Buckling Analysis

Hi everyong,

I'm testing out how varying different cross sections of selected structural elements may affect the deformation of a cross braced building and to test for buckling.

However, while trying to use the buckling mode component, the message "no normal force NII defined for calculating the geometric stiffness of the elements of the model" appears. Moving on, while performing a ThII analysis as recommended, the process halts with a message "the structural system buckles under given loads".

Therefore, the questiong I have is:

How do I identify what causes the buckling, at which point of the structure, and in so, go about rectifying it?

Best,

Tim

Attachments:

CrossBracedBuilding.3dm, 216 KB
CrossBracedBuildingAnalysis.gh, 50 KB

Replies to This Discussion

Permalink Reply by Romain Mesnil on April 6, 2016 at 1:54am

Hi (again) Tim,

you need a second order analysis to perform buckling (which is the moment where linear calculations are not valid anymore). If the load you apply is superior to the buckling load, then the structure is unstable and the analysis does not converge.

You should try to decrease the applied load, and it will converge if the model is good. If it doesn't converge, look at a first order analysis to see if the structure behaves as expected.

Then, buckling is about bending stiffness of the elements and buckling lengths. You have different solutions: using bigger cross-sections, add bracing to decrease the buckling lengths, change the boundary conditions...

Best,

Romain

----

Thinkshell / Design by Data

Permalink Reply by Tim Lum on April 6, 2016 at 3:17am

Hi Romain,

Thanks again for the reply.

Does this mean that rather than trying to determine potential buckling from the amount of displacement of beams of 1st order analysis, it can only be determined from the convergence of 2nd order analysis?

On another note, while trying to understand how the buckling occurs under wind load, the below 2 top views are being compared. Left - model view after being connected to the 2nd order analysis, and right - model view after being connected to the buckling mode component (after 2nd order analysis as well).

Does this mean that deformations that results to the buckling are those reflected on the right image?

Thanks ahead!

Best,

Tim

Permalink Reply by Romain Mesnil on April 7, 2016 at 2:26am

Hi Tim,

You need to perform a second order analysis to perform linear buckling analysis. It is called linear buckling analysis because it solves an eigenvalue problem (you assume the behaviour of the structure to be fully linear before the structure buckles).

Don't take the result of a second order analysis for granted if you don't put geometrical imperfections (in reality, beams are not perfectly straight for example, which will decrease their buckling capacity). Linear buckling analysis captures possible instabilities of the system, it should be your reference in the first steps of the design.

You can read two references on buckling analysis here and here. I used the references of Professor Bathe's book Finite Element Procedures in my Master's Thesis, so I suggest that you have a look at it if you want to have a more complete overview of the problem. There is a chapter in Finite Element Procedures on buckling and nonlinear analysis, it is very well written and it provides all the necessary informations on the topic.

Best,

Romain

----

Thinkshell / Design by Data

RSS