Using virtual parts to optimize the metrology process

In the measurement process, there are many parameters affecting the measurement results: the influence of the probe system, material stiffness of measured workpiece, the calibration of the probe with a reference sphere, the thermal effects. We want to obtain the limits of a measurement methodology to be able to validate a result. The study is applied to a simple part. We observe the dispersion of the position of different drilled holes (XYZ values in a coordinate system) when we change the quality of the part and the method of calculation. We use the Design of Experiment (Taguchi method) to realize our study. We study the influence of the part quality on a measurement results. We consider two parameters to define the part quality (flatness and perpendicularity). We will also study the influence of different methods of calculation to determine the coordinate system. We can use two options in Metrolog XG software (tangent plane with or without orientation constraint). The originality of this paper is that we present a method for the design of experiment that uses CATIA (CAD system) to generate the measured parts. In this way we can realize a design of experiment with a largest number of experimental results. This is a positive point for a statistical analysis. We are also free to define the parts we want to study without manufacturing difficulties.