help in regularizing the mesh

[Asad005]

Dear Stko team,
Can you please that how can I regular the mesh of the attached model without changing the spacing of bars?
Regards

[Clarabella]

Dear Asad,
please send us a picture and explain what you like to do,
Thank you

[Asad005]

The issues in "GHobarah2002 merged2.zip" is that Near to supports and top of the columns it's difficult to make the mesh quadrilateral although I selected the Quad and structural option in Mesh Control. Picture of this model are GHobarah2002 unmerged2, 2a and 2b.

Secondly, in GHobarah2002 merged3.zip, I changed the support shape from 2 triangles to 1 triangular shape but still, the problem of existence of triangular mesh within quadrilateral exists. The pictures of the model along with the model names i.e GHobarah2002 unmerged3, 3a and 3b. are attached herewith for your consideration.

In addition, how can I get the load vs displacement curve for 3d fem model? I tried to get it via Export Chart data option but it gives me a straight line with zero reaction force at all stages of analysis as shown in the attached screenshot.

[STKO Team]

In order to have a full quadrilateral structured mesh, your surfaces must have exactly 4 sides and 4 corners, otherwise, the structured mesh is not possible.

regarding the reaction forces being zero, it is probably due to the fact that you did not select the nodes where the reactions are.

[Asad005]

exactly, As I am meshing beam-column and both have four corners and four sides of the section as shown. But sometimes the size of the column section is 300 mm X300 mm while the bars have the concrete cover of 25 mm. So, in that case, it gets difficult to divide it into equal sections with a concrete cover of 25 mm when the radius of the bar is added (25+10 mm).
In addition, when support is at a location in which nodes of support are within the quadrilateral mesh then the mesh gets also irregular near to support. And the only solution is to move the support from the original position (given in experimental specimen) but in that case, the model will not present the original experimental specimen.

SO, is it possible to get regular quadrilateral mesh with bar (centre-line) placed at 35 mm distance from the face of the column? OR
I have to change the concrete cover to regularize the mesh. But in this case, the geometrical will be not the exact picture of the experimental one.

[STKO Team]

There is always a way of doing a structured mesh... it's just that sometimes it is difficult. In your case you need to extend the edges up to the boundaries, so that you have 4-sided surfaces.

If you can't do it (I know doing a structured mesh is an art), use unstructured triangular meshes.