Analysis of the influence of crack location for diagnosis in rotating shafts based on 3 x energy

Abstract

The aim of condition monitoring is to detect faults before a catastrophic failure occurs. Cracks in rotating shafts are especially critical. The present work studies vibration signals obtained from a rotating shaft under different crack depths and locations. Tests were performed in a rig called Rotokit at steady state at different rotation speeds. Signals obtained are analyzed by means of energy using the Wavelet Theory, specifically the Wavelet Packets Transform. Nine crack depths in the shafts were tested, from 4% to 50% of the shaft diameter. Previous related work showed good reliability for crack diagnosis using 3 x energy for cracks in the middle section. In the present work, previous results are compared to the obtained for a crack in a change of section at one side. In both crack locations, large changes in energy are observed at 3 x at high speeds. Energy levels at this harmonic were used for the inverse process of crack detection, and Probability of detection curves were calculated by thresholding. Cracks with depths above 12% can be detected with reliability in the locations tested using this method.

Wavelet Packets Transform processing and Genetic Neuro-Fuzzy classification to detect faulty bearings

“…In the laboratory, vibration signals of healthy and faulty mechanical elements were measured and a signal processing was accomplished, particularly a WPT, as it was explained in previous sections. Images of the data used and the first processing corresponding with the spectra PSD can be seen in Castejo´n et al 22 and the energy of the wavelet packet decomposition can be tested in Go´mez et al 46 The application of the discrete WT recursively allows for obtaining an adequate number of patterns which describe the dynamic behaviour of the signal. In this study, the level of decomposition is 3 and the mother wavelet selected is DB6; therefore, after this process, there are 1872 data sets, which means 1872 characteristic energy vectors.…”

Crack Detection in Rotating Shafts Using Wavelet Analysis, Shannon Entropy and Multi-class SVM

“…In addition, a method combining WPT and Empirical Mode Decomposition (EMD) was proposed by Bin et al [9] to precisely characteristic features of crack defects in specific frequency bands. Similarly, features related to WPT have been applied in [10], [11]. By using WPT, it is great convenient to decompose a signal into approximation and detail information having the same frequency resolution.…”

“…A grid search method is applied to search best parameters, namely the cost parameter c and the width parameter g in the training phrase. Here, c and g are respectively set between 2 −10 to 2 10 . In addition, after the data training, a 5-fold cross-validation method is used for the validation of the proposed rotating shaft fault diagnostic approach.…”

“…Crack depth and growth were analyzed and identified using envelop analysis and DWT based on acoustic emission in [8]. Apart from CWT and DWT, WPT was also adopted with time-frequency features for crack detection in [9][10][11], which benefits from appropriately decomposing frequency spectrum into the same bands with detail and approximate coefficients.…”

Reducing the Frame Vibration of Delta Robot in Pick and Place Application: An Acceleration Profile Optimization Approach

“…A lot of model and signal based methods have been proposed to solve the fault detection problem for machines [18][19][20] and robots [21,22]. In this paper, the vibration signals are used to find the optimal dynamic motion parameters.…”