Extraction of non-convergent loadstep results

Andrea_22029

Hi everyone,

I'm trying to perform an optimization study in Hyperstudy of an inflatable bending beam. I have some experimental data (displacement-force at beam tip) and I would try to fit the Young module of the material in order to fit the experimental and simulated Force-disp curves.

I set a NLSTAT model in Optistruct, with an enforced displacement history at the tip (tableid1->tload1->dload) and I request in output the Displacement and SPCF at the tip node in order to get the disp/force curve. In Hyperstudy I try to fit the curves via "curve difference" function as in Tutorial HS-1506.

So, the model works, but in Hyperstudy I face with some problems: when the Young module changes, it could happen that the simulation does not converge; in fact, when for E=140 MPa the beam could reach a displacement of 30 mm, with E=170 MPa it could not, because it collapses before. In this case the loadstep is non-convergent, and I set the NLOUT card in order to get the results as well.

However, when I set the Data Source in Hyperstudy, I must specify the name of loadstep-result in h3d file (for example "Bending2"); in the nominal run there are not any problems, but when E changes in such a way that the simulation does not converge, it could not read the h3d file, because the new name of the loadstep is "Bending2 - NON CONVERGENT". In the same manner, if in the nominal run I set the Data Source in the "Bending2 - NON CONVERGENT" loadstep, it does not read the convergent simulation ("Bending2").

I'm interested in get both the convergent and the non-convergent loadsteps, but I cannot find how to do it. I tried to set an output with a OPTI and OUTPUT2 format, but it does not work.

Do you know how to get both convergent and non convergent solutions? Is there something like a .txt tabular output from the model, in order to avoid the h3d output format?

Thank you for your attention.

Andrea

Adriano_Koga

Hi Andrea, in newer versions there is this .monitor and _ld.monitor output files for nonlinear.

Maybe these might help you.

I would investigate further why the models are not converging, maybe limit the DTMAX for them. I don't know. This would avoid these issues.