scenarios (your attached images) in non-linear analysis? If so, how much will the deviation be from buckling analysis? We don't have simulation pro or premium here, so I can't try some of these options. However, I like to know about available options in SolidWorks for different loading conditions.
maybe this will be helpful. Linear buckling gives you the buckling load for the undefromed shape. So called "non-linear buckling" is performing a linear buckling analysis on a deformed shape and it get you closer. However, unless the aforementioned load is the load right before it buckles you are goig to have errors. Generally it goes like this, as the load is applied the shape deforms and the eccentricities get larger which begets larger bending moments which eventually leads to compression in some area of the structure more than some other area. When the compression load gets large enough the stiffness of the sturcture (local or global) reaches a critical value of zero stiffness and it buckles. Thus, in order to capture buckling you need to incrementally load the structure and at each load increment you need to compute the stiffness of the deformed structure, then using the new stiffness you can apply some more load and then do the same thing. You do this till you get where you want to go. This in general is non-linear analysis - the stiffness isn't constant so you have to keep computing the new structural stiffness as you go. The case of "post buckling" - going past the buckling point is required to get close estimates of the actual buckling load. In other words you have to buckle it. Buckling is a tad tricky numerically due to the rapid rate of change of stiffness that occurs. Specialized controls are required to guide the numerics - Simulation uses the term arc length control which has a couple of variations - Crisfield & Riks (named after the guys that wrote the papers covering their respective approaches - not sure which one is used in Simulation). Further, these methods apply a load pattern multiplier so all the loads are considered in the same proportion regardless of hte load magnitude at any particular step so you can have multiple loads in any combination of directions if that describes your guess at the intended reality. Getting tight estimates of buckling loads is non-trivial in most cases.
The ASME uses a buckling load fator of 4 for pressure vessels on a linear buckling estimate as far as I am aware, and be advised I am not a pressure vessel guy so don't take that number as gospel but at least it gives you some sort of gauge on the uncertainty of linear buckling estimates - they can be pretty large.
Anyway I hope that helps and provides you a frame work to get more specific with your questions.