Wear Fault Research Articles

Research on early warning of rolling bearing wear failure based on empirical mode decomposition

In order to solve the problems of low precision and long time-consuming of traditional methods, this paper designs a rolling bearing wear fault early warning method based on empirical mode decomposition (EMD). Based on the wear reason of rolling bearing, the acceleration sensor is used to collect its vibration signal, and the EMD algorithm is used to stabilise the signal to obtain multi-scale signal. Each multi-scale signal is decomposed into sub-band to get multi-scale sub-band signal, then the multi-scale sub-band sample entropy is obtained, and the optimisation function of local preserving projection algorithm is constructed to obtain the eigenvalue and eigenvector of wear failure fault, and finally the fault early warning is realised. The simulation results show that the signal denoising effect of this method is good, the early warning accuracy is always above 94%, and the average alarm time is close to 0.27 s.

Research on Dynamic Response Characteristics and Condition Indicators of Gear Wear Fault

Research on Dynamic Response Characteristics and Condition Indicators of Gear Wear Fault

Research on the variation of mesh stiffness and transmission error for spur gear with tooth profile modification and wear fault

Tooth profile modification and wear are unavoidable, even intentional tooth profile deviations. Both of them have a significant influence on gear mesh stiffness and quasi-static transmission error (QSTE). In this paper, to efficiently and accurately obtain the two internal excitations of spur gear pairs with the above two kinds of deviations together, improved analytical mesh stiffness and QSTE models are proposed. And the effects of corner contact and structure coupling are considered together to overcome the defect of previous works. By comparing with the results of the finite element method, the established models are verified. Through the models, the mesh stiffness and QSTE of the gear pairs with the tooth profile modification and wear simultaneously can be obtained. The results show that both of the two kinds of deviations have significant effects on the limit value and distribution of the mesh stiffness and QSTE. Besides, it can be found that there is a competition between tooth profile modification and wear, and the influence extent of wear is gradually increasing that eventually can override the modification effect. Moreover, wear can affect the selection of tooth profile modification.

An on-board detection framework for polygon wear of railway wheel based on vibration acceleration of axle-box

The polygon wear of railway wheel (PWRW) is a wear fault that is ubiquitous in railway vehicles. PWRW can induce a strong periodic excitation to both vehicle and track, which not only decreases passenger comfort but also is detrimental to the operational reliability and safety. Both the degree and the order of PWRW are important parameters used to quantify the fault. Because the fault-related components distribute at a wide range in the frequency domain, it is easy to alias with some radiated vibrations from vehicle and track components, which makes the on-board detection for both parameters of PWRW very difficult. To address the practical engineering problem, this paper proposes a detection framework based on the angle domain synchronous averaging technique (ADSAT). The detection method employs the vertical axle-box vibration acceleration (ABVA), which is easy to obtain and can also be used to monitor the conditions of axle-box bearings. The paper compares the proposed and traditional methods. The results reveal that the proposed method not only achieves the order detection which the traditional method cannot, but also mitigates the influence of background noise. The feasibility and effectiveness of the proposed method to improve the detection accuracy of PWRW is demonstrated through simulation and real field investigations.

Simulation-Driven Deep Learning Approach for Wear Diagnostics in Hydrodynamic Journal Bearings

Abstract Early diagnosis in rotating machinery has been a challenge when looking toward the concept of intelligent machines. A crucial and critical component in these systems is the lubricated journal bearing, subjected to wear fault by abrasive removing of material in its inner wall, mainly during run-ups and run-downs. In extreme conditions, wear faults can cause unexpected shutdowns in rotating systems. Consequently, advanced condition monitoring is an essential procedure in the wear diagnosis of journal bearings. Although an increasing number of data-driven condition monitoring approaches for rotating machines have been proposed in the past decade, they mostly rely on substantial amounts of experimental data for training, which is expensive and time-consuming to obtain. The objective of this work is to develop a framework using a deep learning algorithm to classify wear faults in hydrodynamic journal bearings using simulated vibrations signals. Numerically simulated data sets under different wear severity levels and operating conditions were used to train and test the diagnostics framework. The results show that the proposed framework can be a promising tool to diagnose wear faults in journal bearings.

