Hi all,
I have an issue as shown in the attached picture that occurs at the very beginning of a static non-linear column buckling analysis. I am not sure why my mesh has a local distortion there which causes the stress to be concentrated and subsequently causes the column to buckle locally instead of globally. Note: In the attached picture, the location where this is happening is not at the midheight of the column, which doesn't make sense too.
Does anyone know how I can make sure my mesh is smooth (i.e. no out-of-plane deformation) at the very start of the analysis? I have double-checked my mesh in the mesh module and it was perfectly smooth, however, when I look at it in the visualization module at increment 0 (before loading is applied) my mesh became distorted.
Could it be the global imperfection in the shape of a half-sine wave that I am imposing on the column that caused this little indent on my column? Or What could be the reason that caused this indent in my column?
Many thanks for your help,
Heng
Hello Si Hwa Heng ,
it would be helpful if you could provide *.inp file or a more precise description of the model. It's very hard to make any judgments just by looking at a picture. But looking at the illustration provided, I can recommend increasing the number of elements in the thickness direction.
Sincerely,
Pavlo
Hi Pavlo Pavliuchenko ,
Thank you for your reply.
I think I have found the issue to my problem. I will write it here so in case someone came across this and has a fix.
So the problem was actually my fault. I did 2 separate analyses and 1 for Eulerbuckling, the other for non-linear buckling analysis with geometric imperfection input from the previous Eulerbuckling analysis.
Here is the problem: My mesh for Eulerbuckling analysis is slightly different to the mesh for non-linear buckling analysis and therefore the when ABAQUS overlap the normalised first buckling mode obtained from Eulerbuckling analysis to my non-linear buckling model as a geometric imperfection, it distorted the initial geometries of my mesh for non-linear analysis which causes the stress to concentrate at these distorted locations.
Advice: Always copy the model from Eulerbuckling analysis (or any step you will be referring to later in your next analysis) and modify it for non-linear buckling analysis so 2 models are the same except for the step and loading.
Regards,
Heng
- Badges
- Science topic
- Similar topics
- Analytical Chemistry
- Column