Byeong Keun Choi

Rolling element bearing fault detection using acoustic emission signal analyzed by envelope analysis with discrete wavelet transform

Acoustic Emission (AE) what is a non-destructive testing technique is widely used for the early detection of faults in rotating machine in these days, because its sensitivity is higher than normal accelerometers and it can detect low energy vibration signals. The faults in rotating machines are generally occurred at bearing and/or gearboxes which are one of the principal parts of the machines. To detect the bearing fault, envelope analysis was studied and presented for several decade years. And the researches spoke that AE has a possibility of the application in condition monitoring system using envelope analysis for the rolling element bearing. And the peak ratio (PR) was developed for expression of the bearing condition in condition monitoring system using AE. The noise level is needed to reduce to get exact PR value because PR is calculated from total root mean square (RMS) and the harmonics of the defect frequencies. Therefore, in this paper, the discrete wavelet transform (DWT) was added in envelope analysis to reduce the noise level in AE signals. And then, the PR was calculated and compared with general envelope analysis result and the result of envelope analysis added DWT.

Robust Optimal Positioning of Strain Gages on Blades

This paper focuses on the formulation and validation of an automatic strategy for the selection of the locations and directions of strain gages to capture at best the modal response of a blade in a series of modes. These locations and directions are selected to render the strain measurements as robust as possible with respect to random mispositioning of the gages and gage failures. The approach relies on the evaluation of the signal-to-noise ratios of the gage measurements front finite element strain data and includes the effects of gage size. A genetic algorithm is used to find the strain gage locations-directions that lead to the largest possible value of the smallest modal strain signal-to-noise ratio, in the absence of gage failure, or of its expected value when gage failure is possible. A fan blade is used to exemplify the applicability of the proposed methodology and to demonstrate the effects of the essential parameters of the problem, i.e., the mispositioning level, the probability of gage failure, and the number of gages.

Development of Rotordynamics Program Based on the 2D Finite Element Method for Flywheel Energy Storage System

Flywheel energy storage system (FESS) is defined as a high speed rotating flywheel system that can save surplus electric power. The FESS is proposed as an efficient energy storage system because it can accumulate a large amount of energy when it is operated at a high rotating speed and no mechanical problems are encountered. The FESS consists of a shaft, flywheel, motor/generator, bearings, and case. It is difficult to simulate rotor dynamics using common structure simulation programs because these programs are based on the 3D model and complex input rotating conditions. Therefore, in this paper, a program for the FESS based on the 2D FEM was developed. The 2D FEM can model easier than 3D, and it can present the multi-layer rotor with different material each other. Stiffness changing of the shaft caused by shrink fitting of the hub can be inputted to get clear solving results. The results obtained using the program were compared with those obtained using the common programs to determine any errors.

Robust optimal positioning of strain gauges on blades

This paper focuses on the formulation and validation of an automatic strategy for the selection of the locations and directions of strain gauges to capture at best the modal response of a blade in a series of modes. These locations and directions are selected to render the strain measurements as robust as possible with respect to random mispositioning of the gauges and gauge failures. The approach relies on the evaluation of the signal-to-noise ratios of the gauge measurements from finite element strain data and includes the effects of gauge size. A genetic algorithm is used to find the strain gauge locations-directions that lead to the largest possible value of the smallest modal strain signal-to-noise ratio, in the absence of gauge failure, or of its expected value when gauge failure is possible. A fan blade is used to exemplify the applicability of the proposed methodology and to demonstrate the effects of the essential parameters of the problem, i.e. the mispositioning level, the probability of gauge failure, and the number of gauges.© 2002 ASME

Feature Analysis for Fault Diagnosis according to Gearbox Failure

In the development of a fault diagnosis and condition monitoring in the gearbox, the research is a quantitative analysis and a test of the effect of gear damage on the vibration of the gearbox. The lab-scale gearbox test device that builds the several types of fault such as gear tooth breakage, misalignment and looseness occurred by gearbox fault simulator. This paper presents feature analysis through the PCA (principal component analysis), GA (genetic algorithm) and SVM (support vector machine) of machine learning, the performance of feature classification is evaluated by faults on the gearbox. In addition, the trend of selected features of combined fault indicates clustering at the same point on three dimensions. Therefore, the results of feature-based analysis considering gearbox faults are about 99 %, which verifies the performance for gearbox fault diagnosis.

Signal Processing Technology for Rotating Machinery Fault Signal Diagnosis

The acoustic emission (AE) technique is widely applied to develop early fault detection systems, on which the problem of a signal-processing method for an AE signal is mainly focused. In the signal-processing method, envelope analysis is a useful method to evaluate the bearing problems and the wavelet transform is a powerful method to detect faults occurring on rotating machinery. However, an exact method for the AE signal has not been developed yet. Therefore, in this chapter two methods are given: Hilbert transform and discrete wavelet transform (IEA), and DET for feature extraction. In addition, we evaluate the classification performance with varying the parameter from 2 to 15 for feature selection DET and 0.01–1.0 for the RBF kernel function of SVR; the proposed algorithm achieved 94 % classification accuracy with the parameter of the RBF 0.08, 12 feature selection.

Analysis of Dynamics Characteristics of Water Injection Pump through the 2D Finite Element

After drilling operations at the offshore plant to production to crude oil to high pressure. After that time the low pressured of pipe inside when the secondary produce so oil recovery is reduced. At that time injection sea water at the pipe inside through water injection pump that the device Increase recovery so to be research and development at many industry. So developing 3-stage water injection pump at the domestic company. A variety of mathematical analysis during the detailed design analysis was not made through the dynamics characteristic. In this paper, a 2D finite element analysis is performed through the dynamics of the present study was the validation of the model.

Analysis of Dynamic Characteristics of Water Injection Pump

Water injection pump outputs oil with high pressure during this process, seawater is injected into the well to recover the well pressure and maintain high productivity. A water injection pump has high productivity, and therefore, it serves as a key piece of equipment in marine plants. In this light, water injection pumps are being studied widely in industry. In this study, the rotor dynamics is analyzed to determine the natural frequency according to the bearing stiffness and operation speed change. This study aims to establish the pump reliability through critical speed, stability, and unbalance response analysis.

Eigenvalue Analysis of Roots Type Vacuum Pump Using FE-Method

The demand and importance of vacuum pumps are increasing, because the necessity of dry vacuum pump is proven in semiconductor industry. Various types of vacuum pumps have been developed, and in this research a roots type vacuum pump has been analyzed with a variety of mathematical analysis. Analysis of eigenvalue by FEM analysis and Modal Test were performed through the Roots type vacuum pump and using the results of irregular vibration and instability an assessment has been performed.

Signal Processing Method (HT and HHT) of AE for Complex Fault Detection

In the signal processing method, envelope analysis is a useful method to evaluate the rolling element bearing problems and Wavelet transform is a powerful method to detect faults occurred on gearboxes. However, exact method for AE signal is not developed yet. Therefore, in this paper, two methods, which are Hilbert Transform (HT) and Hilbert-Huang transforms (HHT), will be compared for development a signal processing method for early fault detection system by using AE. HHT has better advantages than HT because HHT can show the time-frequency domain result. But, HHT needs long time to process a signal, which has a lot of data, and has a disadvantage in de-noising filter.