Riccardo Rubini

Diagnosis of Bearing Faults in Induction Machines by Vibration or Current Signals: A Critical Comparison

Early diagnosis of faults in induction machines is an extensively investigated field, for cost and maintenance savings. Mechanical imbalances and bearing faults account for a large majority of faults in a machine, especially for small-medium size machines. Therefore their diagnosis is an intensively investigated field or research. Recently many research activities were focused on the diagnosis of bearing faults by current signal. Stator current components are generated at predictable frequencies related to the electrical supply and mechanical frequencies of bearing faults. However their detection is not always reliable, since the amplitude of fault signatures in the current signal is very low. This paper compares the bearing fault detection capability obtained with vibration and current signals. To this aim a testbed is realized that allows to test vibration and current signal on a machine with healthy or faulty bearings. Signal processing techniques for both cases are reviewed and compared in order to show which procedure is best suited to the different type of bearing faults. The paper contribution is the use of a simple and effective signal processing technique for both current and vibration signals, and a theoretical analysis of the physical link between faults and current components including torque ripple effects. As expected because of the different nature of vibration and current, bearing fault diagnosis is effective only for those fault whose mechanical frequency rate is quite low. Experiments are reported that confirm the proposed approach.

Diagnosis of Bearing Faults of Induction Machines by Vibration or Current Signals: A Critical Comparison

Mechanical imbalances and bearing faults account for a large majority of the faults in a machine, particularly for small-medium size machines. Therefore, their diagnosis is an intensively investigated field of research. Recently, many research activities were focused on the diagnosis of bearing faults by current signals. This paper compares the bearing fault detection capability obtained with the vibration and current signals. The paper contribution is the use of a simple and effective signal processing technique for both current and vibration signals, and a theoretical analysis of the physical link between faults, modeled as a torque disturbance, and current components. The focus of the paper is on the theoretical development of the correlation between torque disturbances and the amplitude of the current components, together with a review of fault models used in the literature. Another contribution is the re-creation of realistic incipient faults and their experimental validation. Radial effects are visible only in case of large failures that result in air-gap variations. Experiments are reported that confirm the proposed approach.

Detection of Generalized-Roughness Bearing Fault by Spectral-Kurtosis Energy of Vibration or Current Signals

Generalized roughness is the most common damage occurring to rolling bearings. It produces a frequency spreading of the characteristic fault frequencies, thus making it difficult to detect with spectral or envelope analysis. A statistical analysis of typical bearing faults is proposed here in order to identify the spreading bandwidth related to specific conditions, relying on current or vibration measurements only. Then, a diagnostic index based on the computation of the energy in the previously defined bandwidth is used to diagnose bearing faults. The proposed method was validated experimentally with vibration signals, with robust and reliable results. The same procedure can be extended to current signals.

Bearing Fault Model for Induction Motor With Externally Induced Vibration

This paper investigates the relationship between vibration and current in induction motors operated under external vibrations. Two approaches are usually available to define this relationship. The former is based on airgap variations, while the latter is based on torque perturbation. This paper is focused on the airgap variation model. The ball bearing fault is modeled by contact mechanics. External vibrations often occur in many industrial applications where externally induced vibrations of suitable amplitude cause cyclic radial loading on the machine shaft. The model is validated by experiments, owing to a dedicated test setup, where an external vibration source (shaker) was employed, together with ball bearing alterations in order to decrease the stiffness of the support along the radial direction. To maximize the effects of externally induced vibrations, the frequency chosen was near the flexural resonance of the rotor (determined by finite-element method analysis). The direction of the external vibration is radial with respect to the axis of the electric machine under test. During tests, both stator phase currents and vibration of the machine were sampled. The test setup allowed one to vary the machine speed and load, vibration amplitude, and bearing stiffness (damage level). Radial effects are usually visible only in the case of large failures that result in significant airgap variations, as confirmed by experiments.

Fault Detection of Linear Bearings in Brushless AC Linear Motors by Vibration Analysis

Electric linear motors are spreading in industrial automation because they allow for direct drive applications with very high dynamic performances, high reliability, and high flexibility in trajectory generation. The moving part of the motor is linked to the fixed part by means of linear bearings. As in many other electric machines, bearings represent one of the most vulnerable parts because they are prone to wear and contamination. In the case of linear roller bearings, this issue is even more critical as the rail cannot be easily fully enclosed and protected from environmental contamination, unlike the radial rotating bearing counterpart. This paper presents a diagnostic method based on vibration analysis to identify which signature is related to a specific fault.

