Typically we check slot position using Bi-Lateral Position. We had a high and consistent number of rejected parts on our CMM, even though the parts mount onto their assemblies with more than adequate clearance. Our supplier uses angle and symmetry to describe the slot positions. I decided to compare the 2 methods.

Part: Disc with slots about the parts center register (Z axis, named datumB) in the plane (xy, named datumA) 

Bi-Lateral Method:

1) A cloud is taken from each side of the slot so that a slot can be created from the 2 clouds.

2) The part is aligned to the first slot created.

3) The cloud measurements and the slot creation are duplicated about datumB using the Transform function during programming.

-A best fit circle is created from the slots duplicated. When running the program the circle is created from the measurements of those slots.

4) The true position of the circle is reported to show how far the slot-pattern deviates from the register (datumB).

5) The bi-lateral position of each slot is taken with respect to datumA and the slot-pattern. This is because we are only concerned with the tangential position of the slot and not its radial position.

Symmetry Method:

1) A cloud is taken from each side of the slot so that planes can be created from the 2 clouds.

2) A plane is created between the slot.

3) This plane is duplicated via the transform method about the register to reflect the preferred position of the slots.

4) Step 1 is duplicated via the transform method about the register so that the planes are a result of the measurements taken. 

5) The sides of the slots are reported as symmetry to the planes created in step 2 & 3. 

I am trying to understand why I am seeing differences of up to .2mm between bi-lateral position and symmetry on a part with a symmetry call out of .10mm with the bi-lateral position being up to a diameter of .4mm.