Homogenization with Finite Size

  1. Eren Açar (Student)

    Hello all, 

    I am having a problem with homogenization of a squarepack (block with a hole) using SwiftComp ANSYS Workbench GUI. I tried both symmetrical mesh and default quadratic mesh for both 2D and 3D models of my unitcell. However, I could not get any homogenization results for the 2D model. For the 3D model "Solid Homogenization" cases with aperiodic in Y direction is selected, I always got the same result (which seems like isotropic properties) for fully periodic homogenization and aperiodic in one direction.

    I could only achieve fully periodic solid homogenization with default quad mesh. Other than that, I cannot obtain finite size effect even if I use symmetrical mesh. 

    I showed the errors and homogenization results in the attached document. Any suggestion or solution is highly appreciated.

    Thank you in advance

  2. Banghua Zhao

    Hi Eren,

    For 2D model, if I remember correctly, the in-plane directions is y and z directions. Does the 2D model in the y-z plane? (The error shows that the shape of model is abnormal. So I guess the coordinate might be incorrect) Please double check with that.

    For 3D model, it is surprising that the symmetric mesh gives a result of isotropic material, which, however, is expected to be the same as Quad mesh. I would suggest to double check with Abaqus SwiftComp GUI.

    Best Regards

  3. Eren Açar (Student)


    Thank you for your response.

    I used the same directions for 3D and 2D model. I will check the directions as you said.

    But, for the same model, is it possible to analyze aperiodicity in in-plane direction or is it only possible for out-of plane(x in tihs case)? i.e. if in-plane directions is yz, is it possible to get correct result if I select aperiodicity for Y direction?


    Best regards

  4. Anonymous

    I used TEXGEN to conduct a macro-homogenization of woven composites. I chose a 2x2 weave to determine the mechanical properties (E11, E22, G12, etc.). After changing the number of layers (thickness direction) of the weave, I noticed that the G values remained almost the same.


    Thank you for your help.

Add Post

User photo

You must be logged in to comment.

Please keep comments polite and on topic. Offensive posts may be removed.