Learning Forum

[spamanon]

I am dealing with a simulation (static structural) where the forces are near the predicted buckling load for a columnar component. The model converges fine until about time value 0.97 then diverges. To remedy this, I added constant stabilization energy (ratio = 1e=4) to the nonlinear settings for the restart from time value 0.97. This got the model over the finish line, but I know I need to compare the strain energy to the stabilization energy before accepting the results. The problem is I cannot find what exactly it is that I need to compare. 

If I look at the ratio of the Maximum values listed in the Tabular Data of the contour plots, I get: (3.2301e-010) / (6.8774e-007) = 0.05%

If I look at the ratio of the Total values listed in the Tabular Data of the contour plots, I get: (3.0384e-006)/(5.0492e-003) = 0.06%

When I create a user defined result as a contour plot (=STEN/ENERGYPOTENTIAL * 100) the contour shows the Maximum value is 3.5% and the Average value is 0.08%

The documentation says that the value of the ratio should be very small and says "for example, within a 1.0 percent tolerance" but it does not define how to calculate this.  By this criterion, my measures are all good except the Maximum value of the contour plot in the user defined result. 

How can I tell if the results are acceptable with these values?  Any insight would be appreciated.

[dlooman]

It sounds like the stabilization energy is relatively small.  Since you ran the analysis to 0.97 without stabilization, could you compare the stabilization run results with that run at load factor 0.97?  You could also run a second load step at the final load to see if the results are the same.  The stabilization forces should dissipate during the second load step.