Hello,
I have convergence problem when using NLGEOM key word for a structural simulation of an elastomer submitted to large deformation. To try to identify the issue, I have simulated something extremely simple:
A 3mmx3mmx30mm sample submitted to uniaxial tensile stress. One of the two small areas is fixed in every direction, and a force is applied on the opposite surface to induce an elongation of the probe. To simplify, an isotropic material model is taken, with a Young's modulus of 0.8MPa and a Poisson ratio of 0.485. The idea would be to stretch this sample to >2 times its initial length (the material in question can sustain stretch ratios up to 10).
I have no convergence problems when I do not use NLGEOM, but the results are correct only for strains <~5%. Above, I need to use NLGEOM, but past a stretch ratio of ~1.6, I can not obtain convergence anymore.
I have tried to refine the mesh, to change from mapped to free meshing, changing the number of substeps, use ramped loading, change the element (solid185, solid186), without success. Theses modification only slightly changed the last converged substep number.
The situation here is trivial: No buckling, no contact elements, no complex shape. I can analytically calculate the results, but yet it is impossible to simulate this simple situation with ANSYS?
I would welcome any suggestion as to what I should try to do to obtain convergence for this simple case.
I thank you very much in advance.
Samuel
I have convergence problem when using NLGEOM key word for a structural simulation of an elastomer submitted to large deformation. To try to identify the issue, I have simulated something extremely simple:
A 3mmx3mmx30mm sample submitted to uniaxial tensile stress. One of the two small areas is fixed in every direction, and a force is applied on the opposite surface to induce an elongation of the probe. To simplify, an isotropic material model is taken, with a Young's modulus of 0.8MPa and a Poisson ratio of 0.485. The idea would be to stretch this sample to >2 times its initial length (the material in question can sustain stretch ratios up to 10).
I have no convergence problems when I do not use NLGEOM, but the results are correct only for strains <~5%. Above, I need to use NLGEOM, but past a stretch ratio of ~1.6, I can not obtain convergence anymore.
I have tried to refine the mesh, to change from mapped to free meshing, changing the number of substeps, use ramped loading, change the element (solid185, solid186), without success. Theses modification only slightly changed the last converged substep number.
The situation here is trivial: No buckling, no contact elements, no complex shape. I can analytically calculate the results, but yet it is impossible to simulate this simple situation with ANSYS?
I would welcome any suggestion as to what I should try to do to obtain convergence for this simple case.
I thank you very much in advance.
Samuel
