We need to evaluate the roundness from the highest points of an outside knurl. This knurl luckily always has the same amount of "teeth".
We scan this outside knurl and then use a script to find all the highest points.
We then create a least square circle from the highest points and evaluate this circle.
The script was created by Hexagon (I don't know how it works).
In it's current state only works on an outside geometry but you could adjust it to your needs.
Explanation how you need to have your routine setup for the script to work:
Whatever the method, the problem is the scan density.
Maybe there's no hit on the highest / lowest points.
Without using a script, I believe you can do the same with an array :
(Assuming CIRC1 is the max inscr circle for an ID or the min circ for an OD)
C1=COMMENT/INPUT, enter the number of tooth
ASSIGN/V1=SQRT(DOT(SCAN1.HIT[1..SCAN1.NUMHITS].XYZ-CIRC1.XYZ,SCAN1.HIT[1..SCAN1.NUMHITS].XYZ-CIRC1.XYZ))
ASSIGN/V2=MININDEX(V1)..........................................or maxindex for an OD, it's used to align on the hihest / lowest hit)
ASSIGN/V3=SCAN1.HIT[V2].XYZ
F1=GENERIC/POINT,DEPENDENT,CARTESIAN,OUT,$
NOM/XYZ,<V3.X,V3.Y,V3.Z>,$
MEAS/XYZ,<V3.X,V3.Y,V3.Z>,$
NOM/IJK,<0,0,1>,$
MEAS/IJK,<0,0,1>,$
A2=ALIGNMENT/START,RECALL:A1,LIST=YES
ALIGNMENT/LEVEL,ZPLUS,PLN1
ALIGNMENT/ROTATE_CIRCLE,CIRC1,F1
ALIGNMENT/TRANS,XAXIS,CIRC1
ALIGNMENT/TRANS,YAXIS,CIRC1
ALIGNMENT/TRANS,ZAXIS,PLN1
ALIGNMENT/END
ASSIGN/ANG_HIT=RAD2DEG(ATAN(SCAN1.HIT[1..SCAN1.NUMHITS].Y/(SCAN1.HIT[1..SCAN1.NUMHITS].X+0.0000000001)
ASSIGN/COR_TAN=(((SCAN1.HIT[1..SCAN1.NUMHITS].X-ABS(SCAN1.HIT[1..SCAN1.NUMHITS].X))/((SCAN1.HIT[1..SCAN1.NUMHITS].X-ABS(SCAN1.HIT[1..SCAN1.NUMHITS].X)))*90.....(array of 0 or 180)
ASSIGN/ANG_COR=ANG_HIT+COR_ANG
ASSIGN/ANGL_TH=360/C1.INPUT
V4 =LOOP/START,ID=YES,NUMBER=C1.INPUT,START=1,SKIP=,
OFFSET:XAXIS=0,YAXIS=0,ZAXIS=0,ANGLE=0
ASSIGN/ANGL1=ANGL_TH*(V4-1)
ASSIGN/V5=MININDEX(ABS(ANG_COR-ANGL1))................................................searching the closest hit of the actual tooth
ASSIGN/V6=SCAN1.HIT[V2].XYZ
F2=GENERIC/POINT,DEPENDENT,CARTESIAN,OUT,$
NOM/XYZ,<V6.X,V6.Y,V6.Z>,$
MEAS/XYZ,<V6.X,V6.Y,V6.Z>,$
NOM/IJK,<0,0,1>,$
MEAS/IJK,<0,0,1>,$
LOOP/END
CIR2 =FEAT/CIRCLE,CARTESIAN,IN,LEAST_SQR,YES
THEO/<0,0,0>,<0,0,1>,50
ACTL/<0,0,0>,<0,0,1>,50
CONSTR/CIRCLE,BF,2D,F2[1..C1.INPUT],,
OUTLIER_REMOVAL/OFF,3
FILTER/OFF,UPR=0
Sorry, I missed the bit where he said he was running 2019 R2. The Extreme point was added in 2021.2 to make it easier for people wanting to do this kind of thing. It was a highly voted for item on the ideas centre because most lesser mortals aren't as proficient with assignments and arrays as
JEFMAN
