hexagon logo
Register Sign in
  • Replies 4 replies
  • Subscribers 221 subscribers
  • Views 58 views
  • Users 0 members are here
Options
Related

BFRE construction and probes used

dwade
When using BFRE on a constructed feature I need to be sure that it is done while the probe that was used for measuring is loaded.
This is correct, yes?

I also am believing that one cannot use BFRE when more than one probe is used to measure features used in a construction such as individual points measured for construction of a feature.
This is correct, yes?​
Parents
  • What kind of feature do you want to construct ?
    I believe you can check the construction using RAWHIT and PROBEDATA("diam").
    For a circle, for example, you could create a feature set with points measured with diameter 1 and another with points measured with diameter 2 (and so on).
    Construct a BFRE circle CIRC1 from points, and then :
    ASSIGN/V1=UNIT(SCN1.RAWHIT[1..SCN1.NUMHITS].XYZ-CIRC1.XYZ)
    ASSIGN/V2=UNIT(SCN2.RAWHIT[1..SCN2.NUMHITS].XYZ-CIRC1.XYZ)
    ASSIGN/V3=DOT(V1,SCN1.HIT[1..SCN1.NUMHITS].IJK)
    ASSIGN/V4=DOT(V2,SCN2.HIT[1..SCN2.NUMHITS].IJK)
    ​If V3 and V4 are array of ones, then the demon didn't change the vector, so I think the compensation is wrong (or the hits are exactly perp to the surface - in this case BF and BFRE should be exactly the same)

    If the dot product of 2 unit vectors is 1, then the cos of the angle between vectors is also 1, so angle is 0)
Reply
  • What kind of feature do you want to construct ?
    I believe you can check the construction using RAWHIT and PROBEDATA("diam").
    For a circle, for example, you could create a feature set with points measured with diameter 1 and another with points measured with diameter 2 (and so on).
    Construct a BFRE circle CIRC1 from points, and then :
    ASSIGN/V1=UNIT(SCN1.RAWHIT[1..SCN1.NUMHITS].XYZ-CIRC1.XYZ)
    ASSIGN/V2=UNIT(SCN2.RAWHIT[1..SCN2.NUMHITS].XYZ-CIRC1.XYZ)
    ASSIGN/V3=DOT(V1,SCN1.HIT[1..SCN1.NUMHITS].IJK)
    ASSIGN/V4=DOT(V2,SCN2.HIT[1..SCN2.NUMHITS].IJK)
    ​If V3 and V4 are array of ones, then the demon didn't change the vector, so I think the compensation is wrong (or the hits are exactly perp to the surface - in this case BF and BFRE should be exactly the same)

    If the dot product of 2 unit vectors is 1, then the cos of the angle between vectors is also 1, so angle is 0)
Children
  • in reply to JEFMAN
    Toying with changing plane and cylinder measurement and constructions. A family of implant trials that are a PITA to measure with our standard probe configurations as we have 17 CMMs and need to keep them standardized in as many ways as possible as management doesn't want "special needs probes" on just one CMM
© 2024 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement