Outer Race Research Articles

IMAGE RECOGNITION FOR BEARING FAULT DETECTION BASED ON CONVOLUTIONAL NEURAL NETWORK

Bearing is an essential component for rotary machinery. The bearing serves as a fixture for position and provides stability for rotation. Bearing failure has detrimental effects on production schedules and operations. Consequently, detecting and diagnosing bearing issues in advance ensures the safety and reliability of rotating equipment systems, which will definitely save production costs and time. Therefore, this paper proposes the use of image recognition based on a convolutional neural network (CNN) for machine fault detection. Recent years have seen the development of deep learning-trained artificial intelligence, which aims to reduce human-induced errors and expenses. Initially, we acquire the vibration signals of the bearing from a test rig under four different conditions. We consider four bearing conditions: healthy bearings, inner race defects, outer race defects, and ball bearing defects. Each of the four conditions is recorded in the vibration time-series data, then converted into spectrogram images before feeding it to the CNN model for training. The performance of the CNN model is based on the comparison of two different models, which are Model A and Model B. Model B is developed based on the performance of Model A, where hyperparameter tuning is implemented to improve the performance. The result shows that the proposed model is capable of detecting and classifying the bearing faults up to 99.9% accuracy.

Non-linear Vibration Analysis of a Ball Bearing Having an Outer Race Defect Using 2-DOF Mathematical Model

Non-linear Vibration Analysis of a Ball Bearing Having an Outer Race Defect Using 2-DOF Mathematical Model

Bearing fault diagnosis using combined complete empirical mode decomposition and TKEO methods

This paper proposes a new method for bearing fault detection in induction motors by combining Complementary Ensemble Empirical Mode Decomposition (CEEMD) and the Teager-Kaiser Energy Operator (TKEO). The proposed method addresses the challenges of analyzing non-stationary and non-linear signals in motor current signature analysis. CEEMD is employed to decompose the stator current signals into Intrinsic Mode Functions (IMFs), effectively reducing mode mixing issues common in traditional EMD. The TKEO is then applied to extract amplitude and frequency modulations with high temporal resolution, enabling precise identification of fault-specific frequencies. Experimental validation was conducted on a test bench for various bearing fault conditions: inner race, outer race, and ball defects. Results demonstrate that the combination of CEEMD-TKEO successfully detects characteristic fault frequencies with improved accuracy compared to conventional methods. The method identified specific frequency components and their sidebands, correlating with theoretical calculations. Furthermore, correlation coefficient analysis between original signals and their IMFs provided effective mode selection for fault diagnosis. The proposed approach shows promising results for industrial applications, offering both computational efficiency and reliable fault detection capabilities.

Vibration reduction of rotor systems using magneto-rheological elastomeric supports

Support characteristics (stiffness and dissipation) are well understood to influence the dynamic behaviour of rotors. Supports play an essential role in deciding the stability and response of the rotors. Therefore, frequency-dependent support characteristics are beneficial in determining the rotor response as the frequency of excitation keeps changing. An elegant choice of making rotor supports is by using magneto-rheological elastomeric (MRE) materials. The magnetic field application results in a change in the stiffness and dissipation behaviour of MREs and consequently influences the dynamics of the rotor–shaft system. An attempt is made in this work to conceptualize the MRE as viscoelastic sectors put between the outer race of a bearing and the bearing housing to make the rotor support. A rigorous analytical formulation is done to obtain the stiffness and dissipation of the support so formed in terms of a viscoelastic model of the material under the effect of dispersed magnetic material, the geometry of the sectors, the intensity of the magnetic field, and the frequency of forcing function the supports are subject to. The supports show promising results by reducing the vibrations at resonance by suitably applying the magnetic field. Therefore, MRE materials may be proposed as adaptive supports to both structures and rotors to induce intended dynamic behaviour.

