hexagon logo

Body axis scan question

Hello everyone, I’m using body axis scan to create 3D lines then intersect the lines. One line horizontal and the other roughly vertical.
My concern is the lines are not where I expect they would be.
For example I set a body axis scan for Z-6.000, I carefully scan 2 hits roughly 2 inches apart and create the 3D line. When I analyze the line it’s Z-6.017. Shouldn’t the Z-6.000 be a fixed dimension, am i doing something wrong?
Parents
  • Manually with an arm? You'll never be exact.


    This isn't necessarily true. PCDMIS creates the body axis scan hits by:

    - setting body axis plane
    - capturing scan
    - identifying last hit prior to plane crossing and first hit afterwards
    - drawing a line between the two hits
    - calculating the body axis scan hit as the pierce point between the line and the body axis plane.

    by that logic (and by my experience), it should be exact at Z= -6!! Are you using a CAD model? If so, you should have FINDNOMs turned on, with an appropriate tolerance. The presence of the CAD model will affect how your probe compensation takes place (towards the CAD!).

    Lacking a CAD model, things get trickier. PCDMIS wants to comp your hits, so it will do so in the closest Axis you have the arm pointed in. This is usually bad for me, as I'm scanning a 2D complex surface, and my IJK isn't as simple as <-1,0,0> in this case. In that situation, I tend to turn off Probe Compensation (under F10 > Probing), and capture my scan hits from the center of the ball. Unfortunately, then you are off by the ruby's radius, but you can always construct features (offset by the ruby radius) to represent the actual surface.

    Long story short, doing any sort of scan without a model is tough with a Romer in PCDMIS, mostly due to probe compensation. Try scanning with probecomp off and constructing your features from there. I suspect that's the source of your issue.
Reply
  • Manually with an arm? You'll never be exact.


    This isn't necessarily true. PCDMIS creates the body axis scan hits by:

    - setting body axis plane
    - capturing scan
    - identifying last hit prior to plane crossing and first hit afterwards
    - drawing a line between the two hits
    - calculating the body axis scan hit as the pierce point between the line and the body axis plane.

    by that logic (and by my experience), it should be exact at Z= -6!! Are you using a CAD model? If so, you should have FINDNOMs turned on, with an appropriate tolerance. The presence of the CAD model will affect how your probe compensation takes place (towards the CAD!).

    Lacking a CAD model, things get trickier. PCDMIS wants to comp your hits, so it will do so in the closest Axis you have the arm pointed in. This is usually bad for me, as I'm scanning a 2D complex surface, and my IJK isn't as simple as <-1,0,0> in this case. In that situation, I tend to turn off Probe Compensation (under F10 > Probing), and capture my scan hits from the center of the ball. Unfortunately, then you are off by the ruby's radius, but you can always construct features (offset by the ruby radius) to represent the actual surface.

    Long story short, doing any sort of scan without a model is tough with a Romer in PCDMIS, mostly due to probe compensation. Try scanning with probecomp off and constructing your features from there. I suspect that's the source of your issue.
Children
No Data