Crazy Diameter (math error)

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

  • Crazy Diameter (math error)

    Edit: reducing the hits to 3 INCREASED the accuracy of my measurements. Special thanks to UKCMM!




    So I was measuring a cylinder and I got some pretty crazy results (See the diameter). Trying to figure out why this happened beyond some good ol' fashion PC DMIS math magic: none of the hits seem to be varying a crazy amount (at least not WHOLE INCH level of crazy) so I was wondering if any of the experts out there had some insight into how this could have happened so I can avoid it in the future without having to just remeasure the feature until it works.


    Code:

    LEFT_CUTOUT =FEAT/CONTACT/CYLINDER/DEFAULT,CARTESIAN,IN,LEAST_SQR
    THEO/<-3.6115,0.601,3.8812>,<0.943801,0,0.3305144>,0.2,-0.125
    ACTL/<-3.4513,20.3725,4.1557>,<0.9510529,-0.0114305,0.3088165>,39.6259,-0.125
    TARG/<-3.6115,0.601,3.8812>,<0.943801,0,0.3305144>
    START ANG=0,END ANG=-180
    ANGLE VEC=<-0.3305144,0,0.943801>
    DIRECTION=CW
    SHOW FEATURE PARAMETERS=YES
    VOID DETECTION=NO
    REMEASURE=NO,USE THEO=NO
    SURFACE=THICKNESS_NONE,0
    MEASURE MODE=NOMINALS
    RMEAS=LEFT_NOTCH,LEFT_EDGE,LEFT_NOTCH
    AUTO WRIST=NO
    CIRCULAR MOVES=CIRCULAR
    GRAPHICAL ANALYSIS=NO
    FEATURE LOCATOR=NO,NO,""
    SHOW CONTACT PARAMETERS=YES
    NUMHITS=8,NUMLEVELS=4,DEPTH=0.02,END OFFSET=0.02,PITCH=0
    SAMPLE METHOD=SAMPLE_HITS
    SAMPLE HITS=0,SPACER=0
    AVOIDANCE MOVE=NO,DISTANCE=0.3937
    FIND HOLE=DISABLED,ONERROR=NO,READ POS=NO
    SHOW HITS=YES
    HIT/BASIC,<-3.6257,0.6045,3.9821>,<0.330313,-0.0348995,-0.943226>,<-3.615,0.5995,3.9847>
    HIT/BASIC,<-3.622,0.6466,3.9718>,<0.2941154,-0.4562107,-0.8398619>,<-3.6114,0.6426,3.9746>
    HIT/BASIC,<-3.6128,0.6801,3.9455>,<0.2021837,-0.791071,-0.577346>,<-3.6024,0.6772,3.9482>
    HIT/BASIC,<-3.5998,0.6986,3.9084>,<0.0719386,-0.9760254,-0.2054245>,<-3.5888,0.6971,3.9098>
    HIT/BASIC,<-3.5854,0.6986,3.8673>,<-0.0719386,-0.9760254,0.2054245>,<-3.5738,0.6981,3.8662>
    HIT/BASIC,<-3.5724,0.6801,3.8301>,<-0.2021837,-0.791071,0.577346>,<-3.5599,0.6782,3.828>
    HIT/BASIC,<-3.5632,0.6466,3.8038>,<-0.2941154,-0.4562107,0.8398619>,<-3.5507,0.6434,3.8012>
    HIT/BASIC,<-3.5596,0.6045,3.7935>,<-0.330313,-0.0348995,0.943226>,<-3.5473,0.5999,3.7912>
    HIT/BASIC,<-3.5329,0.6045,3.8029>,<-0.330313,-0.0348995,0.943226>,<-3.5205,0.5999,3.801>
    HIT/BASIC,<-3.5365,0.6466,3.8132>,<-0.2941154,-0.4562107,0.8398619>,<-3.524,0.6432,3.811>
    HIT/BASIC,<-3.5457,0.6801,3.8394>,<-0.2021837,-0.791071,0.577346>,<-3.5334,0.678,3.8378>
    HIT/BASIC,<-3.5587,0.6986,3.8766>,<-0.0719386,-0.9760254,0.2054245>,<-3.547,0.6975,3.8759>
    HIT/BASIC,<-3.5731,0.6986,3.9177>,<0.0719386,-0.9760254,-0.2054245>,<-3.5619,0.6968,3.919>
    HIT/BASIC,<-3.5861,0.6801,3.9549>,<0.2021837,-0.791071,-0.577346>,<-3.5756,0.6768,3.9571>
    HIT/BASIC,<-3.5953,0.6466,3.9812>,<0.2941154,-0.4562107,-0.8398619>,<-3.5844,0.6421,3.9833>
    HIT/BASIC,<-3.5989,0.6045,3.9915>,<0.330313,-0.0348995,-0.943226>,<-3.588,0.5995,3.9936>
    HIT/BASIC,<-3.5722,0.6045,4.0009>,<0.330313,-0.0348995,-0.943226>,<-3.5608,0.5994,4.0016>
    HIT/BASIC,<-3.5686,0.6466,3.9905>,<0.2941154,-0.4562107,-0.8398619>,<-3.5574,0.6417,3.9917>
    HIT/BASIC,<-3.5594,0.6801,3.9643>,<0.2021837,-0.791071,-0.577346>,<-3.5485,0.6761,3.9658>
    HIT/BASIC,<-3.5463,0.6986,3.9271>,<0.0719386,-0.9760254,-0.2054245>,<-3.5355,0.6962,3.9285>
    HIT/BASIC,<-3.5319,0.6986,3.886>,<-0.0719386,-0.9760254,0.2054245>,<-3.5203,0.6964,3.8851>
    HIT/BASIC,<-3.5189,0.6801,3.8488>,<-0.2021837,-0.791071,0.577346>,<-3.5067,0.6772,3.8475>
    HIT/BASIC,<-3.5097,0.6466,3.8226>,<-0.2941154,-0.4562107,0.8398619>,<-3.4975,0.643,3.821>
    HIT/BASIC,<-3.5061,0.6045,3.8122>,<-0.330313,-0.0348995,0.943226>,<-3.4937,0.5997,3.8103>
    HIT/BASIC,<-3.4794,0.6045,3.8216>,<-0.330313,-0.0348995,0.943226>,<-3.4683,0.5997,3.8234>
    HIT/BASIC,<-3.483,0.6466,3.8319>,<-0.2941154,-0.4562107,0.8398619>,<-3.4716,0.6416,3.8327>
    HIT/BASIC,<-3.4922,0.6801,3.8582>,<-0.2021837,-0.791071,0.577346>,<-3.4805,0.6749,3.8589>
    HIT/BASIC,<-3.5052,0.6986,3.8954>,<-0.0719386,-0.9760254,0.2054245>,<-3.4936,0.6943,3.8956>
    HIT/BASIC,<-3.5196,0.6986,3.9364>,<0.0719386,-0.9760254,-0.2054245>,<-3.5083,0.6942,3.9372>
    HIT/BASIC,<-3.5326,0.6801,3.9736>,<0.2021837,-0.791071,-0.577346>,<-3.5215,0.6746,3.9743>
    HIT/BASIC,<-3.5418,0.6466,3.9999>,<0.2941154,-0.4562107,-0.8398619>,<-3.5299,0.6405,3.999>
    HIT/BASIC,<-3.5454,0.6045,4.0102>,<0.330313,-0.0348995,-0.943226>,<-3.5338,0.5994,4.0107>
    ENDMEAS/
    Attached Files
    Last edited by derekvegeta; 01-09-2020, 09:11 AM.

