Greetings
My team has recently been battling a profile check and we've discovered that the culprit was the use of lines rather than auto-vector points. I was wondering if anyone else had experienced this, had some mental triggers as to when to choose points over lines, etc.
We started with scanning the profile and it also gave us some unreliable data, so we decided to see what kind of success we'd have by building a set from multiple scans and/or hits of individually built auto-features. Pretty straight forward. A few lines and a few radii. For some reason we were getting even less reliable data. Hard checks were disproving the results we were getting on the CMM, but we couldn't figure out why the profiles looked so off. The strangest result was on the top face. We'd have 3 corners showing equal/level to each other, and one corner showing way higher or lower than the other 3. Indicator showed that there was no such deviation.
All of this to say that once we switched to vector points (note, with the same exact xyz nominals of the auto-lines), the profiles were near perfect, consistently.
Any thoughts or similar experiences?
