Composite tolerance - true position evaluation

[David Prochazka]

Hi fellow engineers,

I am a newbie to GD&T and I have a problem with evaluation of composite tolerance of true position for pattern of features.

Let's say i have pattern of 4 holes with a composite tolerance:

4x ØD ± DTOL
⌖ ØPOS1 | A | B | C
ØPOS2 | A

I know how composite tolerance works - the first segment as a regular position to the selected DRF A|B|C. The second segment as evaluation of position & orientation between the features and orientation to the selected DRF (in my case perpendicular plane to cylindrical holes).

I have a little problem with comprehension of the evaluation of this tolerance though. Should the output be 4 separate values for the first segment (each hole separately, as would the tolerance been specified for each hole separately) and then 4 values for the second segment? That's how the PC-DMIS returns the results (see attach). I really don't understand how this evaluation works - is there a first step of setting the tolerance area for the first segment for the DRF and then somehow best-fitting the actual feature set inside the tolerance field to check the position between the features for the second segment? That's why there are different results for the second segment for each feature - each one is differently positioned to the tolerance fields in the best-fit alignment? And shouldn't be there just one value for the first segment and one for second (the worst one) as it is just one geometrical dimension for a set of features?

Thanks for your patience and reply.

 

[Sem_D220]

The first segment of the composite control creates 4 tolerance zones in fixed locations relative to the datums. When you verify the holes to the first segment requirement you check how far each hole axis is from its true position and whether it's within the tolerance zone. That's why 4 sets of results are reported. At this stage, you know the axis locations. The second segment creates a pattern of tolerance zone cylinders at a fixed spacing, and this tolerance zones pattern is movable. You can best fit it over the actual hole axes until as many of them as possible are entirely within the tolerance cylinders. Then, you have again four sets of results reporting how far each axis is from the center of its respective second-segment tolerance zone cylinder.