Publication: Análisis y evaluación de los niveles de energía obtenidos mediante WPT para defectología en ejes
Loading...
Identifiers
Publication date
2012-10
Defense date
2012-10-15
Authors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
En la última década, la detección de defectos incipientes en elementos mecánicos rotatorios, ha motivado numerosos estudios de investigación en toda Europa. La industria cada vez está más concienciada de la importancia de conocer el estado de sus máquinas antes de producirse un fallo catastrófico. Este espíritu de mejora de la calidad y de la seguridad, ha llevado a la demanda de sistemas de detección de fallos en componentes mecánicos mediante sistemas en línea de monitorización de su estado. Estos métodos de diagnóstico no destructivos se basan, principalmente, en el análisis de señales vibratorias. Dichas señales se consideran de naturaleza estocástica y no estacionaria, por lo que su tratamiento se debe realizar mediante técnicas avanzadas de análisis de señales. Existen diferentes herramientas para tratamiento de este tipo de señales, siendo las más conocidas las relacionadas con la Transformada de Fourier. Actualmente se están investigando nuevas metodologías que van más allá del análisis en frecuencia que proporciona la Transformada de Fourier. Las nuevas líneas de investigación están desarrollando sus estudios basándose en la Transformada Wavelet, una herramienta matemática que permite el estudio de la señal, además de en frecuencia, en el tiempo. Este proyecto busca el desarrollo de un patrón de energía de la señal, que permita la identificación de diferentes tipos de defectos en ejes de ferrocarril. Los estudios están basados en el cálculo de la Transformada Wavelet por Paquetes para las señales de vibración obtenidas de diferentes ensayos. Los ensayos se realizan en un banco de ensayos de laboratorio para un eje de ferrocarril a escala 1/8: para el eje en estado sano (sin defecto) y para el eje con diferentes estados de defecto y a velocidades de giro diferentes. Los resultados muestran que, a pesar de tener una variabilidad alta en los ensayos, se pueden obtener conclusiones que indican que es posible la detección de ciertos defectos a velocidades altas. __________________________________________________________________________________________________________________________
In the last decade, the detection of incipient faults in rotating mechanical components, has led numerous research studies across Europe. The industry is becoming increasingly aware of the importance of knowing the status of their machines before a catastrophic failure occurs. This spirit of improving the quality and safety, has led to demand, fault detection systems of mechanical components through on-line monitoring systems. These non-destructive diagnostic methods are based, mainly on the analysis of vibration signals. These signals are considered stationary and non-stochastic, so its treatment should be performed by advanced techniques of analysis signals. There are different tools for processing such signals, the best known are methods related to the Fourier Transform. Are currently being investigated new methods that are beyond the frequency analysis provides Fourier Transform. The new lines of research are developing their studies based on Wavelet Transform, a mathematical tool that allows the study of the signal, in addition to frequency, in time. This project aims to develop an energy pattern of the signal, which allows the identification of different types of defects in railway axles. The studies are based on the calculation of the Wavelet Packet Transform for vibration signals obtained from different tests. The tests were performed in a laboratory test bench for a railway axle scale 1/8: for axle in healthy state (without blemish) and different defect states with different rotational speeds. The results show that, despite having a high variability in the tests, conclusions can be reached that indicate that it is possible to detect certain defects at high speeds.
In the last decade, the detection of incipient faults in rotating mechanical components, has led numerous research studies across Europe. The industry is becoming increasingly aware of the importance of knowing the status of their machines before a catastrophic failure occurs. This spirit of improving the quality and safety, has led to demand, fault detection systems of mechanical components through on-line monitoring systems. These non-destructive diagnostic methods are based, mainly on the analysis of vibration signals. These signals are considered stationary and non-stochastic, so its treatment should be performed by advanced techniques of analysis signals. There are different tools for processing such signals, the best known are methods related to the Fourier Transform. Are currently being investigated new methods that are beyond the frequency analysis provides Fourier Transform. The new lines of research are developing their studies based on Wavelet Transform, a mathematical tool that allows the study of the signal, in addition to frequency, in time. This project aims to develop an energy pattern of the signal, which allows the identification of different types of defects in railway axles. The studies are based on the calculation of the Wavelet Packet Transform for vibration signals obtained from different tests. The tests were performed in a laboratory test bench for a railway axle scale 1/8: for axle in healthy state (without blemish) and different defect states with different rotational speeds. The results show that, despite having a high variability in the tests, conclusions can be reached that indicate that it is possible to detect certain defects at high speeds.
Description
Keywords
Tecnología ferroviaria, Ejes ferroviarios, Elementos mecánicos rotatorios, Vibraciones, Detección de defectos