Loading...
Publication
Análise e diagnóstico não invasivo de avarias em rolamentos de motores de indução trifásicos
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 5.11 MB | Adobe PDF |
Authors
Abstract(s)
O presente trabalho de dissertação foca-se no diagnóstico de avarias em rolamentos, componentes mecânicos essenciais nas máquinas elétricas rotativas, destacando-se o motor de indução como uma das máquinas elétricas mais utilizadas a nível industrial devido às suas características operacionais, tais como simplicidade, elevado rendimento e facilidade de manutenção. Nos motores de baixa e média potência, a maior taxa de avarias ocorre nos rolamentos, e para o seu diagnóstico, propõe-se a implementação de métodos não invasivos, onde a corrente elétrica do estator se apresenta como uma das opções mais viáveis. Para esta análise, foram aplicadas duas técnicas: a Short-Time Fourier Transform (STFT) e o Extended Park’s Vector Approach (EPVA), ambas fundamentadas no domínio da frequência, sendo que a STFT permite, adicionalmente, uma análise no domínio do tempo. Com o objetivo de otimizar os resultados do diagnóstico, propõe-se uma abordagem híbrida que combina as vantagens das duas técnicas, permitindo uma avaliação precisa do comportamento da frequência ao longo do tempo e a identificação do tipo de avaria que afeta o rolamento. A eficácia destas técnicas, bem como a sua combinação, foi validada através de ensaios num motor de indução trifásico de 2,2 kW, equipado com rolamentos que apresentam avarias típicas, tais como o pitting. Os resultados foram satisfatórios, mostrando que cada método foi capaz de detetar sintomas provocados pelas avarias, como o aumento de energia nos primeiros harmónicos e distorções ao longo do tempo nas frequências caraterísticas do rolamento. Este estudo reforça a importância do diagnóstico não invasivo, que oferece vantagens significativas como simplicidade, baixo custo e utilidade em aplicações de difícil acesso, onde os métodos convencionais de diagnóstico não seriam viáveis.
This dissertation focuses on the diagnosis of bearing failures, essential mechanical components in rotating electrical machines, highlighting the induction motor as one of the most widely used electrical machines at an industrial level due to its operational characteristics, including simplicity, high efficiency, and ease of maintenance. In low and medium power motors, the highest failure rate is found in the bearings, and for their diagnosis, the implementation of non-invasive methods is proposed, where the stator's electric current is presented as one of the most viable options. For this analysis, two techniques were applied: the Short-Time Fourier Transform (STFT) and the Extended Park’s Vector Approach (EPVA), both based on the frequency domain, with the STFT additionally allowing analysis in the time domain. To optimize the diagnostic results, a hybrid approach is proposed, combining the strengths of both techniques, allowing for an accurate evaluation of frequency behavior over time and the identification of the specific type of fault affecting the bearing. The effectiveness of these techniques, as well as their combination, was validated through tests on a 2.2 kW three-phase induction motor, equipped with bearings exhibiting typical faults such as pitting. The results were satisfactory, showing that each method successfully detected symptoms caused by the faults, including increased energy in the first harmonics and time-varying distortions in the bearing’s characteristic frequencies. This study highlights the importance of non-invasive diagnostics, offering significant advantages such as simplicity, low cost, and applicability in hard-to-reach environments, where conventional diagnostic methods would not be feasible.
This dissertation focuses on the diagnosis of bearing failures, essential mechanical components in rotating electrical machines, highlighting the induction motor as one of the most widely used electrical machines at an industrial level due to its operational characteristics, including simplicity, high efficiency, and ease of maintenance. In low and medium power motors, the highest failure rate is found in the bearings, and for their diagnosis, the implementation of non-invasive methods is proposed, where the stator's electric current is presented as one of the most viable options. For this analysis, two techniques were applied: the Short-Time Fourier Transform (STFT) and the Extended Park’s Vector Approach (EPVA), both based on the frequency domain, with the STFT additionally allowing analysis in the time domain. To optimize the diagnostic results, a hybrid approach is proposed, combining the strengths of both techniques, allowing for an accurate evaluation of frequency behavior over time and the identification of the specific type of fault affecting the bearing. The effectiveness of these techniques, as well as their combination, was validated through tests on a 2.2 kW three-phase induction motor, equipped with bearings exhibiting typical faults such as pitting. The results were satisfactory, showing that each method successfully detected symptoms caused by the faults, including increased energy in the first harmonics and time-varying distortions in the bearing’s characteristic frequencies. This study highlights the importance of non-invasive diagnostics, offering significant advantages such as simplicity, low cost, and applicability in hard-to-reach environments, where conventional diagnostic methods would not be feasible.
Description
Keywords
Avarias em Rolamentos Correntes Elétricas Extended Park’S Vector Approach (Epva) Motor de Indução Trifásico Short-Time Fourier Transform (Stft)