The thickness tool helps highlight areas that are too thin. 

NOTE: The concept of “thickness” is not as well-defined as we might like. This tool chooses just one of many approaches to computing it. We estimate thickness by shooting a number of random rays from each vertex into the “interior” of the surface, and measuring the distance to the closest ray intersections. The average of these distances is our measure of thickness. 

Cone Samples specifies how many rays are fired from each vertex, and Cone Angle specifies the distribution of rays. We ignore ray intersections where the ray direction is nearly perpendicular to the normal of the triangle it hits. Grazing Angle is the threshold angle used to cull these intersections

We refer to regions that are too thin as “defects”. Min Thickness is the tolerance thickness for a region to be labeled as a defect. Vertices below the min thickness are rendered in red.

 

Regions must be larger than Min Defect Size to be labeled as defects. This can be used to filter out small-scale bumps that might be erroneously identified as defects.

We also use rendered spheres to highlight the thin regions, as it can sometimes be difficult to see every red region in a complex model. Clicking a sphere will select the triangles in the defect region.