plate internal forces

[Jore-biw]

Hello everyone,
I am studying civil engineering and am currently writing my final thesis. As part of this thesis, I am supposed to create a program that enables the evaluation of the load-bearing capacity and the life cycle assessment of floor structures. I would like to use Pynite for the calculation of the internal forces, as it is the only Python library that offers plate formulations. It is about the calculation of point supported slabs. Unfortunately, I don't get such good results for the support moments of the center columns, as I work with a uniform mesh of the add_rectangle_mesh method with quad elements. Unfortunately, I could not find a possibility for local mesh refinement. Field moments and support forces provide good results. Unfortunately, I only have basic knowledge of FEM and programming in Python, which is why I have not managed to implement my own mesh refinement. Are there any approaches or solutions that I have overlooked?
Thank you very much in advance!

[JWock82]

Plates subject to point loads will report localized inaccurate stresses at the load that approach infinity as the mesh gets further refined. To put it another way, a point load (or column reaction in your case) is a force acting over an infinitely small area. The solution is to spread the load over several nodes (increasing the force’s area to something other than zero), using stiff/short members linking the column to multiple nearby nodes in the mesh.

[Jore-biw]

Thank you very muc for your answer. I'm aware of the need to use a finer mesh and to connect the columns to multiple nodes. If I use a finer mesh, which would solfe this problemthe calculation time is not accaptable anymore. Therefore I wanted to use a local mesh refinement around the columns, since the results for the field-momentums are good. Is there a solution for this? I was not able to programm a net with local refinement myself.

[JWock82]

A finer mesh will only make the plate stress results appear worse locally. You have a singularity in the problem you are trying to solve.

Pynite does not have a way to locally refine the mesh. At some point I’d like to integrate GMSH with Pynite to help with meshing.