PC-DMIS requires a defined Auto Circle target and can only look in this pre-defined location for the edge. The problem that I run into is the diameters are often not very round. The target boundaries are two concentric circles with a dashed nominal diameter circle midway between them. These are perfectly round circles. When the circle to be measured is out of round enough to leave this target area, no points are recorded. Expanding the Auto Circle target boundaries is often not an option, especially with top lighting involved. A lot of points get recorded that are not on the edge of the circle that is supposed to be measured.
There’s a solution to keeping the target area small when the circle is round and automatically expanding it to a controlled diameter that’s dependent on how much it’s out of round. This is based on a pre-measurement of the circle with a few points around the diameter for an estimate of the out of round condition, as well as giving the nominal diameter and location.
In the Auto Circle code there’s a line for HIT TARGET.
HIT TARGET=EA1,0,1,LOW, 1.659
The number at the end, 1.659 is the radial distance between the dashed nominal diameter, and the two target diameters. Changing this number to a variable that uses the amount of reported roundness in the pre-measured circle (LOC2.RN) controls the size of the hit target.
HIT TARGET=EA1,0,1,LOW, 0.2+(LOC2.RN.MEAS*0.8)
A study that lasted 6 hours gave me a variable that works pretty well.
The 0.2 is the size that will be used when the reported roundness is zero. As the (out of) roundness grows, the hit target limits grow proportionally.
Are there any comments, questions, or other solutions?
Attached Files
