Comparación de Señales de Vibración, Acústica y de Corriente para la Detección de Fallos en Rodamientos Mediante la Transformada Rápida de Fourier, Envolvente y Cepstrum

Bearings are fundamental elements in rotating machinery; a fault in them generates a decrease in the performance of the machine and its useful life. Therefore, maintenance programs more frequently apply condition monitoring techniques to avoid a severe fault in the elements of the machinery. In order to compare the detection of failures located in the bearings of a reciprocating compressor, using condition monitoring techniques with acoustic, vibrational, and current signals, the damage was simulated in the inner race, outer race, and rolling element in a conical bearing located in a reciprocating compressor. The acquired signals were processed using the Fast Fourier Transform, the envelope analysis, and the Cepstrum analysis. Then the spectra were compared in normal condition and failure condition, obtaining that the spectra in normal condition are recognizable as the characteristic frequencies of the machine and that in fault condition there are changes in the spectrum, it is concluded that a detectable failure will depend on the type of fault, the type of signal acquired and the processing algorithm used.