Search

[ES] En este Trabajo Fin de Master se pretende comprobar la validez de la técnica basada en el análisis de flujo de dispersión para el diagnóstico de averías electromecánicas en motores asíncronos. La técnica de diagnóstico basada en análisis de flujo, si bien conocida desde hace años, ha concitado reciente interés en la industria debido al progresivo abaratamiento de los sensores de flujo y a las ventajas inherentes de la misma (simplicidad, carácter no invasivo ). Sin embargo, la literatura relativa a su aplicación con fines de diagnóstico es todavía limitada y su uso se ha visto relegado hasta la fecha a aplicaciones concretas en régimen estacionario. El objetivo de este Trabajo Fin de Master es profundizar en la aplicación de la técnica en lo que respecta a su aplicación para la detección de fallos en el rotor y otras averías electromecánicas. El objeto es no solamente aplicar técnicas basadas en el análisis de señales en régimen permanente, sino también extender el análisis al estudio de señales en régimen transitorio, tales como señales capturadas durante el arranque. A tal efecto, tras una revisión de la literatura técnica en el área, se pretenden desarrollar múltiples ensayos de laboratorio en motores con diferentes niveles de avería en el rotor, así como motores con otros posibles fallos. En cada ensayo, se registrará la señal de fuerza electromotriz inducida en una bobina externa colocada en diferentes posiciones sobre la carcasa del motor. Las señales se registrarán tanto durante el arranque como en régimen permanente. Las señales capturadas se analizarán posteriormente con herramientas aptas para cada régimen (transformada de Fourier para permanente y transformadas wavelet en régimen transitorio). Posteriormente se interpretarán los resultados obtenidos, evaluando la fiabilidad de cada análisis y la idoneidad de cada posición estudiada., [EN] The aim of this Final Master's Thesis is to verify the validity of the technique based on stray flux analysis for the diagnosis of electromechanical faults in asynchronous motors. The diagnostic technique based on magnetic flux analysis, although well known for years, has drawn recent interest in the industry due to the progressive decrement of the cost of flux sensors as well as to the inherent advantages of the technique itself (simplicity, non-invasive nature ...). However, the literature related to its application for diagnostic purposes is still limited and its use has been relegated to specific applications in steady state. The objective of this Master's Thesis is to deepen the application of the technique with regard to its application for the detection of rotor failures and other electromechanical failures. The object is not only to apply techniques based on the analysis of signals in steady state, but also to extend the analysis to the study of signals in transient regime, such as signals captured during startup. To this end, after a review of the technical literature in the area, we intend to develop multiple laboratory tests on motors with different levels of rotor failure, as well as motors with other possible failures. In each test, the induced electromotive force signal will be captured in an external coil attached to different positions on the motor housing. Signals will be recorded during both start-up and steady-state. The captured signals will be analyzed later with tools that are suitable for each regime (Fourier transform for steady-state and wavelet transforms for transient regime). Subsequently, the results obtained will be interpreted, evaluating the reliability of each analysis and the suitability of each position studied.

[ES] El objeto de este trabajo es estudiar la efectividad de las técnicas basadas en análisis de flujo de dispersión para la detección de asimetrías en el devanado del rotor en máquinas asíncronas de rotor bobinado. Estas máquinas se emplean en procesos industriales con requerimientos de par de arranque elevados y es frecuente encontrarlas en el rango de altas potencias. La avería que se considera en el proyecto es una de las más habituales en este tipo de máquinas y su presencia puede acarrear consecuencias muy negativas para la máquina. El trabajo se centrará en el análisis del flujo de dispersión durante el arranque de la máquina. A tal fin se emplearan herramientas especiales de análisis de señal (transformadas tiempo-frecuencia) y se relacionarán los resultados de la aplicación de las mismas con el nivel de avería en la máquina. Para el desarrollo del proyecto, se realizarán múltiples ensayos con máquinas de laboratorio en los que se registrará durante el arranque la señal de fuerza electromotriz inducida en una bobina externa, situada en diferentes posiciones de la carcasa de la máquina, para diferentes niveles de avería en el rotor. Posteriormente, estas señales se analizarán con una herramienta tiempo-frecuencia, la transformada wavelet, y se interpretarán los resultados obtenidos, evaluando la fiabilidad de cada tipo de análisis. Se compararán los resultados con los obtenidos con las herramientas clásicas de diagnóstico, basadas en el análisis en régimen permanente., [EN] The purpose of this work is to study the effectiveness of the techniques based upon the stray flux analysis for the detection of rotor winding asymmetries in wound rotor induction motors. These machines are commonly used in industrial processes with high starting torque requirements. They are often found in the high-power range. The fault considered in the project is one of the most common in this type of machines, and its presence might have very negative consequences for the machine. The work will focus on the stray flux analysis during the start-up of the machine. Special signal analysis tools (time-frequency transforms) will be used and their results will be related to the condition of the machine. For the development of the project, multiple tests will be developed with laboratory machines in which the electromotive force induced in an external coil sensor will be registered during the start-up of the machine, locating the sensor at different positions of the motor frame, for different levels of damage in the rotor winding. Afterwards, these signals will be analyzed with a time-frequency tool, the wavelet transform, and the results will be interpreted, evaluating the reliability of each analysis. The results will be compared to the classic diagnostic tools, based upon the permanent regime analysis., [CA] L’objecte d’aquest treball és estudiar la efectivitat de les tècniques basades en anàlisi de flux de dispersió per a la detecció d’asimetries en el debanat del rotor en màquines asíncrones de rotor bobinat. Aquestes màquines s’utilitzen en processos industrials amb requeriments de par d’arrancada elevats i és freqüent trobar-les en el rang d’altes potències. L’avaria considerada en el projecte és una de les més habituals en aquest tipus de màquines i la seua presència pot comportar conseqüències molt negatives per a la màquina. El treball es centra en l’anàlisi del flux de dispersió en l’arrancada de la màquina. Així, s’utilitzen ferramentes especials d’anàlisi de senyal (transformades temps-freqüència) i es relacionen els resultats de l’aplicació de les mateixes amb el nivell d’avaria de la màquina. Per al desenvolupament del projecte, es realitzen múltiples assajos amb màquines de laboratori en els quals es registra en l’arrancada la senyal de força electromotriu induïda en una bobina externa, situada en diferents posicions de la carcassa de la màquina. Posteriorment, aquestes senyals s’analitzen amb una ferramenta temps-freqüència, la transformada wavelet, i s’interpreten els resultats obtinguts, avaluant la fiabilitat de cada tipus d’anàlisi.

