True position error on an inclinated bores pattern

[[Ru...]]

ood morning everyone,

I am currently working on a spherical part featuring several inclined holes, and I am looking for the best way to measure their position error.

My base alignment is set as follows:

• Origin: Centered on the central hole.
• Rotation (Y-axis): Oriented using one of the outer holes.
• Z-axis: Levelled to the external plane of the sphere.

For both the central and the orientation holes, I measure them as cylinders and then calculate the intersection point between the cylinder axis and the sphere surface to define the alignment features.

All subsequent holes are also measured as cylinders. As mentioned above, their true position is evaluated at the intersection point between each hole's axis and the sphere surface.
I only measured the holes located along the Y-axis line, and then I created radial patterns for the rest.

The problem I am encountering is that as I move further away from the hole used for the Y-axis orientation, the position error increases exponentially (as shown in the attached graph).

I have already tried using the intersection with the plane instead of the sphere surface, but the results remain unchanged.

Does anyone have suggestions on how to achieve more accurate and reliable values? Is there a better way to handle the alignment or the projection of these inclined axes?"