Assessment of measuring processes by mathematical method

The major risks of the measurement metrological assurance system are the acquisition of measurement information not possessing the properties adequate to make the required decisions. The occurrence of these risks results in an incorrect decision, which can potentially cause damage to the enterprise, so these risks must be predictable. The probability of making an incorrect decision during the tolerance control of the part dimensions is affected by the combination of the measurement error and the controlled parameter dispersion. The article describes a mathematical model for assessing the influence of the measurement error on the probabilistic characteristics of measurement results during tolerance control. Delthe7 environment-based software has been developed to calculate probabilistic sorting parameters, where the characteristics of the controlled parameter and measurement uncertainty distribution laws are used as input data, and the assessment is performed to the left and right of the tolerance field boundary for any possible positions of the middle of size dispersion relative to the middle of the tolerance field. The output is probabilities of classifying good parts as defective parts and defective parts as good parts as risks of making an incorrect decision. The program was tested on the control of crank and pin journals of the bent shafts of YaMZ-238 engines machined to repair size. It has been revealed that the dynamics of measuring the probabilistic characteristics of sorting parameters is common for both crank and pin journals of the bent shafts; the probabilities and risks increase with the measurement error. The quantitative value of these parameters, other conditions being equal, largely depend on the distribution characteristics not only of the measured value, but also of the real error of the measuring instruments used.