Optimized CNN model for identifying similar 3D wear particles in few samples

Typical wear particles can be considered as distinctive indicators of on-going wear faults in machines. However, the small number of samples has limited the identification accuracy for similar fault particles. Besides, the three-dimensional (3D) characterization of wear particles may face huge challenges due to excessive surface parameters. Focusing on the problems of high similarity particles and few samples, a CNN-based particle classification method is developed with an example set of fatigue and severe sliding particles, which are the product of severe wear of machines. For data reduction, imaged 3D particle surfaces are firstly converted into 2D depth maps without losing surface information. Virtual fault particle images are then synthesized using a Conditional Generative Adversarial Networks (CGAN), according to the particle generation mechanism and distinctive particle features. Furthermore, a non-parametric particle identification model is established with the optimization of the CNN structures and training method, and the network is further optimized with the particle image standardization and the network visualization. Validation experiments reveal that the proposed method can accurately identify all tested fatigue and severe sliding particles with their typical characteristics.

A parametric model to identify hydrodynamic bearing wear at a single rotating speed

For dynamic systems, fault detection and diagnosis involve the observation of vibration signals. Faults associated with hydrodynamic bearings are one of the most common causes of forced shutdowns in rotating machinery. If it is a wear fault, the bearing clearance is affected changing the rotating system characteristics. Therefore, the time response of a rotating machine subjected to worn hydrodynamic bearings is analysed to observe the vibration behaviour in presence of bearing wear. Spectral analysis is used to understand intact and worn bearing characteristics and to identify wear severity. For the wear, an improved and new parametric model is generated to increase simulation accuracy. The new approach indicates the severity of wear in the studied hydrodynamic bearings.

Feature extraction of the hydraulic pump fault based on improved Autogram

Abstract Autogram is a good tool to extract the fault feature information from a fault vibration signal of rolling element bearings. Structure of a hydraulic pump is more complicated than the rolling element bearings, and its vibration signal is more contaminated by heavy Gaussian and non-Gaussian noises if slipper wear fault happens, and too many noise amplitudes can be introduced into computation of kurtosis in the time domain, and a data source of containing rich fault feature information cannot be successfully selected by the kurtosis, and it is ineffective application of the hydraulic pump. Aiming to resolve the above problems, an improved Autogram called PSE-Autogram is proposed. Its key and different selection of the data source is completed by the power spectral entropy (PSE) rather than the kurtosis, and the fault feature information can be made highlighted and noise can be suppressed by PSE in the frequency domain, and shortcomings of the original Autogram can be effectively overcame. A simulated signal and a slipper wear fault signal are tested and verified, and results demonstrate that PSE-Autogram performs better than Autogram and traditional fast kurtogram based on the assessment criterion of feature energy ratio.

A novel approach for predicting tool remaining useful life using limited data

Wear, fracture, and other tool faults affect the quality of a machined workpiece and can even damage machine tools. The accurate prediction of remaining useful life (RUL) can prevent a tool from suddenly failing, an ability of significance for ensuring machining quality and providing effective predictive maintenance strategies. Most current approaches for predicting tool RUL are based on historical failure and truncation data. However, for new types of tools or when a similar tool has just launched, such failure and truncation data are limited or even unavailable, making RUL prediction a challenge when using previously proposed methods. To address this problem, a novel method for the prediction of tool RUL using limited data is proposed in this study. A time window is constructed to track the tool condition using sensor data, and its size can be dynamically adjusted according to the wear factor and increase rate. Then, a deep bidirectional long short-term memory (DBiLSTM) neural network in which sequential data are predicted and smoothed by forwards and backwards directions, respectively, is developed to encode temporal information and identify long-term dependencies. On this basis, multi-step ahead rolling predictions are then employed to predict tool RUL. Finally, the effectiveness of the proposed method is verified using the results of milling experiments. These results show that the proposed method is able to predict tool RUL with high accuracy using only limited data.

Fluorescence-enhanced microfluidic sensor for highly sensitive in-situ detection of copper ions in lubricating oil

Sensitive on-line detection of Cu2+ in lubricating oil of marine power plant is important for displaying equipment wear status and fault pre-judgment in real time. However, there are few reports on in-situ, portable and highly sensitive Cu2+ detection devices in organic media before. In this report, CsPbBr3 perovskite quantum dots (PQDs) are employed as fluorescence probe for Cu2+ detection in microfluidic chip. The size effect of the CsPbBr3 QDs on the detection performances is firstly investigated. To obtain preferable sensitivity and detection limit, polymethyl methacrylate opal photonic crystals (PMMA OPCs) film is developed as microfluidic sensor substrate, which can tremendously improve the fluorescence intensity of PQDs as high as 26-fold. The enhancement effect can be attributed to the coupling between photonic crystal stopband effect and the excitation light field. Moreover, the use of photomultiplier module for fluorescence signal measurement enables the detection device to be portable. The novel strategy offers the advantages of higher sensitivity (8.62 μM–1) and lower detection limit (0.40 nM). Otherwise, this sensor also provides wide detection range and good selectivity, thus, it will have potential in oil quality in-situ detection and mechanical wear prediction in the future.

