Composite position, threaded hole pattern

Wouldn't all holes be located to datum B as they are controlled with basic dimensions? They way I see the dimensioning position is two circles, one with a .038 diameter and the other with .018 diameter centers are the same. Your holes are out of tolerance if they fall in the area between the .018 and .038 diameters for the composite but are still within tolerance for the .038. Both direction from datum A and datum B are still used to locate holes for the .018 and any tilt of the hole also takes away from your location tolerance. Is this not the way the dimensioning should be interpreted?

My dilemma is that just looking at the basic dimensions, I've added into the program (for analysis): distance along the z axis between one level of holes and the B datum (basic is 4.642"), distance from the first set of 4 holes to the next set (3.707" basic from the first level), angle between holes ( I constructed 2D lines from the axis of Datum A out to each hole, then dimensioned the angles between them). All this looks good, reasonable and held very close to the basic dimensions. But I'm befuddled as to how to use Position, and Alignment to agree with those numbers. When I used a Best Fit 2D alignment, using just the holes, and then dimensioning the position within that alignment @ .018 TP, 7 out of 8 holes fails, and badly. It makes no sense. So, I questioned the alignment I used,,,,first an alignment level to A, translated to B. Then within that, a 2D best fit alignment, least squares, Z axis, rotate & translate (this is the way I usually handle composite position call-outs.

But it just doesn't seem to work.

I don't think the part is bad, based on the analysis dimensioning I did. But I need to report it accurately, and I have about a dozen more of these to measure, so, I need to make the correct adjustments to the program. I'm stumped.

Running version 3.7MR3

Wouldn't all holes be located to datum B as they are controlled with basic dimensions? They way I see the dimensioning position is two circles, one with a .038 diameter and the other with .018 diameter centers are the same. Your holes are out of tolerance if they fall in the area between the .018 and .038 diameters for the composite but are still within tolerance for the .038. Both direction from datum A and datum B are still used to locate holes for the .018 and any tilt of the hole also takes away from your location tolerance. Is this not the way the dimensioning should be interpreted?

My dilemma is that just looking at the basic dimensions, I've added into the program (for analysis): distance along the z axis between one level of holes and the B datum (basic is 4.642"), distance from the first set of 4 holes to the next set (3.707" basic from the first level), angle between holes ( I constructed 2D lines from the axis of Datum A out to each hole, then dimensioned the angles between them). All this looks good, reasonable and held very close to the basic dimensions. But I'm befuddled as to how to use Position, and Alignment to agree with those numbers. When I used a Best Fit 2D alignment, using just the holes, and then dimensioning the position within that alignment @ .018 TP, 7 out of 8 holes fails, and badly. It makes no sense. So, I questioned the alignment I used,,,,first an alignment level to A, translated to B. Then within that, a 2D best fit alignment, least squares, Z axis, rotate & translate (this is the way I usually handle composite position call-outs.

But it just doesn't seem to work.

I don't think the part is bad, based on the analysis dimensioning I did. But I need to report it accurately, and I have about a dozen more of these to measure, so, I need to make the correct adjustments to the program. I'm stumped.

Running version 3.7MR3