| Basics of Coordinate Metrology Unit 12: Evaluation of the Measurement - Fit Algorithms | Step 2 of 6 |
| Whenever a standard geometric elements is measured using more points than the minimum number of probing points, scattered values are obtained. These scattered values are an indication of the quality of the measurement and of the part. However, at the same time, the standard geometric element has been overdetermined, since it was probed using more points than mathematically necessary. This is why a suitable substitute element must be calculated using a so-called fit algorithms. Various algorithms are available, which are best explained using the example of a "circle": | |||
 |  |  |  |
| The fit algorithm according to Gauss simply calculates a circle which lies as accurately as possible "in the middle" of all measured points. In the Gauss algorithm, all points have equal weight. | The fit algorithm called "circumscribed circle" calculates a circle such that all measured points are inside the circle, and the circle is as small as possible. This algorithm is usually used for determining the geometric data of shafts when pair dimensions are checked. | The fit algorithm called "inscribed circle" calculates a circle such that all measured points are outside the circle, and the circle is as large as possible. This algorithm is usually used for determining the geometric data of bores when pair dimensions are checked. Attention: The center of the circumscribed circle is in a different place than the center of the inscribed circle. | The fit algorithm called "minimum zone circle" calculates two circles such that one is outside all points and the other one inside all points but that both have the same center. Then the distance of the two circles (the difference in diameters, shown on the picture in red) is the form deviation. The minimum zone circle (Chebychev circle) is then the middle circle of these two circles. This algorithm is used almost exclusively for the calculation of form errors. |