2D Microstructure varying grain size when checking z dimensions or resolution.

[William Noh]

Hello,

I had a question about initializing the synthetic volume. I am a bit confused about the relationship between the dimension/resolution and the grain sizes generated from the stats generator. I understand that the dimension is the number of elements(voxels) you set for the volume: e.g. 348x348x348 would be 348^3 elements. The resolution sets the length units for the voxels: eg. res 1x1x1 = 348x348x348 lengths units and 0.5x0.5x1 = 174x174x348 length units. 

My objective is to create a "2D" microstructure with the following grain statistics:

I initially created a 348x348x1 dimensioned volume with resolution 1x1x1, but this create a microstructure that is not reflective of the grain statistics, not even a mean grain size of 30(um): 

This is obviously not correct. I suspect that the thinness is interfering with the 3D nature of the algorithm. Does anyone have any advice to correct this while maintaining the 1 voxel thickness? Technically you could create 348x348x348 volume and take slices, but I am looking for alternative ways while maintaining 1 voxel thickness.

My main priority is to stay true to the statistics I set along the flat plane, so if 1 voxel thickness is impossible, please let me know!

One idea I did have is to change the resolution to 1x1x348, but does this maintain the statistics I set? See below:

Thank you for your time!
WN