hexagon logo

True position bad math

Either I'm having a huge brainfart or PC-DMIS has a new bug after this last update (2017 R2 SP9). Look at the actual hole locations and then look at the positions. I don't know what kind of common core math it's doing to get those numbers.

The only factor that I can think of, which shouldn't be a factor at all, is that each cylinder is constructed out of two circles. The Z nominal of the bottom circle is at the bottom of the part with a vector of 0,0,-1, as I intended to get a full length cylinder. As it turned out, I was unable to reach the bottom of the part and instead took it from the top at a much shallower depth with a vector of 0,0,1. I don't see how this would affect anything but it's the only thing I can think of.




Parents
  • Can you report the Z axis too. I think that the xact measure will use all applicable axis to calculate the position regardless of reported axis. It looks like z is off in CYL_24_11. This is a long shot by the way.


    It shouldn't take the third axis into account, unless the feature isn't normal to the alignment (ie. 45 degree hole).
  • Try punching out a legacy dimension for one of the "bad" cylinders and create a few different ones with "Use worst", "use best", "use start", "use end", whatever, and see if you get values that are similar to the Xact dimensions for any of them.

    I have a feeling that my earlier post is the problem.
Reply Children
No Data