AI-Driven Thermography-based Fault Diagnosis in Single-Phase Induction Motor

Single-phase induction motors (SIMs) are commonly used in industrial applications. The extensive industrial usage of SIMs requires proper maintenance and fault detection. Among various faults, the most common mechanical faults in SIMs are bearing faults. Thus, detecting these faults during motor operation is essential for preventing damage. Many fault diagnosis methods have been proposed to detect bearing faults based on contact sensors, but they have accessibility problems. This paper presents a novel Infrared Thermography (IRT) based fault diagnosis method, leveraging both conventional machine learning (CML) and deep learning (DL) techniques for motor condition classification. In CML, Statistical and Gray Level Co-occurrence Matrix (GLCM) features are extracted from thermal images. The support vector machine recursive feature elimination (SVM-RFE) method is used to select the most relevant high-score features from the extracted features. Support vector machine (SVM) with linear and radial basis function (RBF) kernel functions and K-nearest neighbours (KNN) classifiers are applied to the selected interpretable features to classify machines into five classes of healthy, inner race, outer race, missing ball with low lubrication, and no lubrication bearing faults. Furthermore, in DL, a convolutional neural network (CNN) of few simple convolutional layers with several filters is also used as a healthy and faulty bearings classifier. The proposed method achieves classification accuracy of 98.29% using CML and 100% using DL. Moreover, our IRT-based SIMs bearing fault diagnosis methods demonstrate a reduction in architectural complexity compared to prevailing state-of-the-art approaches.

FIR noise reduction using GWO and Fast Kurtogram for extracting fault characteristic frequencies in rolling bearings

Abstract Fast Kurtogram (FK) is a widely recognized method for detecting characteristic fault frequency bands. Nevertheless, its sensitivity to noise often causes diagnostic errors. This research seeks to enhance FK’s identification performance by implementing an adaptive FIR filter that mitigates noise. To validate this approach, single-point fault signals were obtained from the outer race, inner race, and rolling elements of bearings operating at four distinct motor speeds. These raw signals were then subjected to statistical analysis after normalization. To design the adaptive FIR filter, the study employs an enhanced Grey Wolf algorithm along with the Parks-McClellan algorithm to determine the optimal filter order and impulse response coefficients. Subsequently, band-pass filtering is applied to the fault characteristic frequency bands identified through FK. The fault characteristic frequencies are then extracted using envelope spectrum analysis. The bandwidth and fault characteristics derived from the processed signals are compared against those obtained from the unfiltered original signals. The findings indicate that optimal normalization preprocessing is crucial before filter design. The determined optimal order and impulse response coefficients enable the adaptive design of FIR filters. As a result, the FIR-filtered signal exhibits significantly fewer noise-induced anomalous peaks, thereby improving the signal-to-noise ratio and enhancing FK’s robustness against noise interference. Implementing an adaptive FIR filter for noise reduction significantly aids in more precise analysis and diagnosis of bearing conditions. Looking ahead, this study proposes integrating machine learning algorithms to further enhance the accuracy in identifying characteristic frequency bands associated with single-point faults in rolling elements.

Inner race dynamics of a high-speed ball bearing with elastic ring squeeze film damper

Abstract This paper aims at the inner race dynamics of a high-speed ball bearing incorporated with Elastic Ring Squeeze Film Damper (ERSFD). Dynamic model of the inner race with three degrees of freedom (DOFs) is established in an ERSFD integrated ball bearing. Ball contact force and operating contact angle are characterized through an improved quasi-static model where the geometric relationship between the bearing displacement and elastic deformation is developed by considering the outer race movement. Numerical simulations were performed for bearings with or without ERSFD. The model was validated using Campbell diagram, measured displacement response and center trajectory of the inner race at different rotor speeds. The results show that the ERSFD provides adequate stable traction on the ball and limits excessive increase in the inner race contact angle. Such load-carrying condition benefits for decrease in ball orbital revolution skid at high rotor speeds. The ERSFD helps rotor unbalance force dominate bearing motion, leading to a stable periodic motion for the inner race at high rotor speeds. For ball bearings without ERSFD inner race possess stable motion at low and moderate rotor speeds, but is more prone to instability at high rotor speeds.

Detection of Damage on Inner and Outer Races of Ball Bearings Using a Low-Cost Monitoring System and Deep Convolution Neural Networks

Bearings are vital components in machinery, and their malfunction can result in equipment damage and reduced productivity. As a result, considerable research attention has been directed toward the early detection of bearing faults. With recent rapid advancements in machine learning algorithms, there is increasing interest in proactively diagnosing bearing faults by analyzing signals obtained from bearings. Although numerous studies have introduced machine learning methods for bearing fault diagnosis, the high costs associated with sensors and data acquisition devices limit their practical application in industrial environments. Additionally, research aimed at identifying the root causes of faults through diagnostic algorithms has progressed relatively slowly. This study proposes a cost-effective monitoring system to improve economic feasibility. Its primary benefits include significant cost savings compared to traditional high-priced equipment, along with versatility and ease of installation, enabling straightforward attachment and removal. The system collects data by measuring the vibrations of both normal and faulty bearings under various operating conditions on a test bed. Using these data, a deep neural network is trained to enable real-time feature extraction and classification of bearing conditions. Furthermore, an explainable AI technique is applied to extract key feature values identified by the fault classification algorithm, providing a method to support the analysis of fault causes.

