hexagon logo
Register Sign in
  • Replies 5 replies
  • Subscribers 31 subscribers
  • Views 127 views
  • Users 0 members are here
Options
Related

Leapfrog and scanning

JFrohriep
Software: PC-DMIS Portable
Equipment: Absolute Arm Romer

We have an IP set up in the shop and we have to scan the whole part for exporting a cloud of points for surfacing. We can get most of the part but have to move the Romer arm to complete the scan. After doing the leapfrog alignment the scans do not match up, they are in completely different areas. We also did a 321 alignment before hand and that also does not match up with the scan. When we have a probe head on the arm, it matches up with the alignment on the screen so we know that is correct, but as soon as we take the probe off and switch it to scan, the scan head on the screen jumps to a new position. How can we leapfrog and continue scanning and have our scan match up to one another?
Parents
  • You come up with a workaround that sacrifices as little as possible. No software tool is perfect, in my experience.

    In this case, I know that because he is scanning the part for the purposes of surfacing, he shouldn't be using PCDMIS at all. The PCDMIS Laser pointcloud is XYZ only, and not XYZIJK, as true reverse engineering software would be. So the pointcloud that we're coming up with isn't going to be sufficient to RE the part anyway.

    Now...if we were, say...INSPECTING the part...

    I would align to a single set of features, scan what I could, and save the program.

    then i would move the Arm, re-align to the same set of features, Equate the old/new alignments, and continue scanning.

    I would continue this process until I had scanned everything I could.
Reply
  • You come up with a workaround that sacrifices as little as possible. No software tool is perfect, in my experience.

    In this case, I know that because he is scanning the part for the purposes of surfacing, he shouldn't be using PCDMIS at all. The PCDMIS Laser pointcloud is XYZ only, and not XYZIJK, as true reverse engineering software would be. So the pointcloud that we're coming up with isn't going to be sufficient to RE the part anyway.

    Now...if we were, say...INSPECTING the part...

    I would align to a single set of features, scan what I could, and save the program.

    then i would move the Arm, re-align to the same set of features, Equate the old/new alignments, and continue scanning.

    I would continue this process until I had scanned everything I could.
Children
No Data
© 2024 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement