Detección de Fallos en Engranajes a Través del Análisis de Firma de Corriente del Motor

Rotating machines that generally work 24 hours a day can fail, causing economic losses for the company. For this reason, maintenance techniques are used for the early detection of these failures. Condition monitoring is one of the techniques used, where through the measurement of mechanical and electrical magnitudes or the performance of non-destructive tests, the condition of the elements that make up the mechanical systems can be determined. Thus, the present project deals with the analysis of motor current signatures (AFCM) used to detect faults in a gearbox, being these faults tooth breakage and tooth pitting at different levels of severity. The implemented methodology consists of acquiring current signals from the motor, these signals are acquired with the machine working in a first instance in normal condition (without faults) and then working in condition with artificially implanted faults. Once the signals are acquired, they are processed by means of the Fast Fourier Transform, then an analysis of the frequency spectra is performed where the characteristic frequencies of the machine are determined and through the comparison of the spectra the condition of the machine can be determined. The results obtained show that a change in the frequency spectrum is generated depending on the failure analyzed, it can also be observed that this spectrum is modified depending on the severity of the failure, this methodology can be used to detect failures with a mild to severe severity.