An improved approach on ball bearing characteristic frequencies

Bearing Characteristic Frequencies (BCF) are crucial for diagnosing faults in rotary systems, as they depend on the number of balls, bearing geometry, and shaft speed. However, divergences often arise between theoretical and actual BCF values due to additional influencing factors. This paper addresses these divergences by incorporating two key factors into the BCF equation. First, the speed variation among the shaft, inner race, outer race, and rolling elements is considered, as fault conditions can alter these speeds. Second, the fit factor, representing the impact of bearing mounting and component fits, is introduced to account for variations in BCF. The initial section of the paper reviews existing BCF equations proposed by various researchers, while the latter part presents modified equations that integrate these additional factors, offering a more accurate range of BCF values.

An FBG-based method for load measurement of a cylindrical rolling element bearing

Abstract Contact forces between raceways and rolling elements of a bearing stand as crucial operational aspects defining the bearing’s performance. While monitoring bearing load through load cells or strain gauges on the shaft or housing is feasible, it may not precisely reflect the distributed load transmitted by the rollers directly to the raceways. This paper introduces a method to calculate these contact forces, relying on a linear assumption correlating the loads to the measured strains on the outer race of a bearing. In our research, we conducted both static and dynamic tests on cylindrical roller bearings through simulations and experiments. We designed an experimental test rig to conduct both static and dynamic experiments and utilized FBG optical fiber sensors for contact force measurement in bearings due to their multiple advantages, including high sensitivity, resistance to corrosion and magnetic interference, and ease of installation on the bearing outer race due to their shape and size. Our findings indicate a consistent linear relationship between contact forces and strains measured on the bearing’s outer race. Furthermore, we calculated the linear coefficient K values from three test groups: static tests under simulation study, static tests under experimental study, and dynamic tests under experimental study. The K values obtained from static tests (simulation and experimental studies) and dynamic tests (experimental study) align consistently. Following this, we calculated contact forces in both static and dynamic experiments by multiplying the measured strains with K. Our calculations resulted in an error percentage of less than 2% for static tests and below 5% for dynamic tests, highlighting the accuracy of our approach in determining these contact forces.

Localized Bearing Fault Analysis for Different Induction Machine Start-Up Modes via Vibration Time-Frequency Envelope Spectrum.

Bearings are the most vulnerable component in low-voltage induction motors from a maintenance standpoint. Vibration monitoring is the benchmark technique for identifying mechanical faults in rotating machinery, including the diagnosis of bearing defects. The study of different bearing fault phenomena under induction motor transient conditions offers interesting capabilities to enhance classic fault detection techniques. This study analyzes the low-frequency localized bearing fault signatures in both the inner and outer races during the start-up and steady-state operation of inverter-fed and line-started induction motors. For this aim, the classic vibration envelope spectrum technique is explored in the time-frequency domain by using a simple, resampling-free, Short Time Fourier Transform (STFT) and a band-pass filtering stage. The vibration data are acquired in the motor housing in the radial direction for different load points. In addition, two different localized defect sizes are considered to explore the influence of the defect width. The analysis of extracted low-frequency characteristic frequencies conducted in this study demonstrates the feasibility of detecting early-stage localized bearing defects in induction motors across various operating conditions and actuation modes.

An interpretable hybrid framework combining convolution latent vectors with transformer based attention mechanism for rolling element fault detection and classification

Failure of industrial assets can cause financial, operational and safety hazards across different industries. Monitoring their condition is crucial for successful and smooth operations. The colossal volume of sensory data generated and acquired throughout industrial operations supports real-time condition monitoring of these assets. Leveraging digital technologies to analyze acquired data creates an ideal environment for applying advanced data-driven machine learning techniques, such as convolutional neural networks (CNNs) and vision transformer (ViT) to detect faults and classify, enabling accurate prediction and timely maintenance of industrial assets. In this paper, we present a novel hybrid framework based on the local feature extraction ability of CNN with comprehensive understanding of transformer within a global context. The proposed method leverages the complex weight-sharing properties of CNNs and ability of transformers to understand the larger context of spatial relationships in large-scale patterns, making it applicable to datasets of varying sizes. Preprocessing methods such as data augmentation are used to train the model on the Case Western Reserve University (CWRU) dataset in order to increase generalization through computational efficiency. An average fault classification accuracy of 99.62% is accomplished over all three fault classes with an average time-to-fault detection of 38.4 ms. MFPT fault dataset is used to further validate the method with an accuracy of 99.17% for outer race and 99.26% for inner race. Moreover, the proposed framework can be modified to accommodate alternative convolutional models.