[EN] The reliable assessment of the rotor condition in induction motors is a matter of increasing concern in the industry. Although rotor damages are more likely in line-started motors operating under high inertias, some cases of broken rotor bars in motors supplied via soft starters have been also reported in the industry. Motor current signature analysis (MCSA) is the most widely spread approach to diagnose such failures. However, its serious drawbacks in many real industrial applications have encouraged investigation on alternative methods enhancing the reliability of the diagnosis. This paper extends a recently introduced diagnosis methodology relying on the startup current analysis to the case of soft-starter-operated motors. The approach has proven to provide very satisfactory results, even in cases where the classical MCSA does not lead to correct diagnosis conclusions. However, its extension to operation under soft starters was still a pending issue. The experimental results shown in this paper ratify the validity of the proposed diagnosis approach in soft-starteroperated induction motors., This work was supported by the Spanish “Ministerio de Economía y Competitividad” (MINECO) in the framework of the “Proyectos I+D del Subprograma de Generación de Conocimiento, Programa Estatal de Fomento de la Investigación Científica y Técnica de Excelencia” under Grant DPI2014-52842-P

To enable further growth of wind turbine dimensions and rated power, it is essential to decrease structural loads that wind turbines experience. Therefore a great portion of research is focused on control algorithms for reduction of wind turbine structural loads, but typically wind turbine rotor is considered to be perfectly symmetrical, and therefore such control algorithms cannot reduce structural loads caused by rotor asymmetries. Furthermore, typical approach in the literature is to use blade load measurements, especially when higher harmonics of structural loads are being reduced. In this paper, improvements to standard approach for reduction of structural loads are proposed. First, control algorithm capable of reducing structural loads caused by rotor asymmetries is developed, and then appropriate load transformations are introduced that enable presented control algorithms to use load measurements from various wind turbine components. Simulation results show that proposed control algorithm is capable of reducing structural loads caused by rotor asymmetries.

[EN] This paper introduces a speed-sensorless method for detecting rotor asymmetries in wound rotor induction machines working under nonstationary conditions. The method is based on the time-frequency analysis of rotor currents and on a subsequent transformation, which leads to the following goals: unlike conventional spectrograms, it enables to show the diagnostic results as a simple graph, similar to a Fourier spectrum, but where the fault components are placed always at the same positions, regardless the working conditions of the machine; moreover, it enables to assess the machine condition through a very small set of parameters. These characteristics facilitate the understanding and processing of the diagnostic results, and thus, help to design improved monitoring and predictive maintenance systems. Also these features make the proposed method very suitable for condition monitoring of wind power generators, because it fits well with the usual non stationaryworking conditions ofwind turbines, and makes feasible the transmission of significant diagnostic information to the remote monitoring center using standard data transmission systems. Simulation results and experimental tests, carried out on a 5-kW laboratory rig, show the validity of the proposed method and illustrate its advantages regarding previously developed diagnostic methods under nonstationary conditions., This work was supported by the Spanish Ministerio de Economia y Competitividad in the framework of the Programa Estatal de Investigacion, Desarrollo e Innovacion Orientada a los Retos de la Sociedad under Project (DPI2014-60881-R).

