Marcelo Perez-Alonso

Practical Aspects of Mixed-Eccentricity Detection in PWM Voltage-Source-Inverter-Fed Induction Motors

In this paper, we present the results of an ongoing investigation relating to the detection of static and dynamic eccentricity in cage induction motors fed by a pulsewidth-modulation frequency converter using line-current spectral analysis. Two different motors fed by different supply sources (utility voltage and two different voltage converters at different assigned frequencies) were tested. A statistical analysis of the results obtained was carried out. These results allow us to present practical conclusions relating to the detection of mixed eccentricity.

Transient detection of close components through the chirplet transform: Rotor faults in inverter-fed induction motors

Up to now, detection of rotor faults in inverter-fed induction motors has received very little attention. This fault is difficult to be detected, since the fault-related components are too close to the fundamental (the inverter usually operates at low slip). Moreover, classic techniques cannot be applied since steady states are not common. The causes of this type of fault are analyzed in the paper, showing its importance. Particularly, cases of real faults in electric traction are exposed. Then, the paper explores the use of linear time-frequency transforms to detect the time-frequency evolution of the fault-related components. It is shown how the most common linear transforms, such as the Short Time Fourier Transform, do not enable the fault detection. The Chirplet Transform (which has never been used for diagnosing purposes), is proposed to obtain the components evolutions, even if they are too close in the time-frequency plane. The technique is validated through startup tests, in which the presence of the fault is quantified when analyzing the stator current.

Assessment of the lubrication condition of a bearing using spectral analysis of the stator current

Bearing lubrication is very important to ensure a satisfactory and long operation of bearings. An excess of lubrication can be as damaging as a lack of lubrication. An excess of oil or grease has damaging effects in the short term such as difficult heat evacuation, sliding balls and greater current consumption. As MCSA is widely recognized as a useful and reliable tool for condition monitoring of induction motors, we have used it to identify signs in the stator current spectrum of an excess of lubrication.

Bearing lubrication assessment using an statistical analysis of the stator current spectrum

Bearing lubrication is very important to ensure a satisfactory and long operation of bearings. An excess of lubrication can be as damaging as a lack of lubrication. An excess of oil or grease has damaging effects in the short term such as difficult heat evacuation, sliding balls and greater current consumption. There are some procedures, such as acoustic monitoring and methods based in charts provided by manufacturers, to decide when to regrease a bearing and how much grease is needed. As Motor Current Signature Analysis is widely recognized as a useful and reliable tool for condition monitoring of induction motors, it has been used, along with an Analysis of Covariance, to find signs in the stator current spectrum of an excess of lubrication. The statistical analysis shows that the variation observed in some integer harmonics of the supply frequency is due to different levels of bearing lubrication.

Experimental identification of induction motor broken bar characteristic frequencies using a programmable power source

The introduction of Voltage Source Inverter fed motors has produced significant changes in the field of diagnostics being necessary further research to overcome various challenges. To enhance the diagnosis of motor faults it is advisable to consider as many fault signatures as possible, so, in this paper it is analyzed the detection of broken bars using the spectra around harmonics introduced by the power supply. Since the harmonic content of the voltage waveform and the amplitude of the harmonics depend on the voltage supply, in order to establish diagnosis criteria that are independent of the voltage supply, an induction motor has been tested being supplied by a Programmable Power Source, which has allowed to completely control the harmonic content of the input voltage waveform. The results are discussed and some guidelines for condition monitoring are suggested.

Some practical aspects of eccentricity detection in induction motors for different supply conditions

In this paper, we advance results of an ongoing investigation related with the detection of static and dynamic eccentricity in cage induction motors fed by frequency converter using line current spectral analysis. A statistical analysis of the results obtained with the laboratory motors for different supply conditions has been carried out. These results allow us to present some preliminary conclusions.

Identification of eccentricity related-faults characteristic frequencies in induction motors using a programmable power source

With the aim of developing an expert system for condition monitoring of cage rotor induction motors fed by power inverters, laboratory tests have been performed on a motor with artificially created eccentricity in order to identify as many fault signatures as possible. The technique used to detect faults is Motor Current Signature Analysis and in this paper it is analyzed if the spectra around the time harmonics 5,roduced by an inverter are suitable for condition monitoring. monitoring. But, taking into account that the harmonic content of the voltage waveform and the amplitude of the harmonics depend on the voltage supply, in order to establish diagnosis criteria independent of the voltage supply, a programmable power source has been used as a motor supply. The results are discussed and some guidelines for condition monitoring are suggested.