Tesis
Análisis de modos y efectos de falla en rodamientos por defectos de lubricación a través de ultrasonido en un simulador electroneumático
Fecha
2020-03-10Registro en:
Ausay Baquero, César Aurelio; Estrada Tualombo, Willam Enrique. (2020). Análisis de modos y efectos de falla en rodamientos por defectos de lubricación a través de ultrasonido en un simulador electroneumático. Escuela Superior Politécnica de Chimborazo. Riobamba.
Autor
Ausay Baquero, César Aurelio
Estrada Tualombo, Willam Enrique
Resumen
The main aim of this project was the construction of a test simulator to facilitate data analysis of bearings lubrication conditions are made up of power, transmission, electrical, hydraulic and test system, the goal of which was to help characterize the failures of the incipient bearings based on calculations of the amount of grease, type of grease and load to be applied. Then Measurements were carried out under normal operating conditions by ULTRA PROBE 15000 ultrasound analyzer VIBRACHECK 200 vibration analysis and TPI 9080 smart trend meter determining the baseline for an ultrasound that is less than -47 dB and in vibrations. It is between 0 - 50 BDU. Whose values were the baseline for monitoring conditions of the bearings under normal conditions calculated amount of grease, excess grease, and shortage of grease managing to determine the failure conditions that were for ultrasound greater than -25 dB and vibrations greater than 100 BDU. After the measurements in the 15 bearings were obtained 15 values, they were evaluated using the Shapiro Wilk parametric normality test. It is resulting in two of the three conditions that do not behave according to the normal distribution. Therefore, the Wilcoxon Mann Whitney test of the non-parametric type so concluding that the failure times when the bearings have a shortage of grease, excess grease, and normal lubrication conditions are different. Finally, a statistical analysis of the bearing service life was carried out a period during which the bearings in an incipient failure mode were performed using the Weibull distribution. This last calculation is supported by the Kolmogorov-Smirnov goodness-of-fit test, which confirms that the times obtained from the measurements are adjusting to the Weibull distribution with 95% reliability.