A hybrid LSTM random forest model with grey wolf optimization for enhanced detection of multiple bearing faults

Bearing degradation is the primary cause of electrical machine failures, making reliable condition monitoring essential to prevent breakdowns. This paper presents a novel hybrid model for the detection of multiple faults in bearings, combining Long Short-Term Memory (LSTM) networks with random forest (RF) classifiers, further enhanced by the Grey Wolf Optimization (GWO) algorithm. The proposed approach is structured in three stages: first, time and frequency domain features are manually extracted from vibration signals; second, these features are processed by a dual-layer LSTM network, which is specifically designed to capture complex temporal relationships within the data; finally, the GWO algorithm is employed to optimize feature selection from the LSTM outputs, feeding the most relevant features into the RF classifier for fault classification. The model was rigorously evaluated using a dataset comprising six distinct bearing health conditions: healthy, outer race fault, ball fault, inner race fault, compounded fault, and generalized degradation. The hybrid LSTM-RF-GWO model achieved a remarkable classification accuracy of 98.97%, significantly outperforming standalone models such as LSTM (93.56%) and RF (98.44%). Furthermore, the inclusion of GWO led to an additional accuracy improvement of 0.39% compared to the hybrid LSTM-RF model without optimization. Other performance metrics, including precision, kappa coefficient, false negative rate (FNR), and false positive rate (FPR), were also improved, with precision reaching 99.28% and the kappa coefficient achieving 99.13%. The FNR and FPR were reduced to 0.0071 and 0.0015, respectively, underscoring the model’s effectiveness in minimizing misclassifications. The experimental results demonstrate that the proposed hybrid LSTM-RF-GWO framework not only enhances fault detection accuracy but also provides a robust solution for distinguishing between closely related fault conditions, making it a valuable tool for predictive maintenance in industrial applications.

Wind turbine condition monitoring dataset of Fraunhofer LBF

Fraunhofer wind turbine dataset contains monitoring data from a 750 W wind turbine, including accelerometers and tachometer, to capture structural response, bearing vibrations and rotational velocity. Additionally, temperatures, wind speed and wind direction have been measured, while weather conditions have been acquired from selected sources. Various damage scenarios, including mass imbalance, and aerodynamic imbalance as well as damages on bearings’ outer race, inner race and roller element have been implemented. The availability of time series data makes the dataset well suited for both machine learning and signal processing-based condition monitoring applications. The availability of heterogeneous sensors has created a dataset particularly suited for information fusion, data fusion, multi-sensor approaches, and holistic monitoring. Experiments were conducted in real-world conditions outside of a controlled laboratory environment, thereby introducing challenges such as variable rotor speed, noise, overloads, and other environmental factors. Consequently, the dataset is qualified for tasks involving uncertainty quantification and signal pre-processing. This document will detail the test equipment, experimental procedures, simulated damage cases and measurement parameters.

Characterization of bearing health in a noisy environment using vibroacoustic approach

The bearing fault detection localized at outer race on a rotating machine operating in the same environment with another machine (whose wind tunnel generates a different noise) is investigated using acoustic and vibration analysis. The impact of increased acoustic and vibratory background noise on outer race defect detection is studied through spectral kurtogram representation and envelope analysis. The spectral power densities of these two types of noise clearly show that the nature of their associated radiation (airborne or solid-borne) has a different impact. The presence of a strongly amplified background noise demonstrates that the acoustic modality can detect the outer race defect peak at higher levels than the vibratory modality.

Diagnosing multiple faults in rotating machinery using empirical mode decomposition and blind source separation methods

It is a challenging task to accurately diagnose multiple faults in a rotating machinery. In this paper, the authors have used empirical mode decomposition (EMD) and blind source separation (BSS) methods for this purpose. A portable rotating machinery setup is made in the laboratory and faults such as bearing inner and outer race faults, angular misalignment are created into it. The vibration data is collected using SPM made LEONOVA DIAMOND(r) FFT analyzer and processed using CONDMASTER RUBY(r) and MATLAB(r) softwares. Firstly, wavelet transform is applied to the signal to denoise it. After this, EMD is implemented to decompose the signal into different intrinsic mode functions (IMFs). After preprocessing these IMFs using the principal component analysis (PCA), BSS is used to extract fault-related information from the IMFs. A comparison of the results obtained from EMD and PCA-BSS is also presented in the paper. Keywords: Rotating machinery, Multiple faults, Vibration data, Wavelet transform, Empirical mode decomposition, Blind source separation, Principal component analysis.

