Hey all you wonderful people, I just have a question in regard to CMM performance. Does anyone do routine wrist checks? What we do is measure the sphere at T1A0B0 and align to it. Then come back down and measure again at all your 90° Tip angles and report the offsets in XYZ. If you have done something like this, what kind of deviation are you seeing? I know it obviously depends on the machine/Temperature/Stylus Setup. Using my Global Performance 12.15.10 that is equipped with a Hexagon HH-AS8-T2.5 head, and a Leitz LSPX1H-T Analog scanning sensor, I am seeing around 2-3 microns deviation between all the tips for my latest checks. I am using a 4by20mm probe. The Temperature while measuring was stable at approx. 21.3°C, and I made sure the artifact adapted to this environment (part sat in the room for approx. 2.5 hours). The program that came with the Hexagon Artifact we purchased last year was using a tolerance of 2X the sphere tolerance for the artifact. So, to calculate the tolerance, you use the following formula based off the CMM spec, sphere diameter, & uncertainty of the sphere. It looks like this.
Tol=(A + ((B*L)/333))*.001 + Sphere Uncertainty
Tol=(1.8+((0*24.98725)/333))*.001+.00025.
Tol=.0027 x 2
Tol=.0054mm.
I am within the tolerance they supplied with the artifact, but I am just curious to as of how they came up with this value. On the spec sheet for this head, it says the positional repeatability is .5 micron. Also, from run to run I seem to vary a few microns. I am assuming this could be due to the fact the artifact might not be positioned on the table in the same location. But I'm just curious to see if anyone has run tests like this before, and what kind of offsets you are seeing.
