Estudo da influência do fator de recobrimento na detecção de falhas em engrenagens por análise de sinais vibratórios

Abstract

Preventive maintenance is gaining space, since it provides gain of competitiveness as companies, reducing the time and the quantity of interventions. A vibration analysis is a process of maintenance, being responsible for the identification of the faults. This work has as main factor the failure of gears by means of the analysis of vibratory signals. For this purpose, we performed tests on a bench with the purpose of achieving an induction of slack, misalignment, wear and breaking tooth. For an analysis of the acquired vibratory signals, a program developed in the MATLAB® platform was implemented the techniques of RMS, crest factor and kurtoses, as well as a spectral analysis through the FFT. The increase in the production factor degradation produced vibrational changes with RMS values between 11 and 80 times higher than the reference condition, values of crest factors between two and three, and the low variation in kurtosis. More expressive values for the statistical techniques and a behavior closer to the vibration spectra are expected in tests performed with the application of stresses in the gears.