How are “Mean distance” and “Distance standard deviation” weighted in ZEISS INSPECT surface comparison?

[[Já...]]

I am evaluating internal gap using a surface comparison in ZEISS INSPECT/GOM.

For a virtually fitted implant component (local best fit with tolerance/outside target element), practically all valid deviations are positive because penetration is prevented by the alignment. Therefore, I expected the software-reported signed mean distance to be approximately equal to:

Integrated signed distance / Area of valid distance

and also approximately equal to:

Integrated absolute distance / Area of valid distance

However, in one example I obtained:

• Mean distance: 0.0858 mm
• Integrated distance: 31.1040 mm³
• Integrated absolute distance: 31.1040 mm³
• Area of valid distance: 175.5282 mm²

Thus:

31.1040 / 175.5282 = 0.1772 mm

which is about twice the reported mean distance, despite the absence of relevant negative deviations.

Could ZEISS please clarify the exact calculation of the following surface-comparison statistics?

1. Is Mean distance calculated as an unweighted average of geometry inspection values, mesh vertices, CAD vertices, or another set of evaluation points?
2. Are individual values weighted by the associated triangle or surface area?
3. How is Distance standard deviation calculated, and does it use the same weighting as Mean distance?
4. Is avg(inspection.value, index='geometry') mathematically identical to the Mean distance shown in the properties?
5. Is there a built-in method to calculate a genuinely surface-area-weighted mean and standard deviation?
6. Does the tessellation density or triangle-size distribution of the nominal CAD influence Mean distance and standard deviation?

My aim is to report the physical mean internal gap over the complete valid surface. At present, I calculate this as integrated distance divided by valid area, but I would like to understand precisely how this differs from the built-in statistics.