management had me run a test. I taped a 2,4,6 block to the table. took the top plane, x+ and y- and created my alignment. I then called them datum a top b x and c y. I measured 2 holes on each side and did a position in geotol. I used the same measured data in legacy and got a .010 difference. How can this be?
What management is saying is the position formula. 2 x (dx^2 + dy^2)^1/2 when you figure it manually using the measured values matches legacy perfectly but it is not matching in GEOTOL.
I understand this is rough and dirty test but I can not explain why parts we measure good with second methods and legacy are good. GEOTOL (which our customer uses and requires) is failing the part. I have asked Hexagon this question and keep being told that GEOTOL is correct to the standard. Which is great but the standard use the 2 x (dx^2 + dy^2)^1/2 formula and is passing this but failing in GEOTOL.
I am not bashing PCD Hexagon or anything but I can not explain it to management.
Please post your code, including the results. I'd assume, initially, that your alignment doesn't match the coordinate system established by the datum reference frame.
So let me try and give you what I see. the numbers are made up but give you an idea.
Top plane is zero or datum a. drop down to the first hole at .050. measures .0001 dev.
front plane or datum b to the hole is .050 measures .0002 dev
position of .0004
legacy calls it .0004
GEOTOL calls it .0012
Please validate whether you have DEFAULT or LEAST SQUARES setup on GeoTol. This may give different results based on the algorithm, datum setup, number of points taken, etc.
Are your datums using three planes? How many points per plane, and more importantly, is your hole measurement set to max_insc(assuming you are measuring a hole), cuz your geo_tol will do that automatically changing your center point. Also if you're messing around with going back and forth too much, it can mess up your nominal, make sure they are the same on both legacy and geo tol. I would run it from scratch anytime you make any changes.
Yes 3 planes. 3 points per plane. Yes it is. Now that is also something they are questioning. Who cares how it's figure out the center. The center is measuring. 0001 in 1 direction and.0002 in the other. The position formula does not care about anything other then that. So how can it report 2 different values?
if the form of your hole you are measuring is out of round, it will make a lot of difference on your center point. Can you report out the form of the circle you are measuring? or are you saying your actuals are the same but your tps are different? Need more info, pic of your report perhaps?