Non linear optimization

Hi all,

I'm working on a component with non linearities (many contact, large deformation). It works on Radioss and I have results, now I'd like to run an optimization (composite shuffle) for find the best stacking sequence. How can I do? Are there any examples to watch?

Best regards

Find more posts tagged with

English

OptiStruct

Accepted answers

All comments

Altair Forum User

Hi @Virgilio Lo Presti

You can do optimization within RADIOSS. Starting from 13.0 optimization is added in RADIOSS.

You can refer to below example on RADIOSS help for methodology

<?xml version="1.0" encoding="UTF-8"?>

Altair Forum User

Thank you very much!

Altair Forum User

Is the composite shuffle optimization supported?

Altair Forum User

@Virgilio Lo Presti

No, it is not in the current version of HW. Can you explain your use case and what you are trying to optimize so that I can check and suggest the best way to deal with it?

Altair Forum User

I've a B-Pillar reinforced with cfrp. There is an intruder (similar to the figure posted from you) that have an applied constant velocity (1m/s). I run an explicit radioss simulation from 0s to 0,2s (0.2m intrusion), and I compared the reaction force on the intruder. So, I have a graph force-time of this reaction force of the reference model and I need to obtain a similar curve with the cfrp removing local metal reinforces. Actual curve is under the reference curve and I set up a possible stacking sequence (there are 15-20 UD plies): can I optimize the stacking sequence to reach the best sequence?

Altair Forum User

@Virgilio Lo Presti

ESLM can be used for optimization under dynamic loads. But this method only supports PSHELL and PSOLID. So if you are using any orthotropic property like PCOMP(G), PCOMPP, it is not supported right now.

Altair Forum User

Ok...thanks for your support....I'm using hyperstudy

Faisal

/index.php?/profile/45116-virgilio-lo-presti/&do=hovercard' data-mentionid='45116' href='<___base_url___>/index.php?/profile/45116-virgilio-lo-presti/' rel=''>@Virgilio Lo Presti

No, it is not in the current version of HW. Can you explain your use case and what you are trying to optimize so that I can check and suggest the best way to deal with it?

Hello,

Does it support now the current version of the hypermesh optistruct? Shuffle optimization with large deformation non linear laod case.