3D synthentic microstructure based on 2D EBSD or Dummy 3D synthetic microstructure

[NM]

Hello all,

              I am new to using Dream3D. I do not have serial sectioned EBSD data yet. I have read in one of the topics that it is not possible to construct a 2D synthetic microstructure with triangular meshing elements using 2D EBSD. 

1. Is it possible to construct a 3D synthetic microstructure based on 2D EBSD?
   
2. Alternatively, is it possible to construct a random 3D synthetic microstructure (with different columnar grain thickness and different sized equiaxed grains for an FCC structure) without any experimental EBSD data?
   

[Paul Kern]

1. Dream.3D may be unable to do this, but it should be relatively easy for you to create a mesh from your 2d scans. I would extract the grain boundaries as discrete points, use a Delaunay triangulation to create triangles, find the edges between different grains (referring back you your EBSD data) and preserving only those edges from your triangulation. You should then be able to pass this information to something like GMSH to create a 2D mesh of the desired density. Alternatively, you can create a 2.5d mesh by repeating your single EBSD in multiple layers and then follow the usual steps to get a 3d mesh.

2. You can use StatsGenerator (in your Dream3d distribution folder) to create a .dream3d file with the necessary statistics for successful synthetic reconstructions. If you have some information about the anticipated 3d shape of your grains (columnar and equiaxed it sounds like?) you can also look into applying the Saltykov and other stereology methods to transform your 2D grain size distributions into 3d ones. Then you just make sure your aspect ratios and shapes are correct in StatsGenerator and follow the usual synthetic microstructure steps.