Convergence anf mesh refinement

[mvilfour]

Hello everybody,

I'am trying to create a model of a small connecting rod . It is like a dumbbell with hole at each end to install spherical rod bearing. The boundary conditions are axial load only.
I have created a 2D model on ansys workbench. The axial load is apply with a pressure on line. This pressure is applied to one half of the spherical bearing location. This structural analysis is linear and static.
My problem is when I solve the simulation with finer mesh, there is no convergence of Von mises stresses which increase dramatically as the mesh is refined.
So I have check the maximal principal stresses and shear stresses, and the shear stresses inscrease dramatically and the principal stresses converge.

May you help me to solve this problem of shear stress which increase ?

Thank you very much for your help,

Best Regards,

Mickael

 

[Dave442]

Hi Mickael,

Have you looked at the location of the maximum stresses for each mesh? If they are located next to the spherical regions of your model it is possible that you have encountered a stress concentration. If you employ local refinement in these regions, the stress values should eventually converge. If the maximum stresses are observed next to a sharp corner or within the same location that you have applied a point load/constraint however, it is possible that you have encountered a singularity. Unfortunately these stresses will increase to infinity as you refine your mesh.

If you search this forum and google you will find a lot more details/discussion of these issues.

Good luck,
Dave

 

[mvilfour]

Hi Dave,

Thank you for your response.
The maximum stress location is the same for each mesh: The nodes on the boundary between no displacement boundary condition nodes and nodes with axial load. I have tried to refine the mesh at this location but there is no convergence.

Thanks a lot!
Best Wishes!

Mickaël

 

[shaun8567]

Posting a screen shot of the high stress location would help a great deal (including loads, constraints, and geometry). However, you stated that, "maximum stress location is the same for each mesh: The nodes on the boundary between no displacement boundary condition nodes and nodes with axial load.". If I understand this correctly, then you're saying that the peak stress happens at the constraint. If so, then this is a singularity as well, since at this location you have an infinite discontinuity in the stiffness (from the Young's Modulus to the constraint, which has infinite stiffness).