Here are five things CMM operators wish everyone knew

chenfu1

“If you don’t know what you’re dumb at — you’re dangerous.”
& u, N9 Q, o; p- y% V  W Few people would disagree that a coordinate measuring machine (CMM) is often the most expensive item of inspection, measuring, or test equipment an industrial facility has in its quality arsenal. Very few people in those same facilities — outside of the CMM operator or programmer — realize the CMM is probably the most misunderstood item of inspection equipment. In some cases, the CMM can be used incorrectly or under utilized.
# T5 D" y6 m% h0 r( V2 i 1. A is for algorithm. — CMMs calculate measurement values by connecting dots that form features (planes, circles, lines, etc.) from points that are probed or constructed. In the CAD or engineering world, every plane is flat and every circle is round. In the real world, nothing is flat, round, or parallel. For example, when a circle is measured from five data points, there are at least three circles that can be generated from the data points because the five–point circle is not perfectly round — the smallest circle (sometimes called best fit), the largest circle, and the average circle. CMM software might calculate the average circle by default; but to measure within geometric dimensioning and tolerancing (GD&T) standards, the algorithm (or calculation) will need to be changed for outside diameters and inside diameters. If you use the wrong algorithm for the specific application or feature, you’ll get an incorrect reading.
2. Quit asking, “Why probe so many points?” — CMMs can calculate or construct a feature from the minimum number of points required by the laws of geometry. A plane or circle can be constructed from three points and a line from only two points. In the world of metrology, using the minimum number of points is ignorant. If a CMM operator is measuring a plane (that will serve as a datum from which several features will be compared), five points would be much better than three, while seven or nine points would be better than five. To be totally honest, 11 or 13 would be even better yet. When you get to around 17 or 19 points, you’ll be approaching the curve on which the time needed to probe additional points doesn’t compensate for the improvement of the measurement.
/ j; _/ Q( F* k5 D  P; N. }. R
  X) I# c- d' `8 r( ^+ B7 s8 `1 Khttp://www.manufacturingcenter.com/tooling/archives/0207/Departments/0207_quality.gif 3. It is what it is: CMM’s tolerance by the GD&T book. — If you really want to know how well someone understands GD&T on the parts he makes or inspects, I recommend creating the GD&T setups on a CMM and observing the calculations made by the software between different features. CMMs will also help because explanations of these calculations can be found in the user manual or help file. You’ll never view tolerances such as concentricity, positional location, or perpendicularity the same after you really “see” how they work within your CMMs calculations.

4. Very seldom is a reliable measurement made “quickly.” — A few years ago NASA implemented the slogan “Better, Faster, Cheaper,” only to realize later that better is seldom faster and faster is rarely cheaper. Respect how long a CMM operator says it will take him to confirm your part, and don’t argue with him as if you know anything about CMMs if you don’t. After all, it’s his initials on the inspection report. A very quick–witted CMM programmer once told me that a machinist had asked him why it would take so long to confirm a part. The CMM programmer had complemented the machinist on the complexity of the part and had asked him how long it took him to program the mill to machine the part. Very confidently, the machinist said, “About a day–and–a–half.” The CMM programmer smiled and said to the machinist, “See you in a day–and–a–half.”
# A3 {; z( D9 I0 M) u' P 5. Check the drawing. — Nothing annoys a CMM operator more than spending 30 to 40 minutes measuring what somebody thinks should be measured. If you need a feature (or features) of a part confirmed, always bring along a copy of the part drawing to make sure your getting what you need before someone does it twice. Sometimes your CMM operator can even alert you to other geometry of the part that might need to be measured.
8 c# p* O2 ?4 N  y( h Chances are your CMM operator has already been trying to explain these items to people with little or no success. In my personal experience, the foolish are too busy thinking of the next question to listen to the answer being given. In other words, when you don’t really know what you’re talking about, speak with your ears!