I am trying to model a deep tunnel in rock mass in ABAQUS. I am using an elastic-plastic UMAT model. In that model, pressures are applied from top horizontal surface and vertical right surface. Condition of symmetry is applied along vertical left surface and rollers are applied at bottom. The domain width is 15 times of tunnel radius. Displacement (U) contour of the model analysis is provided with the figure attached below. In that figure, top right corner elements distorted heavily. It is like, those elements were squeezed out from FE domain.
Can anyone help me how to solve that distortion of top right corner element in ABAQUS?
Hi Dipaloke,
The deformed shape looks as though u1=u2=0 were prescribed at the top right corner. Could that really be the case?
Victor
Hi Victor,
At top right corner nodes/elements, the u1 and u2 should not be equal to zero. That element should move downwards ( - u2) due to pressure from top surface and also move towards left ( - u1) due to pressure from right hand side surface. Hence, I provided no boundary conditions at that nodes for moving freely under the loads. But, due to concentration of high stress the corner element is getting squeezed out which need to be prevented.
Dipaloke
Your model is to simple and the edges are never firmly fixed. BEM in an infinite block would be better.
You can try more "stiff" scheme using kinematic coupling constraints (i.e. the same u2 at all nodes of the top boundary, the same u1 at all nodes of the right boundary, the corner node is the reference node).
Victor
Your deformation scale factor is 7. Set it to 1 if you want to interpret the plots (somewhere in the common plot options).
Prof. Nichols,
Thank you for your reply.
By 'BEM' if you are indicating to boundary element method, that may not be useful in my case. As my main objective is to show application of my material model through finite element modelling. Besides, one can surly use infinite elements in ABAQUS. But, in my case that may not be used as loads are applied applied on surfaces where infinite elements can be placed.
Prof. Аdlucky,
Thank you for your suggestion. I will reply you after trying your solution.
Prof. Baeker,
Thank you for your reply.
I have checked those results with scale factor equal to 1. However, in those attached figures with the question, I used scale factor greater than 1 (In one figure it is 7) for making displacements/distortions visible to everyone. The distortion/displacements of those corners elements are affecting my results especially displacement contours and plastic strain contours.
- Badges
- Science topic