hexagon logo

How to construct circles with different math methods from hit points data?

Happy Friday! Can not wait to watch my boy play Hockey game tomorrow afternoon. Slight smile
From other threads, I knew we can construct features from points data with different math methods.
The core manual is perfect, but I could not help try right away.

I had a circle C1 now, since I could not construct circle with F9, so constructed a circle with other points and replaced with hit data, but it did not work well.
Can we type those construct command ? And how to work on with hit data for those?

Thank you for any help





Attached Files
  • Better off using a closed scan then using those points to construct your various circles.


    SC_B1 =FEAT/SCAN,LINEARCLOSE,NUMBER OF HITS=103,SHOW HITS=NO,SHOWALLPARAMS=NO
    MEAS/SCAN
    BASICSCAN/LINE,NUMBER OF HITS=103,SHOW HITS=NO,SHOWALLPARAMS=NO
    ENDSCAN
    ENDMEAS/
    CIRCLE1 =FEAT/CIRCLE,CARTESIAN,OUT,LEAST_SQR,NO
    THEO/<-0.0005,0,-8>,<0,0,1>,8.235
    ACTL/<-0.0005,0,458.5778>,<0,0,1>,8.235
    CONSTR/CIRCLE,BF,2D,SC_B1.HIT[1..SC_B1.NUMHITS],,
    OUTLIER_REMOVAL/OFF,3
    FILTER/OFF,UPR=0
    CIRCLE1 =FEAT/CIRCLE,CARTESIAN,OUT,MAX_INSC,NO
    THEO/<-0.0005,0,-8>,<0,0,1>,8.235
    ACTL/<-0.0005,0,458.5778>,<0,0,1>,8.235
    CONSTR/CIRCLE,BF,2D,SC_B1.HIT[1..SC_B1.NUMHITS],,
    OUTLIER_REMOVAL/OFF,3
    FILTER/OFF,UPR=0
  • Better off using a closed scan then using those points to construct your various circles.


    SC_B1 =FEAT/SCAN,LINEARCLOSE,NUMBER OF HITS=103,SHOW HITS=NO,SHOWALLPARAMS=NO
    MEAS/SCAN
    BASICSCAN/LINE,NUMBER OF HITS=103,SHOW HITS=NO,SHOWALLPARAMS=NO
    ENDSCAN
    ENDMEAS/
    CIRCLE1 =FEAT/CIRCLE,CARTESIAN,OUT,LEAST_SQR,NO
    THEO/<-0.0005,0,-8>,<0,0,1>,8.235
    ACTL/<-0.0005,0,458.5778>,<0,0,1>,8.235
    CONSTR/CIRCLE,BF,2D,SC_B1.HIT[1..SC_B1.NUMHITS],,
    OUTLIER_REMOVAL/OFF,3
    FILTER/OFF,UPR=0
    CIRCLE1 =FEAT/CIRCLE,CARTESIAN,OUT,MAX_INSC,NO
    THEO/<-0.0005,0,-8>,<0,0,1>,8.235
    ACTL/<-0.0005,0,458.5778>,<0,0,1>,8.235
    CONSTR/CIRCLE,BF,2D,SC_B1.HIT[1..SC_B1.NUMHITS],,
    OUTLIER_REMOVAL/OFF,3
    FILTER/OFF,UPR=0


    Thank you so much.

    But How can we input those? :

    CIRCLE1 =FEAT/CIRCLE,CARTESIAN,OUT,LEAST_SQR,NO
    THEO/<-0.0005,0,-8>,<0,0,1>,8.235
    ACTL/<-0.0005,0,458.5778>,<0,0,1>,8.235
    CONSTR/CIRCLE,BF,2D,SC_B1.HIT[1..SC_B1.NUMHITS],,
    OUTLIER_REMOVAL/OFF,3
    FILTER/OFF,UPR=0
    ?​
  • Select VIEW from the main menu and make sure EDIT WINDOW and COMMAND MODE are selected. Put your cursor at the place in your routine where you want to add the construction, press enter and start to type CONSTR. As soon as you get as far as CONST it will auto recognise and insert a construction command (press the tab key to accept). You can continue to tab and type or you can hover over the various toggles until a box appears, then click to expand and set the various options. You can also click on them and use F7/F8 to cycle through the available options. You want to set the feature type to CIRCLE, the construction method to BFRE (you should always use BFRE when constructing features from existing hit data), 2D/3D should = 2D (make sure you have the correct workplane active). The hits (or previous features) that you want to use for the construction come next (immediately after the 2D,). Put the cursor on that field and type SC_B1.HIT[..]. This tells PC-DMIS to use all of the hits from feature SC_B1.
  • Neil, I want to question this "the construction method to BFRE (you should always use BFRE when constructing features from existing hit data)" because I have seen no difference in the reported results between BF and BFRE. I have been told over the years to always use BFRE for existing data but have forgotten to toggle to it and when I did saw nothing different in the reported results. I scan ID/OD features a lot and use the scan data for constructing planes/lines/circles/cylinders and have played with BF and BFRE and not seen a difference.
  • The only time there should not be any difference between BF and BFRE constructions would be when a non-contact sensor was used to gather the data (laser scanner / vision camera). For all tactile sensors, there should always be a difference. This is because BF constructions take the previously compensated hits and construct a feature through them, whereas BFRE take the hit centroids (the RAWHITS), construct the feature through those centroids and then apply compensation based on the newly computed feature. The amount of difference will vary depending on how "good" the original hits were - it's a bit like removing the cosine error.
  • Thank you Neil. This is how to input commands with keyboard. I will keep it Slight smile