additional metric: hex characteristic side length (CSL?)

[hovey]

@mrbuche

Context

Voxelated Meshes

For direct voxel to hex conversion, the side length is easily described:

Conforming Meshes

For the conforming meshes, there is a fuzzy element size parameter that is set at time of mesh generation.

Wish List

It would be nice to have a more apples-to-apples comparison of element characteristic side length, comparing

• the voxel mesh (a Dirac delta type of histogram with a single side length and height equal to the number of elements in the mesh) to
• the element characteristic of a conforming mesh (a histogram with some dispersion around a mean).

Can we envision a side length parameter for a general hex that takes the average of all the 12 side lengths? Would this be useful (I think so, but would like to know what you think).

I think then I could say more clearly what the element length of the conforming meshes were and how they compared on a convergence plot (with element size length as the independent variable).

[mrbuche]

that sounds great to me! i bet the side lengths within a hex dont vary a ton either, and we can probably show that too, so average side length would be a good measure

[hovey]

@mrbuche I will implement this one if that suits you?