Learning Forum

[Ravil Gizatulin]

Please help me figure out why I can't achieve convergence. Nonlinear strength analysis, compression of a thin-walled profile. Length 800 mm, thickness 1 mm (photo 1). Material "Structural Steel NL", the large displacement was turned on, real imperfections (the model was obtained by scanning and reverse engineering). Boundary conditions: constraints on one side displacements XYZ, displacements XY on the other, displacement X rotation YZ on the longitudinal edges. Mesh size 4 mm. Load applied to the edge as a displacement of 10 mm. 

Once I managed to obtain a solution with compression up to 5 mm (photo 2 - force(N)/deformation(mm) diagram). Then I couldn't repeat the calculation with the same mesh size and step settings (unfortunately, the first solution was not saved, only the graph was exported). Specifically, at the point where the force drop occurs on the graph, convergence does not occur in further attempts to solve. What I tried to do:

• refine the mesh in general, refine the mesh in the plastic hinge zone
• increase the number of steps
• simplify the mesh by changing the "Capture curvature" settings in the sizing mesh
  ERROR message I have: "excessive thickness change", The solver engine was unable to converge on a solution for the nonlinear problem as constrained.  Please see the Troubleshooting section of the Help System for more information.
  
  WARNING message I have: The reference convergence value may be less than the threshold, you can overwrite the minimum reference value by specifying it under the nonlinear controls of analysis settings.
  
  Previously, with a profile obtained by extrusion of the contour of the cross-section and imperfections based on LBA, the calculation was successful, there was no such drop on the graph.
  
  

[Ravil Gizatulin]

UPD: the problem was solved by applying the stabilization method. The stabilization energy - strain energy ratio should be taken into account. 

The semi-implicit method also can be used in the situation of buckling and force drop (https://www.padtinc.com/2019/10/09/ansys-mechanical-semi-implicit/), however, this method is time-consuming and can be effective if the solution did not converged after 80-90 % of analysis.

More info about implicit and explicit analysis: https://caendkoelsch.wordpress.com/2016/05/07/blog-3/

What can help to identify problematic zones is Newton-Raphson Residual in ANSYS. It shows where the mesh could be refined.
More info: https://www.padtinc.com/2012/10/10/overcoming-convergence-difficulties-in-ansys-workbench-mechanical-part-i-using-newton-raphson-residual-information/