[EN] The main objective of this paper is to diagnose the presence of combined faults in induction machines. For this purpose, a methodology based on the application of the Discrete Wavelet Transform (DWT) to the stator startup current is used. This approach was applied in previous works with success to the diagnosis of rotor asymmetries and mixed eccentricities in motors with different sizes and conditions. However, as most of the diagnosis methods hitherto developed, the application of the proposed approach was circumscribed to situations in which a single fault was present in the machine. In addition, the influence of other phenomena such as load torque oscillations or voltage fluctuations was studied, but without considering the combination of these phenomena and the fault in the machine. This work is intended, first, to apply the proposed transient-based methodology to several cases in which different faults (rotor asymmetries, mixed eccentricities and inter-turn and inter-coil stator short-circuits) are simultaneously present in the machine and, second, to apply it to cases regarding faults combined with other phenomena making difficult the diagnosis, such as load torque oscillations. Interesting considerations regarding the preponderance of the effects of some of the faults are also done in the paper. The application of the methodology is focused on induction machines with stator parallel branches; in this sense, the suitability of the use either of the phase current or of the branch current for the diagnosis of each particular fault is analysed. The results look promising with regard to the validity of the methodology for the reliable discrimination of simultaneous electromechanical faults and the diagnosis of faults combined with other phenomena. (C) 2009 Elsevier Ltd. All rights reserved., This work was supported by the Conselleria d'Educacio, Formacio i Ocupacio of the Generalitat Valenciana, in the framework of the "Ayudas para la Realizacion de Proyectos de I + D para Grupos de Investigacion Emergentes", project reference GV/2012/020.

[EN] There exists no single quantity able to diagnose all possible failures taking place in induction motors. Currents and vibrations monitoring are rather common in the industry, but each of these quantities alone can only detect some specific failures. Moreover, even for the specific faults that a quantity is supposed to detect, many problems may rise. As a consequence, a reliable and general diagnosis system cannot rely on a single quantity. On the other hand, it would be desirable to rely on quantities that can be measured in a noninvasive way, which is a crucial requirement in many industrial applications. This paper proposes a twofold method to detect electromechanical failures in induction motors. The method relies on analysis of currents (steady state + transient) combined with analysis of infrared data captured by using appropriate cameras. Each of these noninvasive techniques may provide complementary information that may be very useful to diagnose an enough wide range of failures. In the present paper, the detection of three illustrative faults is analyzed: broken rotor bars, cooling system problems and bearing failures. The results show the potential of the methodology that may be particularly suitable for large, expensive motors, where the prevention of eventual failures justifies the costs of such system, due to the catastrophic implications that these unexpected faults may have.

To enable further growth of wind turbine dimensions and rated power, it is essential to decrease structural loads that wind turbines experience. Therefore a great portion of research is focused on control algorithms for reduction of structural loads, but typically wind turbine rotor is considered to be perfectly symmetrical, and therefore such control algorithms cannot reduce structural loads caused by rotor asymmetries. In this paper, the reduction of wind turbine structural loads caused by rotor asymmetries is discussed. To this aim, suitable transformation of structural loads and control algorithm based on such transformation are proposed. Simulation results show that proposed control algorithm is capable of reducing structural loads caused by rotor asymmetries.

[EN] In this paper, a cutting-edge time-frequency decomposition tool, i.e., the Hilbert-Huang transform (HHT), is applied to the stator startup current to diagnose the presence of rotor asymmetries in induction machines. The objective is to extract the evolution during the startup transient of the left sideband harmonic (LSH) caused by the asymmetry, which constitutes a reliable evidence of the presence of the fault. The validity of the diagnosis methodology is assessed through several tests developed using real experimental signals. Moreover, in this paper, an analytical comparison with an alternative time-frequency decomposition tool, i.e., the discrete wavelet transform (DWT), is carried out. This tool was applied in previous works to the transient extraction of fault-related components, with satisfactory results, even in cases in which the classical Fourier approach does not lead to correct results. The results of the application of the HHT and DWT are analyzed and compared, obtaining novel conclusions about their respective suitability for the transient extraction of asymmetry-related components, as well as the equivalence, with regard to the LSH extraction, between their basic components, namely: 1) intrinsic mode function, for the HHT, and 2) approximation signal for the DWT., This work was supported in part by the Spanish “Ministerio de Educación y Ciencia,” in the framework of the “Programa Nacional de proyectos de Investigación Fundamental,” project reference DPI2008-06583/DPI and in part by “Vicerrectorado de Investigación, Desarrollo e Innovación of the Universidad Politécnica de Valencia” through the Programa de Apoyo a la Investigación y Desarrollo under Contract PAID-06-07.