hexagon logo

PC DMIS 2018 R2 free form alignment

I run a Romer 7535 arm, and have just upgraded to pc dmis 2018 R2 last week. Typically we use a free form alignment (best fit) for the contoured parts we create. The first few inspections with the new software I had no issues. It seems now though I can't get a proper lock in. I'm picking points on datum surfaces and going through the normal steps, but now when I try the free form alignment it doesn't seem to compensate surfaces. For example, it's telling me the right end of the part is .070" long, and the left end is .070" short, instead of compensating for the difference. Typically when I best fit to some points on a couple datums, it will give me next to nominal on those surfaces, then I can see the error in other features relating back to the datums

I've tried using lots of different points. I've tried resetting the probe files. I've also gone back to our old PC DMIS version and had no issues. When I come back to 2018 R2 however, it still doesn't seem to work. Any thoughts?
Parents
  • No print for this one, it's to the Catia model. We do a lot of composite layups and large contoured parts so I'm used to using the best fit method for locking onto parts in A/C coordinates. I have ruled out the surfaces being offset. I called up the same program in PC DMIS 2016 and got locked on 1st try, showing all surfaces well within tolerance. So I proven to myself that it's something to do with the 2018 R2. Must be some setting that's changed somewhere or something.

    For my alignment I'm basically having to pick surfaces that are vertical enough to establish where this sits in X,Y and Z. Typically if a part is long by say .020" and I use both ends for lock in, it will compensate and give me a +.010" on each end. By using only one side, it would read 0 on the lock in side, and +.020 on the other side. However no matter which surfaces I seem to pick, I get random amounts from .030" to .400" away from nominal.

    again using the best fit method in the free form alignment
Reply
  • No print for this one, it's to the Catia model. We do a lot of composite layups and large contoured parts so I'm used to using the best fit method for locking onto parts in A/C coordinates. I have ruled out the surfaces being offset. I called up the same program in PC DMIS 2016 and got locked on 1st try, showing all surfaces well within tolerance. So I proven to myself that it's something to do with the 2018 R2. Must be some setting that's changed somewhere or something.

    For my alignment I'm basically having to pick surfaces that are vertical enough to establish where this sits in X,Y and Z. Typically if a part is long by say .020" and I use both ends for lock in, it will compensate and give me a +.010" on each end. By using only one side, it would read 0 on the lock in side, and +.020 on the other side. However no matter which surfaces I seem to pick, I get random amounts from .030" to .400" away from nominal.

    again using the best fit method in the free form alignment
Children
No Data