Leveling to a line (or cylinder axis)

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Leveling to a line (or cylinder axis)

    Hello,

    I know there have been posts covering this topic but I can`t find any good explanation of the logic behind this. How can you use a line (or cylinder/cylinder axis) to level? Doesn´t it take at least a line and a point to create a plane? I tried to do and and it goes as i would expect - the coordinate system trihedron tilts about the line/cylinder axis and the thus the leveling is incomplete.

    I think i´m missing something here or worse even, i have totally misunderstood the concept of leveling.

    Any explanation on subject is welcome.

    BR,
    PekkeriMan

  • #2
    Leveling Z to a plane locks two directions and one coordinate. The remaining DOF (degrees of freedom) are rotation around Z, and translation in XY.

    Leveling Z to a 3D line (cylinder axis, cone axis, 3D line sphere-sphere, etc) locks two directions and two coordinates. The remaining DOF are rotation around Z and translation in Z.

    Leveling to a 2D line is not recommended - as you need a correct work plane before measuring/constructing a 2D line, this work plane would be the primary candidate for level...
    AndersI
    SW support - Hexagon Metrology Nordic AB

    Comment


    • #3
      You can think of a cylinder axis or 3d line as the vector of a plane. The difference, as Anders mentioned, is what degrees of translation are constrained.
      "This is my word... and as such is beyond contestation."

      Comment


      • #4
        Okay, if I have a cylinder horizontally on table (axis approx parallel to Y). I measure two cylinders at both ends of the cylinder part and construct a 3D line between them. Then I create an alignment by leveling Y+ to the 3D line. Origin Z and X to the 3D line, rotate 3D line to Y plus, and finally origin Y to one end of the cylinder part. Now I only have one DOF unconstrained - rotation around Y.
        Is this correct?
        In this example if this last DOF remains unconstrained will PC-DMIS use machine XY plane to "stabilize" the coordinate system?

        Comment


        • #5
          No, you don't rotate to the 3D line (same vector direction). Don't you have a secondary datum?
          sigpicIt's corona time!
          737 Xcel Cad++ v2009MR1....SE HABLA ESPAÑOL

          Comment


          • PekkeriMan4000
            PekkeriMan4000 commented
            Editing a comment
            But i have to align cyl axis with Y axis..

        • #6
          Levelling uses only vectors. A cvector, by definition, is parallel to himself everywhere.
          In your example, if the part is totally axisymetric, you can measure the cylinder, level Y on it, measure the granite as a plane, rotate on it (X or Z), then origin Xand Z on the cylinder and measure a point or a plane to origin Y somewhere along the part.

          Comment


          • PekkeriMan4000
            PekkeriMan4000 commented
            Editing a comment
            So the "measure the granite as a plane, rotate on it (X or Z)," will constrain my last DOF?

        • #7
          Yes, if the part is totally axisymetric !
          if you have any flat surface, hole or.... which determines another orientation, then use it instead of the granite.

          Comment


          • PekkeriMan4000
            PekkeriMan4000 commented
            Editing a comment
            Okay as a last thing can you please just write step-by-step the forkflow of leveling Y on a cylinder.

        • #8
          Depending on the features that you need to measure and report. And there is nothing on the part to rotate to you may not need to constrain the last DOF.
          Some outputs do not need all 6 DOF constrained.
          Time for the Trolls to leave.

          Comment


          • #9
            It depends on the part...
            When I have to do it, I first measure a cylinder with a few hits, in the right direction.
            Then I create an alignment, and level on the cylinder, origin on the axis (often Z = level, even if the part is horizontally, XY origin)
            Then I measure the granite (X rotate), the start face of the cylinder, origin Z on it. (at this step, I have 3 alignments !).
            Then, I can DCC a circle close to each end, create a line between them in the right order, level Z on it and origin XY.
            If the part is totally axisymetric, I assume that the manual plane is enough for the rotation, I just re-measure the Z origin, and create the alignment (rotate to the granite, origin Z)

            Comment


            • PekkeriMan4000
              PekkeriMan4000 commented
              Editing a comment
              Okay, thank you very much!

          • #10
            You could also take a vector point on top of the front OD. Construct a line between the front circle and the point to rotate Z+.
            There's more than one way to skin the cat
            sigpicIt's corona time!
            737 Xcel Cad++ v2009MR1....SE HABLA ESPAÑOL

            Comment


            • #11
              Why spend running time on measuring a rotation constraint at all if it's arbitrary? Just construct a line from the MCS alignment (XAXIS, YAXIS or ZAXIS) and use that.
              AndersI
              SW support - Hexagon Metrology Nordic AB

              Comment

              widgetinstance 190 (Related Topics) skipped due to lack of content & hide_module_if_empty option.
              Working...
              X