  • #2
    Workplane correct?

    Comment


    • derekvegeta
      derekvegeta commented
      Editing a comment
      It is, though I didn't think that workplane mattered for auto cylinders?

  • #3
    It could be that as the length of the measured cylinder is very short ( 0.080" ) compared to diameter, you are only measuring 180 degree of diameter and it is at an angle then the software maybe having problems resolving the hits.

    With this small feature I have found that less hits can yield better sometimes results.

    You could try reducing the number of hits to 3 and levels to 3, if you cam increase the length of the measured cylinder to the max possible. Also are the cylinder vectors correct they just seem a little odd and is the probe at the right angle to prevent shank out.

    lastly Happy new year

    Comment


    • derekvegeta
      derekvegeta commented
      Editing a comment
      Yeah its a cutout on an angled surface, so the vectors are correct for that surface. The length is .125" with .02" offsets on both ends, is that the .08" length you're referring to? The material thickness is only .125" unfortunately but I will definitely try reducing the hits

    • UKCMM
      UKCMM commented
      Editing a comment
      Yes the 0.080" is the length minus the offsets.

  • #4
    Is it possible the probe is partly missing the cutout or hitting a burr/chamfer its just that I see the start/end angles are 0/180 degree, is that taken from a cad model if so you may have to adjust the angle to allow for burrs/chamfer and position error.

    Comment


    • derekvegeta
      derekvegeta commented
      Editing a comment
      The hits looked to be hitting the feature correctly, but changing it to 3 hits per level per your recommendation made a huge difference, so thanks!

Related Topics

Collapse

Working...
X