Thrust ball bearing fault identification from visualized data using optimized deep network

Bearings are the most common components employed in machine parts. The bearings' movement facilitates the smooth motion. They also help with friction reduction. The faults in bearings are often caused by tribological parameters. Various methods have been developed to identify faults in bearings, but they often fail to predict these accurately. This work has concentrated on designing an effective fault-recognizing model. Therefore, a framework, the Zebra-based Radial Basis Prediction Mechanism (ZbRBPM), was proposed in this work. Initially, the bearing datasets are collected, and mineral oil (MO) lubrication is added to minimize wear and friction. The primary goal of the work is to detect and classify the fault in the thrust ball bearing. The bearing vibration data included a normal vibration signal, a ball fault, and inner and outer race faults. Hence, Zebra fitness is enhanced for the optimization of tribological parameters and fault identification. The proposed model is executed in the MATLAB system. Finally, performance criteria like accuracy, precision, recall, F-score, error rate, computation time, speed, specific wear rate, friction torque, and energy consumption are validated. The performance of the proposed ZbRBPM gains better accuracy rate as 99.5 %, 99 % of precision and f-score and 99.4 % of recall. Also it significantly reduced the prediction error rate into 0.5 % with lower computation time and very low wear and friction.

Robust post-processing time frequency technology and its application to mechanical fault diagnosis

Post-processing synchrosqueezing transform and synchroextracting transform methods can improve TFR resolution for fault diagnosis. The normal and fault signal can be described by infinite variance process, and 1 < α ≤ 2, even the background noise belongs to the process under complex conditions. The effect of traditional SST and SET methods is greatly reduced and even lost in infinite variance process environment. Several robust post-processing methods are proposed including FSET, FSSET, FSOSET and FMSST technology employing infinite variance process statistical model and FLOS, and their mathematical derivation are completed in this paper. The proposed methods are compared with the conventional methods, and the results show that the proposed methods achieve better results than the existing methods. In addition, the new methods are applied to diagnose the bearing outer race DE signals polluted by infinite variance process, the result demonstrates that they have performance advantages. Finally, the characteristics, shortcomings and application scenarios of the improved algorithms are summarized.

Dynamic characteristics of high-speed angular contact ball bearings considering the coupling of outer race waviness and local defects

This paper proposes a novel four degrees of freedom model of angular contact ball bearings (ACBBs) under high-speed effect, in order to characterize the dynamic behavior of rolling bearings with coupling waviness and local defects on the outer race in detail. Firstly, based on bearing quasi-static model, considering centrifugal force, gyroscope moment, the bearing contact stiffness is calculated. Then, the representation model of outer race waviness and the time-varying displacement excitation expression of the bearing with local defects are put forward respectively. On the basis, a dynamic model of ACBBs with coupling factors is built. Finally, the vibration characteristics of the bearing with faults are investigated, the model is verified by the existing experimental results and the influence of different parameters on the bearing dynamics is analyzed. The results show that the bearing fault characteristic frequency is unchanged when the high-speed effect is considered, but the vibration amplitude is smaller than that without the high-speed effect. With the increase of speed, the bearing fault characteristic frequency and vibration amplitude increase. When considering the coupling of outer race waviness and local defects, with the increase of the two factors, the bearing fault characteristic frequency remains unchanged but the vibration amplitude increase.

Inner and Outer Races Bearing Damage Detection using Low-cost Fault Detecting Sensors

Bearings are small components of a machine used in various industries. However, bearings have an important role in transferring energy from the main motor to other parts. Therefore, monitoring the condition of the bearings is very necessary to keep the production process running smoothly. The study provides a new perspective that bearing damages were analyzed by low-cost sensors, namely the Accelerometer sensor (ADXL-335) and Arduino Uno. The results showed that the low-cost sensor developed in this study was able to detect damage to the ball bearing. The Fast Fourier Transform (FFT) signal processing tool works compatible with the low-cost sensor and could be used to determine the type of damage to the bearing by analyzing the signal frequency spectrum. In this process, there were several frequencies that appear with characteristics related to the condition of the bearing. The working frequency of the shaft rotation on the bearing with normal conditions was 10 Hz, the frequency with damage to the inner race of the bearing was 55.52 Hz, and the bearing with damage to the outer race included a frequency of 34.47 Hz.