Finding the Low point of a circle

[zachzwp]

I am working on a program that takes 2 circles and two planes with 1,440 points. I have constructed the planes and circles from the points to find the runouts of these features. Is there a halfway decent way to find the low point of these features or the location of the max and min of the circle. Essentially I want to find the place on the part where the circle's diameter is the closest to the center of the part. I tested a couple ways, but with this many points I am not sure if there is an efficient way to approach this.

[Matthew D. Hoedeman]

if you need a physically measured point, simply align to the circle (say XY origin, XY zero), then measure a point at X0,Y0(-cir1R),z,0,1,0 (rough code)

[louisd]

Change your circles from Cartesian to Polar.
Then, you can use a maxindex and minindex command for each circle to extract the max and the min.

ASSIGN/MAX_PR_1=MAXINDEX(ARRAY(CIR1.HIT[1..CIR1.NUMHITS).PR))
ASSIGN/MIN_PR_1=MININDEX(ARRAY(CIR1.HIT[1..CIR1.NUMHITS).PR))
ASSIGN/MAX_PR_2=MAXINDEX(ARRAY(CIR2.HIT[1..CIR2.NUMHITS).PR))
ASSIGN/MIN_PR_2=MININDEX(ARRAY(CIR2.HIT[1..CIR2.NUMHITS).PR))

Once max/min assigns extract the hit number... You can generate a generic hit, to output the polar angle and polar radius of each max & min point.

MAX_PT_1 =GENERIC/POINT,DEPENDENT,POLAR
NOM/XYZ,<CIR1.HIT[MAX_PR_1].TPR, CIR1.HIT[MAX_PR_1].TPA, CIR1.HIT[MAX_PR_1].TZ>
MEAS/XYZ,<CIR1.HIT[MAX_PR_1].PR, CIR1.HIT[MAX_PR_1].PA, CIR1.HIT[MAX_PR_1].Z>
NOM/IJK,<CIR1.HIT[MAX_PR_1].TI,<CIR1.HIT[MAX_PR_1].TJ,<CIR1.HIT[MAX_PR_1].TK>
MEAS/IJK,<CIR1.HIT[MAX_PR_1].I,<CIR1.HIT[MAX_PR_1].J,<CIR1.HIT[MAX_PR_1].K>

Repeat for MIN_PT_1 and circle 2's MAX_PT_2, and MIN_PT_2.

Output your locations PR and PA for each point.
you have been forewarned, this WILL extract outliers as valid hits. so make sure you have zero oil, dirt FOD Debris on your sample parts and probe tip.

[zachzwp]

I will try this, thanks!

[RandomJerk]

Did they finally add PR to a feature's hits? Cool.

[louisd]

If you designate the coord system to polar, it should present itself as being available, along with PA and Z.
I believe if you keep the coord system as Cartesian, you are stuck with only xyz ijk as data values. Polar data is null.

[RandomJerk]

Another way:

TMP_XY =FEAT/CONTACT/CIRCLE/DEFAULT,CARTESIAN,IN,LEAST_SQR
THEO/<0,0,0>,<0,0,1>,1.5
ACTL/<0,0,0>,<0,0,1>,1.5
TARG/<0,0,0>,<0,0,1>
START ANG=0,END ANG=360
ANGLE VEC=<0,1,0>
DIRECTION=CCW
SHOW FEATURE PARAMETERS=NO
SHOW CONTACT PARAMETERS=YES
NUMHITS=3,DEPTH=1,PITCH=0
SAMPLE METHOD=SAMPLE_HITS
SAMPLE HITS=0,SPACER=1
AVOIDANCE MOVE=BOTH,DISTANCE=3
FIND HOLE=CENTER,ONERROR=NO,READ POS=NO
SHOW HITS=NO

A2 =ALIGNMENT/START,RECALL:A1,LIST=YES
ALIGNMENT/TRANS,XAXIS,TMP_XY
ALIGNMENT/TRANS,YAXIS,TMP_XY
ALIGNMENT/END

ASSIGN/MIN_R=999
ASSIGN/MAX_R=-999

V1 =LOOP/START,ID=YES,NUMBER=TMP_XY.NUMHITS,START=1,SKIP=,
OFFSET:XAXIS=0,YAXIS=0,ZAXIS=0,ANGLE=0

ASSIGN/PNT_R=SQRT(TMP_XY.X^2+TMP_XY.Y^2)
ASSIGN/MIN_R=IF(PNT_R<MIN_R,PNT_R,MIN_R)
ASSIGN/MAX_R=IF(PNT_R>MAX_R,PNT_R,MAX_R)

LOOP/END

[JEFMAN]

louisd , Thx for the info about PR !!!!!
I would say that, if you construct a filtered circle from hits of the measured circles, it doesn't use outliers.

As I didn't know that PR was available in polar only, I usually do it :
ASSIGN/POLAR_RAD=SQRT(DOT(CIRC1.HIT[1..CIRC1.NUMHITS].XYZ-CIR1.XYZ,CIRC1.HIT[1..CIRC1.NUMHITS].XYZ-CIR1.XYZ))
Then
ASSIGN/V1=MAXINDEX(POLAR_RAD)
ASSIGN/V2=MININDEX(POLAR_RAD)
ASSIGN/RMAX=CIRC1.HIT[V1].XYZ
ASSIGN/RMIN=CIRC1.HIT[V2].XYZ

and use RMAX.X, .Y and .Z in a generic feature.