STFT Based Approach for Ball Bearing Fault Detection in a Varying Speed Motor

This paper focuses on the diagnostics of ball bearings in direct-drive motors. These specific AC brushless motors are increasing their importance in automation machineries because they can work with a built-in flexibility. In particular the angular displacement of the shaft is continuously monitored by an embedded encoder while the control system allows to perform complex motion profiles such as polynomial ones, even with the inversion of the rotating direction. Direct-drive motors avoid the presence of a mechanical cams or gearboxes between the motor and the load with a subsequent money-saving. On the other side, unfortunately, the diagnostics of ball bearing in those motors is not trivial. In fact most of the solutions proposed in the literature require a constant frequency rotation of the shaft since the characteristic fault frequencies are directly proportional to speed of the motor. It follows that in a varying speed application the fault characteristic frequencies change instantaneously as the rotational frequency does. In this paper an industrial application is considered, where the direct drive motors are used in the kinematic chain of an automated packaging machine performing a cyclic polynomial profile. The basic idea is to focus on signal segmentation using the position profile of the shaft – directly measured by the encoder – as trigger. Next the single cycles of the machine is analysed in time domain, again using encoder signal machine contribution is deleted. Feature extraction for damage detection is done by applying the Short Time Fourier Transform (STFT), the STFT for each cycle is averaged in time-frequency domain in order to enhance fault signature. Finally, the sum of STFT coefficients is used as a simple indicators of damage.

Diagnosis of mechanical faults by spectral kurtosis energy

Generalized roughness is the most common damage occurring to roller bearing. It produces a frequency spreading of the characteristics fault frequencies, thus being difficult to detect with spectral or envelope analysis. A statistical analysis of typical bearing faults is here proposed in order to identify the spreading bandwidth related to a specific conditions, relying on current measurements only. Then a diagnostic index based on the computation of the energy in the above defined bandwidth is used to diagnose bearing faults. The proposed method was validated experimentally with vibration signals, with robust and reliable results. Subsequently it has been applied to stator currents monitoring.

Kurtosis over Energy Distribution Approach for STFT Enhancement in Ball Bearing Diagnostics

This paper focuses on the diagnostics of ball bearings under time varying speed conditions. Compared to classical demodulation techniques, time-frequency approach allows to take into account transient occurrence or non-stationary phenomena along the timeline. Among the different time-frequency approaches available the simplest is the Short Time Fourier Transform (STFT). From a practical point of view, its implementation in an industrial environment has a main drawback: the industry usually needs a scalar value as output (like a semaphore: green, yellow and red light) to assess the bearing condition, while time-frequency approaches produce a bi-dimensional map that needs to be interpreted. The authors suggest to combine the information gathered by spectral kurtosis and energy distribution for the automatic selection of a filtering band that could extract from the STFT map the most informative component in time domain, reducing the complexity of the output to a mono-dimensional vector. A simple check if the output exceed a given threshold can then be used to obtain a scalar value.

Currents and vibrations in asynchronous motor with externally induced vibration

This paper presents an experimental investigation of the airgap variation model for vibration and current harmonics relationship in induction motors. To this aim, an external vibration source was employed, together with ball bearing alterations in order to decrease stiffness. The direction of the external vibration is radial with respect to the axis of the electric machine under test. To maximize the effect of externally induced vibrations, the frequency chosen was near the flexural resonance of the rotor, determined by FEM analysis. During tests both currents and vibration of the machine were acquired. The test rig allowed to vary speed, vibration level and bearing stiffness. An electromagnetic brake provided a variable output load for the electric machine. The focus of the paper is the review of fault models used in literature. Radial effects are usually visible only in case of large failures that result in air-gap variations, as the experiments confirmed.

Comparison Between Time-Frequency Techniques to Predict Ball Bearing Faults in Drives Executing Arbitrary Motion Profiles

In this paper Wavelet Transform (WT) and Hilbert-Huang Transform (HHT) are used as bearing diagnostics tools in drives executing arbitrary motion profiles. This field is increasingly drawing the attention of the industries because the modern electric motors work as electric cams inducing the shaft to move with a cyclic variable-velocity profile. The literature papers take into account only a constant velocity profile and they are not suitable for such applications. In fact literature methods analyse the signal only in the frequency domain, while in variable-velocity condition the bearing diagnostics should be performed in time domain. Both WT and HHT are time-frequency techniques which describe an input signal as a sum of specific functions. These functions are compared with a signal which simulates the expected vibrations of a bearing with a given fault, e.g. on the outer race. The comparison is done through a cross-correlation between the expected signal and the time-frequency techniques output. WT and HHT are used separately in an industrial case, which consists in bearing fault prediction in an automated packaging machine. In the end of the paper the WT and HHT results are discussed to analyse the different responses.Copyright