Reanalysis of Topology Optimization Results

Altair Forum User
Altair Forum User
Altair Employee
edited October 2020 in Community Q&A

Hello,

i need your help...

 

i have the topology optimization results of different FEA Solvers. Now i would like to reanalyze it in hypermesh.

All optimizations based on the same mesh. For the reanalysis i have the .fem Modell.


For each optimization i have a result file in the form:

element_id  density
1                   0.1
2                   0.121
3                    0.9
...

Its a file with several thousand lines..


I would like to do this steps:

1)load the .fem file
2)load the result file
3)create components for the elements with the same density
4)create properties for every density
5)assign the property with the corresponding density to the elements
--> reanalyse


I dont know how to implement this. I have no experiance with programming in hypermesh. Is there a way to do this?

In return, I will present my results here.

 

Thany you


Regards
Andreas

 

Answers

  • Altair Forum User
    Altair Forum User
    Altair Employee
    edited September 2016

    Hi Andreas,

     

    Instead you can use OSSmooth module to re-analysis.

     

    Refer to the Tip and trick for more information: http://www.altairuniversity.com/learning-hyperworks/tips-tricks/optistruct/optistruct-ttf-reanalysis-for-topology-and-topography-using-ossmooth/ 

  • Altair Forum User
    Altair Forum User
    Altair Employee
    edited September 2016

    Hi Prakash Pagadala,

     

    thank you for your answer. I would like to do a reanalysis without smoothing. With smoothing i cant compare the results because smoothing has a high Impact.

     

     

    Is it possible to create a macro?

     

    Thank you

     

  • Altair Forum User
    Altair Forum User
    Altair Employee
    edited September 2016

    Andreas,

     

    Does your model has solid elements?

     

    I will check if there is any script available which can create properties and components based on a text file which has thickness, element ID etc...

     

    If not, you may have to create your own macro :)/emoticons/default_smile.png' srcset='/emoticons/smile@2x.png 2x' title=':)' width='20'>

  • Altair Forum User
    Altair Forum User
    Altair Employee
    edited September 2016

    Yes, it has solid Elements. It would be great if you find a Script.

     

    Otherwise, is it possible to give me some advise to create a Script in hypermesh? I have no experiance and dont know how to start..

     

    thank you 

  • Altair Forum User
    Altair Forum User
    Altair Employee
    edited September 2016

    Andreas,

     

    I will check if there are any scripts available pertaining to your case.

  • Altair Forum User
    Altair Forum User
    Altair Employee
    edited September 2016

    Hi Prakash Pagadala, 

    have you found a Script to solve my Problem?

     

     

    Thank you

     

    Andreas

  • FabioC
    FabioC Altair Community Member
    edited November 2018

    Dear users

     

    I am aware this is an old post, but this is the most relevant conversation I have found about my problem and I would like to share my solution with those who will have the same doubt in the future.

    My problem is to re-analyse a model assigning the element densities (from file .*sh) generated after topology optimisation.

    The reason for the re-analysis is explained by the fact that some boundary conditions change and I need to apply the new conditions on the optimised structure.

    Obviously the mesh does not change.

    OSSmooth is not a viable solution in my case.

    What I have done is.

    1) Run topology optimisation, introducing an OptiControl on the MINDENS parameter (for example MINDENS=0.1).

    2) Save the .*sh file which relates the elements ID to the density value.

    3) Modify the *.hm and *.fem file updating the boundary conditions as I needed.

    4) Then from OptiStruct I selected restart-->reanalysis and at this point I use the updated *.fem file and the *.sh file previously saved. I also set a density threshold < MINDENS.

         If you check the Optistruct run options you will find that imposing a threshold < MINDENS makes sure that Optistruct restarts from the density values in the *.sh file.

     

    This work fine for me. I do not know if there is a clever solution to this specific problem.

    Anyway, hope this can help.

     

    Regards