I am with @
Kai Kendall on this one. It is probably a better idea to not use MMB from the datum features in general. I don't think he is saying your math is wrong but more that you need to be very careful when using this or you will get wrong results.
When you have datum shifts (which is a modification of the alignment system to suite one or more Geotol dimensions) you end up with two (or more) distinctly different alignment systems. The problem is you don't see the impact of this change in anything you cannot tie together using the Simultaneous Evaluation command. If you had a random vector point measured somewhere you would not see its value change following the datum shift for example. If you simply ask for the XY location of your datum features you won't see the change in location as a result of the datum shift.
A classic example is when you report a series of position dimensions where each shows no deviation in the result. You have the original alignment system of the part and one unique alignment for each separate position dimension.
There should only be one alignment system. If it is modified by a datum shift you should see this effect on everything. The Simultaneous Evaluation command in PC-DMIS tries to address this problem but it doesn't quite go far enough. Unless you are extremely careful your report will show results from the two or more alignment systems that exist.
I have used software where the impact of a datum shift suddenly changes everything on the report. At first I thought it was a bug but in retrospect I understand what they did and why.
I am with @
Kai Kendall on this one. It is probably a better idea to not use MMB from the datum features in general. I don't think he is saying your math is wrong but more that you need to be very careful when using this or you will get wrong results.
When you have datum shifts (which is a modification of the alignment system to suite one or more Geotol dimensions) you end up with two (or more) distinctly different alignment systems. The problem is you don't see the impact of this change in anything you cannot tie together using the Simultaneous Evaluation command. If you had a random vector point measured somewhere you would not see its value change following the datum shift for example. If you simply ask for the XY location of your datum features you won't see the change in location as a result of the datum shift.
A classic example is when you report a series of position dimensions where each shows no deviation in the result. You have the original alignment system of the part and one unique alignment for each separate position dimension.
There should only be one alignment system. If it is modified by a datum shift you should see this effect on everything. The Simultaneous Evaluation command in PC-DMIS tries to address this problem but it doesn't quite go far enough. Unless you are extremely careful your report will show results from the two or more alignment systems that exist.
I have used software where the impact of a datum shift suddenly changes everything on the report. At first I thought it was a bug but in retrospect I understand what they did and why.
I have actually forced correct simultaneous evaluation in Quindos (like where C gets a bonus for certain things, and both B and C on others), but it gets EXTREMELY complex, even with only 2 or 3 features of interest.
Am in the process of developing an improved (simpler) method in Quindos that should also work in PC-DMIS. Give me about 3 years.
