Application of Statistical Methods to Assess the Quality of the Machining Process

Abstract

The finishing processes in mechanical engineering are aimed at creating ideal surfaces of parts satisfying specifications, in particular their dimensions, but these processes carried out on actual equipment are subject to deviations from specified values. Statistical methods of quality control and management are a diagnostic tool for analyzing the causes of variations. The diameters of 20 crankshaft main journals of the ZMZ engine were investigated after machining for the repair size. The quality of machining crankshaft journals was evaluated sequentially using quality tools such as the box-and-whisker diagram, Shewhart control charts, and histograms. It was found from the box-and-whisker diagram that after machining for the repair size all the journals had the diameter within the machining tolerance field, but at the same time there was a scatter between journal diameters of the same crankshaft. An analysis of the Shewhart control charts established that the scatter values within one subgroup (crankshaft) and between subgroups were caused by usual reasons, and this scatter pattern can be considered random. Hence, the process can be considered stable and controllable. The histogram helped to determine the probability of 1.95% for recoverable rejects after machining the engine journal for the repair size and the probability of zero for unrecoverable rejects. The calculated reproducibility index of 0.93 indicates a significant wear of the equipment. The index will decrease over time and the risks of rejects will increase. The authors conclude that the quality of setting for machining crankshaft main bearing journals for the repair size is acceptable. It is recommended that the quality of machining crankshaft main journals for the repair size be checked regularly applying successively all three tools to identify the size variation and inconsistencies, which is particularly important for worn equipment.