Elaboration d’une base de données de signaux issus des défauts d’impacts sur les roulements à billes.

Abstract

The world of industry and transport has increasingly efficient and complex machines and installations. This work is part of the theme of developing a database of signals from impact defects on ball bearings. Is simplicity of assembly, its low purchase cost and its mechanical non-robustness, especially its virtual absence of maintenance, justifies its omnipresence in the industrial sector. They cannot be free from disturbances and failures, affecting the quality of the product, which can cause the immediate shutdown of a machine and impair the proper functioning of an entire production system. The diagnosis of these machines is mainly based on monitoring symptoms related to different degradation conditions. These symptoms can be drawn from various sources of information, among which vibration analysis and acoustic emission occupy a prominent place. Today, many effective fault analysis devices are well established, anchored on powerful tools offered notably by the method of vibration analysis. The complexity of these tools requires an expert to use and interpret them. The continued presence of an expert is difficult and expensive. State indicators of the latter exist in the literature but they are designed under the assumption of perfect operating conditions. The main objective of this thesis is to reduce the recourse to human intervention by proposing strategies to design an optimal indicator which summarizes the information of vibrations by numerical values. Vibration Analysis (A.V) is recognized today for its effectiveness in applications for verifying the integrity of mechanical components. This non-intrusive and non-destructive method can be implemented in particular during the "zero" point during the initial acceptance test and in service, without stopping industrial equipment and under normal operating conditions.