Análise de vibração em motores elétricos com mouse óptico

Abstract

This work aimed at studying a new methodology for detecting and diagnosing mechanical flaws upon vibration analysis in rotating equipment. The use of a laser optic mouse as data collection sensor, associated to signal analysis was the adopted methodology. The technique is experimentally implemented for detecting mechanical flaws based on the time domain analysis by means of temporal response. However, it does not allow us to classify the type of flaw, making necessary the use of other parameters. In this work, we propose the use of statistical equations, in the context of standard recognition, to diagnose mechanical flaws using the vibration signals as input. To validate this technique, we collected data by means of a portable vibration analysis gauge present in the market. All the computational routine to implement the techniques for detection and diagnosis of flaws was conducted at the environment of the MATLAB program. The results showed that the new proposal for detecting and diagnosing flaws is a promising technique for being non-destructive and noninvasive, with the prospect of becoming a valuable tool for predicting flaws in the context of predictive maintenance.