Hello, i am basically modelling a notched tensile specimen in 1/4 symmetry. In this part of my project i have two main problems.
1) When i apply force to model, it only deforms first row of the elements. Which is causing inconsistent results. As you can see in the figures, equivalent plastic strain values of finer meshed model is almost twice the value of coarser meshed model. Hence, i am looking for a way to create a model which would yield a homogeneous strain field rather than a strain field localized in only a single element row.
2) Again, as u can see from the figures strain behavior of the material under same circumstances are completely different from each other. In this case, how can i verify my results, which mesh size should i select to get more trustworthy results, would converting my elements into a tri (3) element type help? How can i avoid high shape distortion in the elements?
Any answer, advice or source is appreciated. Thank you in advance!
Strain gradient is very high in the attached plots, You need to refine the mesh in such areas before you submit the run, or use local adaptive remeshing feature available in Marc. Under the local adaptive remesh criteria you can use equivalent value > plastic strain criteria and define a threshold so that Marc will perform a remesh in areas where the threshold is met. You can refer to Marc_UG example - e008 - Parallelized Local Adaptive Meshing .
Apart from the mesh refinements, make sure large strain is activated, And use proper nonlinear material data.
Strain gradient is high because of the localization, and it becomes higher when i refine the mesh. Unfortunately i can not use local remeshing because i need the nodal values after the simulation which i can not reach if i use local refinement. When i use finer mesh, elemental distortion increases hence results become inconsistent. As an example, finer mesh figure above, elements are clearly more distorted in the fine mesh. Hence, i need a solution for stress and strain concentration in only single row of elements. Instead, i want a way to distrubute to strain in a field no matter how much the mesh is fine.
I also know global or local adaptive remesh is a good choice here however, i do need the nodal data in postprocessing which is not possible to get when remesh is applied. That is why i am looking for a different solution to my problem.
Nodal data can be read after remeshing as well. If you want to retain nodes during remeshing, then specify the hard nodes in 15th data block of ADAPT GLOBAL card.
In case you need further assistance with the model, please log a support ticket and share the mud file.
Hard nodes option is not supported for 2D global remeshing. It is supported only in 3D remeshing. I believe attach > node to point will also not work for 2D remeshing.