four search schemes in reliability analysis

civilmanjun · Post by **civilmanjun** » Wed Dec 12, 2007 2:56 am

Hi everyone,
I used four search schemes,repectively, but only the improved HLRF algorithm run properly, others didn't work. please help me ,it's a big problem for me, and I tried my best to debug them , but failed finally. Please help me, I'll appreciated.

koduru · Post by **koduru** » Wed Dec 12, 2007 4:20 pm

Hi civilmanjun,

Please post the errors with other search algorithms and results you obtained with iHLRF.

I am not sure if you have syntax errors (running OpenSees) with the other search algorithms or the convergence problems (getting correct results).

Depending on the limit-state function in your problem, different algorithms may actually fail to converge.

If you have simply problems with syntax errors then please post them. I think the changes in new version of OpenSees might have something to do with that.

Cheers

civilmanjun · Post by **civilmanjun** » Wed Dec 12, 2007 10:53 pm

[color=red]the result using the improved HLRF algorithm[/color]

design point found!
Done analyzing limit-state function 1, beta=2.69326
FORMAnalysis completed.

koduru · Post by **koduru** » Thu Dec 13, 2007 10:18 am

The steel01 material is bilinear. Therefore the limit-state function may suddenly change before and after the yielding.
In that case, the convergence in FORM analysis is not guaranteed.

From your errors I could see that other search algorithms failed to converge. This is not a problem with the OpenSees. This is the problem with FORM and how your problem is defined.

Perhaps you could used a different uniaxial material with smooth transition to the yielding. This could resolve FORM problem.

civilmanjun · Post by **civilmanjun** » Thu Dec 13, 2007 4:13 pm

koduru
thanks for your advice, I'm so appreciat you answer so many questions for me. thanks.