Asking for suggestions about generateing surface mesh with different mesh size from the voxel size?

[Xiang Zhang]

Hi everyone,

I use DREAM.3D for polycrystal microstructure generation and then generate surface mesh (stl files) and then volume mesh to form a abaqus inpfile.

Currently, it seems to me the surface mesh size is the same as the voxel size which is fixed when generating the microstructure. Now I need  meshes correspond to the same polcycrystal geometry but with different sizes , so some of them will need different mesh size from the voxel size (for example, in the case of a mesh convergence study).

 I can solve  this problem by processing the stl file to insert extra nodes to refine the sufrace mesh, from which the corresponding volume mesh can be refined. Before start writing a code to do this, I am just wondering, whether does DREAM.3D  already have the functions to allow user specified surface mesh size, or is it possible to generate the same polycrystal microstructures but with different voxel sizes?

Thank you and I appreciate your thinkings and comments!

best,

Xiang

[Sean Donegan]

Hi Xiang,

You are correct, the resulting triangle size when generating a surface mesh is a function of the size of each voxel.  Currently, we do not have any built-in tools that allow mesh refinement or decimation, although we may get a mesh decimation tool in the near future.

DREAM.3D does have the capability to down-sample or up-sample an image geometry however, so you can resize a synthetic polycrystal microstructure.  The Filter is called "Change Resolution".  You can, for example, create a synthetic structure that has a resolution of (0.5, 0.5, 0.5) and voxel dimensions of (128, 128, 128), and then feed this into the "Change Resolution" Filter.  In that Filter you could set a resolution of (1, 1, 1), and this would have the result of down-sampling the existing microstructure to a grid that is (64, 64, 64).  The structure itself would be exactly the same, it would just have fewer voxels and thus be a lower resolution.  You could then surface mesh the structures to conduct your convergence study.

hope that helps,

-Sean

[Xiang Zhang]

Hi Sean,

Thank you very much for your confirmation of my guesses and pointing me to a convernient way to get my problem solved. Yes, the "Change Resolution" filter exactly solve my need for different mesh sizes corresponding to the same synthetic structure and is more convenient than my previous thought of writing a surface mesh refinement code. I believe your input will benefit many others as well for those who come to a point like me! Thank you again for your help!

Best,
Xiang