Perfect profile scans?

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

  • Perfect profile scans?

    Good Morning Everyone! Happy Friday! Quick question: I consistently use Linear Open scans for surface profiles, so I'm no newbie. I generate from CAD, and I have always seen some kind of deviation in the max/min of the report. I created scans on a part yesterday, and when the report was generated the max/min was 0/0. Is this even possible? I've NEVER seen that. I zoomed in on the graph, and you can see the surface roughness (not just a perfect curve), and I exported the XYZ and the points are there. Has anyone ever seen this?

    Thank you, Jackarroo
    Leitz Reference XI 10.7.6 with HP-S-XIH
    2017 R2 with SP3

  • #2
    If you have MASTER selected, and then execute the scan on an actual part, then it makes that first scan MASTER, and thus zero deviations. Needs to be set to FIND NOMS and RELEARN
    sigpic
    Originally posted by AndersI
    I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

    Comment


    • #3
      I have verified 3 times I have FIND NOMS, RELEARN, VECTOR selected.
      Leitz Reference XI 10.7.6 with HP-S-XIH
      2017 R2 with SP3

      Comment


      • KIRBSTER269
        KIRBSTER269 commented
        Editing a comment
        Defined, find nominals, vector

    • #4
      I also have CAD COMP, Scan only selected
      Leitz Reference XI 10.7.6 with HP-S-XIH
      2017 R2 with SP3

      Comment


      • #5
        I also have PROBE COMP. Sorry my messages are broken, had to go look and verify.
        Leitz Reference XI 10.7.6 with HP-S-XIH
        2017 R2 with SP3

        Comment


        • #6
          Can you check the cad ?
          Or try ASSIGN/V1=DOT(SCN1.HIT[1..SCN1.NUMHITS].XYZ-SCN1.HIT[1..SCN1.NUMHITS].TXYZ,SCN1.HIT[1..SCN1.NUMHITS].TIJK)
          And look at MAX(V1) and MIN(V1)

          Comment


          • #7
            Jefman, I don't know how to do this. I haven't done anything code wise, I'm trying to learn it. I don't even know what this means.
            Leitz Reference XI 10.7.6 with HP-S-XIH
            2017 R2 with SP3

            Comment


            • #8
              It's just the code to calculate the T value of each hit.
              You can do it just by assigning some values (type them in the edit window) :
              ASSIGN/V1=SCN1.HIT[1..SCN1.NUMHITS].XYZ....................................array of coordinates of actual hits
              ASSIGN/V2=SCN1.HIT[1..SCN1.NUMHITS].TXYZ..................................array of coordinates of theo hits
              ​​​​​​​​​​​​​​ASSIGN/V3=SCN1.HIT[1..SCN1.NUMHITS].TIJK..................................... array of theo vectors
              ASSIGN/V4=DOT(V1-V2,V3)............................................ ........................dot product of actual - theo along theo vector => definition of t value.

              If you don't write usually assignments, go in the edit window, click after the scan, then enter (go to next line !)
              type AS then tab twice (it will write ASSIGN/ then V1= and then type the lines above.
              If something appears red, then there's something wrong.
              Maybe the scan is not SCN1, then replace this by the right name...
              ​​​​​​​Hope this help !

              Comment


              • #9
                I'm going to comment on this, because I do Scans all the time. Now I have an OFFLINE Seat and ONLINE seat, Both running Same version. Only difference is OFFLINE has Visual, anyway was running a part on machine, in the mean time I was programming on my OFFLINE seat. Doing some scans, so when I transferred the program over and ran it on the actual part, same as your problem, perfect readings, so I rewrote the program on the online and same CAD model nothing different, and there it was profile was extremely bad as we expected. Now the only thing I did see different from my ONLINE seat and my OFFLINE seat is "Only Selected" box above the "Use Best Fit" box is Greyed out. I cannot find any differences between the 2 but that. Other than 2 different computers.
                (In Memory of my Loving wife, "Ronda" who I lost March 7, 2016. I love you baby.)
                They say "Nobody's Perfect." I must be Nobody.

                Comment


                • #10
                  OK, guys, I tried DEFINED, FIND NOMS, VECTOR (Thanks Kirbster). No results. If I uncheck CAD COMP, everything work fine, I get results. What does CAD COMP do actually? The previous programmer never used it, but I was trained to use it... What is it? I tried combinations of other things, and they didn't work either. Only disabling CAD COMP.
                  Leitz Reference XI 10.7.6 with HP-S-XIH
                  2017 R2 with SP3

                  Comment


                  • NinjaBadger
                    NinjaBadger commented
                    Editing a comment
                    It's been a while but AFAIK (and I'm happy to be corrected) when you do a normal scan the probe hits are compensated (adjusted from stylus centre to where it contacts the part) in the cut plane.

                    So if you scanned a circle on a conical surface (cone in the Z direction) you will get cosine error (which is why we can't measure a gauge diameter like this).

                    If you use CAD comp the scan is compensated by the CAD surface vector. (Still can't do a gauge diameter because you'll wander off the correct height depending if there's material on/off the part).

                • #11
                  Kirbster, I found your post on 2D/3D with CAD COMP. Can I assume since CAD COMP won't work, there is a model issue here?
                  Leitz Reference XI 10.7.6 with HP-S-XIH
                  2017 R2 with SP3

                  Comment


                  • KIRBSTER269
                    KIRBSTER269 commented
                    Editing a comment
                    I believe the CAD comp uses it to look at each point in 3D where the other looks at it like 2D. I could be wrong, but I never use this. Because I'm looking at my scan like a T-value, high or low

                • #12
                  Thank you Jefman, I will give it a shot.
                  Leitz Reference XI 10.7.6 with HP-S-XIH
                  2017 R2 with SP3

                  Comment


                  • #13
                    Kirbster, that's basically all I'm looking for as well, high or low from surface profile, but with it off won't there be deviation from the tip cosine error? Edit: I'm looking for location also.
                    Last edited by Jackarroo; 03-26-2021, 12:05 PM.
                    Leitz Reference XI 10.7.6 with HP-S-XIH
                    2017 R2 with SP3

                    Comment


                    • KIRBSTER269
                      KIRBSTER269 commented
                      Editing a comment
                      Location is based off of the Datums you call it out, or your alignment if using legacy. also when you check Best Fit, it will eliminate any points it couldn't find nominals for. If your profile is bad it will show more than one place, You'll see it, That's why I always do Analysis View, So I can see what it's doing. If you're sweeping more than one line I get it. hence the 3D a Plane versus a Line , You just said Linear Open scan, I'm just guessing your doing a line?

                  • #14
                    Sorry for the delay. I'm using open linear scans on 1/2 of a precision ball.
                    Leitz Reference XI 10.7.6 with HP-S-XIH
                    2017 R2 with SP3

                    Comment


                    • #15
                      From bottom to top, 8 locations around the ball.
                      Leitz Reference XI 10.7.6 with HP-S-XIH
                      2017 R2 with SP3

                      Comment

                      Related Topics

                      Collapse

                      Working...
                      X