Thesis FACTIBILIDAD DE DETECTAR FALLAS EN UNIDAD DE SOPORTE DE RODADO DE ALTA CARGA Y BAJA VELOCIDAD, MEDIANTE EL ANÁLISIS DE VIBRACIONES
Date
2017
Authors
DREVES BAZAN, TOMAS PABLO
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
European Southern Observatory (ESO) dentro de sus equipos operacionales ubicadosen cerro Paranal (Regi´on de Antofagasta) cuenta con un sistema mec´anico llamadoRotation Mechanism (RTM), equipos encargados de hacer girar los domos de los telescopiospara realizar las observaci´on astron´omica. En los ´ultimos a˜nos, estos equiposhan presentado fallas en sus rodamientos, problema que ha llegado a parar uno de los telescopioshasta por una semana. Por esta raz´on ESO plantea y solicita hacer un an´alisisde vibraciones a los sistemas Rotation Mechanism con el fin de estudiar la factibilidadde detectar de manera anticipada las posibles fallas y planificar el reemplazo de losrodamientos. Esta memoria expone dicho estudio realizado entre Agosto y Diciembredel a˜no 2016.Los RTM operan a muy bajas revoluciones como tambi´en soportan una alta carga,yaque debe sobrellevar de manera completa el peso del domo (ocho por unidad). Estohace que deban emplearse t´ecnicas no tradicionales para el an´alisis de vibraciones.Las t´ecnicas elegidas para el an´alisis de los datos obtenidos fueron: Promedio dese˜nal en frecuencia y an´alisis de kurtosis mediante kurtogramas. La primera de estast´ecnicas se eligi´o debido a que la se˜nal que se obtiene por medici´on presenta una grancantidad de impactos indeterminados que perturban el an´alisis en frecuencia, por loque cada medici´on fu´e separada por estos impactos dejando sub-mediciones aptas paraser trabajadas. La segunda t´ecnica se escoge debido a la capacidad de detectar se˜nalesimpulsivas (efecto campana), tales como las que se producen al tener un rodamiento enestado de deterioro en sus componentes. Para el an´alisis de las se˜nales se elige Pythondonde se aplican los algoritmos. Las mediciones se realizan tomando muestras a unatasa 5555 muestras por segundo permitiendo registrar un giro completo del domo y almismo tiempo mantener una respuesta en frecuencia bajo.Para calibrar las t´ecnicas mencionadas se modelan se˜nales con el fin de simular unafalla en rodamiento del tipo impulsiva y repetitiva en el tiempo, a˜nadi´endole altos nivelesde ruido aleatorio. Ambos m´etodos logran extraer las se˜nales impulsivas de maneraefectiva. Al analizar las mediciones obtenidas no se observan principios de fallas en losrodamientos de los RTM de la Unidad de Telescopio 2, no obstante este an´alisis no esconcluyente ya que existe la posibilidad que el alto ruido de vibraci´on del enclosuresea demasiado alta comparado con la falla asociada al rodamiento. Finalmente se proponencomo futuras t´ecnicas el an´alisis de lubricante y emisiones ac´usticas adem´as dela caracterizaci´on de las cargas y de las fallas hist´oricas dentro del observatorio.
European Southern Observatory (ESO) within its mechanical systems located in CerroParanal (Antofagasta Region) has one called Rotation Mechanism (RTM), equipmentin charge of rotating the domes of the telescopes in order to perform the astronomicalobservation. In recent years, these equipments have failed in their bearings, aproblem that disable one of the telescopes up to a week. For this reason ESO raises andrequests to do a vibration analysis to the Rotation Mechanism systems in order to studythe feasibility of detecting faults in advance and planning the replacement of bearingsat time. This report presents the study carried out between August and December of theyear 2016.The RTMs operate at very low speeds as they also withstand a high load, as they mustfully stand the weight of the dome (eight per unit). This implies that non-traditionaltechniques for the analysis of vibrations must be used.The techniques chosen for the analysis for the obtained data were: Average of signalin frequency and analysis of kurtosis by the use of kurtograms. The first of these techniqueswas chosen because the signal obtained by measurement presents a large numberof indeterminate impacts that disturb the frequency analysis. So that, each measurementwas separated by these impacts leaving sub-measurements suitable for work. Thesecond technique is chosen because of the ability to detect impulsive signals (bell effect),such as those that occur when having a bearing in state of deterioration in itscomponents. For the analysis of the signals, Python is chosen. Measurements are madeby sampling at a rate of 5555 samples per second allowing a full turn of the dome to berecorded while maintaining a low frequency response.Before using the aforementioned techniques, signals are modeled to simulate a rollingfailure of the impulsive and repetitive type over time, adding high levels of randomnoise. Both methods manage to extract the impulsive signals effectively. When analyzingthe measurements obtained, there are no principles of failure in the RTM bearingsof the Telescope Unit 2, however this analysis is not conclusive since there is the possibilitythat the high vibration noise of the enclosure is too high compared to the failureassociated with the bearing. Finally, as future techniques, the analysis of lubricant andacoustic emissions is proposed as well as the characterization of the loads and the historicalfaults within the observatory.
European Southern Observatory (ESO) within its mechanical systems located in CerroParanal (Antofagasta Region) has one called Rotation Mechanism (RTM), equipmentin charge of rotating the domes of the telescopes in order to perform the astronomicalobservation. In recent years, these equipments have failed in their bearings, aproblem that disable one of the telescopes up to a week. For this reason ESO raises andrequests to do a vibration analysis to the Rotation Mechanism systems in order to studythe feasibility of detecting faults in advance and planning the replacement of bearingsat time. This report presents the study carried out between August and December of theyear 2016.The RTMs operate at very low speeds as they also withstand a high load, as they mustfully stand the weight of the dome (eight per unit). This implies that non-traditionaltechniques for the analysis of vibrations must be used.The techniques chosen for the analysis for the obtained data were: Average of signalin frequency and analysis of kurtosis by the use of kurtograms. The first of these techniqueswas chosen because the signal obtained by measurement presents a large numberof indeterminate impacts that disturb the frequency analysis. So that, each measurementwas separated by these impacts leaving sub-measurements suitable for work. Thesecond technique is chosen because of the ability to detect impulsive signals (bell effect),such as those that occur when having a bearing in state of deterioration in itscomponents. For the analysis of the signals, Python is chosen. Measurements are madeby sampling at a rate of 5555 samples per second allowing a full turn of the dome to berecorded while maintaining a low frequency response.Before using the aforementioned techniques, signals are modeled to simulate a rollingfailure of the impulsive and repetitive type over time, adding high levels of randomnoise. Both methods manage to extract the impulsive signals effectively. When analyzingthe measurements obtained, there are no principles of failure in the RTM bearingsof the Telescope Unit 2, however this analysis is not conclusive since there is the possibilitythat the high vibration noise of the enclosure is too high compared to the failureassociated with the bearing. Finally, as future techniques, the analysis of lubricant andacoustic emissions is proposed as well as the characterization of the loads and the historicalfaults within the observatory.
Description
Catalogado desde la version PDF de la tesis.
Keywords
ANALISIS DE KURTOSIS , KURTOGRAMAS , VIBRACION EN BAJA VELOCIDAD