Strange results of building

mgaafar
Advocate
Advocate

Strange results of building

mgaafar
Advocate
Advocate

Hi,

   I was modelling a multi-storey precast building where the walls are not connected together so I applied an edge release (released everything) at the corner between non-connected walls.

I noticed that at the support level, the corners also are free as if robot assumed no support at the corner of both walls), which is fine to me.

In order to minimize the effect of this free corner I refined the bottom mesh at the support level to minimize the free unsupported corner error. When I did that, the results totally changed in an un-realistic behavior.

in particularly two adjacent walls parallel to each other, on is 10.5m long (Wall/panel no. 19) and the other is 7m long (Wall/panel no. 20), in the first run with coarse meshing, the shorter wall attracted about 610 kN shear load while the longer wall attracted 1325 kN shear load which makes sense. now when I refined only the bottom coarse mesh, both walls gave similar shear load value of 715 kN and 798 kN for the shorter and longer wall, respectively.

The only difference is that I refined the lower mesh of the walls (not ever the entire wall)

Does anyone have an explanation?

coarse mesh.jpg  fine mesh.jpg

 

 

 

0 Likes
Reply
441 Views
2 Replies
Replies (2)

Rafal.Gaweda
Autodesk Support
Autodesk Support

Hi @mgaafar

 

These results strongly depend on meshing.

 

"The diagrams are presented along the vertical axis cutting through the gravity centers of the core wall section on every story. Two cuts are performed for each story: above the bottom story edge and below the top story edge. Each cut is made at the CENTROIDS OF SURFACE ELEMENTS as shown in the drawing area."

 

See this help topic.

 



Rafal Gaweda
0 Likes

mgaafar
Advocate
Advocate

Thanks Rafal for your response, I would agree with you if I am viewing the moments on the wall which will then be a function in the section cut location(height), but I am viewing a shear force diagram which is almost constant along the height of the floor.

Also, the mesh is refined for the entire bottom course of all the walls in the model, so the force distribution should remain the same for all walls based on their stiffness which did not happen (shear force for a 10.5m long wall is almost the same as a 7m long wall when the mesh is refined). 

I hope I am explaining the problem clearly.

0 Likes