Diagnostic method and application of low speed sliding bearing wear fault

The static thermal and dynamic characteristics of the oil film affect the motion stability of the rotor system and directly determine whether the entire equipment can operate safely and stably. In this paper, the characteristics of sliding bearing oil film are studied by theoretical analysis, and the bearing wear fault diagnosis model is proposed. The model is verified by the actual production. The results show that the proposed diagnostic model can quickly locate the cause of the fault and ensure the safe and stable operation of the equipment.

Planning and construction of mechanism for surface wear testing and fault analysis in quadril joint prosthetic tribosystem

The problems associated with prosthetic failures and revision surgeries have increased greatly since total joint replacement has also been applied to younger and more active patients, so the need to resolve or reduce wear-related problems is of paramount importance. Based on this premise, it was decided to design, manufacture, and validate a wear test machine of the orbital coupling type, capable of performing the total hip prosthesis (THP) wear test, according to load and movement parameters (flexion/extension, adduction/abduction), established by the Standard ABNT NBR ISO 14242-3. After validation, a sample composed of a 28 mm diameter femoral head of ASTM F138 stainless steel and conventional polyethylene acetabulum (UHMWPE) was performed. The assay lasted for one million cycles (approximately 12 days), equivalent to approximately 1 year of use of the prosthesis in vivo. Two mass and volume loss evaluations were performed throughout the trial, and the results of which were consistent with the results presented by other authors. The evaluation of the wear of the acetabular component (visual and microscopic analysis) and the morphological analysis of the particles generated by the wear, through MEV, provided important information to better interpret the mechanisms of wear (abrasion, adhesion and fatigue) that occurred during the tribological test. It is a fact that these wear mechanisms compromise the useful life of a hip joint prosthetic tribosystem, interfering in the longevity of the prosthetic joint or often leading to total implant failure, resulting in many cases in revision surgeries.

Wear Fault Diagnosis of Aeroengines Based on Broad Learning System and Ensemble Learning

An aircraft engine (aeroengine) operates in an extremely harsh environment, causing the working state of the engine to constantly change. As a result, the engine is prone to various kinds of wear faults. This paper proposes a new intelligent method for the diagnosis of aeroengine wear faults based on oil analysis, in which broad learning system (BLS) and ensemble learning models are introduced and integrated into the bagging-BLS model, in which 100 sub-BLS models are established, which are further optimized by ensemble learning. Experiments are conducted to verify the proposed method, based on the analysis of oil data, in which the random forest and single BLS algorithms are used for comparison. The results show that the output accuracy of the proposed method is stable (at 0.988), showing that the bagging-BLS model can improve the accuracy and reliability of engine wear fault diagnosis, reflecting the development trend of fault diagnosis in implementing intelligent technology.

Ferrograph Analysis With Improved Particle Segmentation and Classification Methods

Abstract Ferrograph analysis has been adopted over decades for determining the root causes of on-going wear faults. After decades of manual operation, this traditional technique is being driven by intelligent algorithms for automatic identification of wear debris. However, the accuracy and robustness of this algorithm remain marginalized when applied in industries due to various types and color blurry of particles. To address this issue, this paper introduces an automatic ferrograph analysis model with a segmentation method and a two-level classification strategy. In order to obtain wear particles from the color ferrograph image, an adaptive Otsu threshold is adopted in three channel images to solve the color blurry in particle segmentation. By grouping particle parameters into shape and morphology ones, a two-level identification strategy is proposed. The first one is to classify rubbing, cutting, and spherical particles, referring to the fuzzy approach degree of shape parameters. In the second level, the shape-close particles are classified with imperceptible textures and back propagation neural network (BPNN). These objective parameters are constructed by applying the principal component analysis into seven texture features and inputted into a BPNN-based model to classify fatigue and severe sliding particles. In order to train the BPNN, more than 100 ferrograph images are sampled together, whereby standard ferrograph analysis is performed on the particle identification. The performance of the identification exhibits an accuracy exceeding 90% for rubbing, cutting, and spherical particles, whereas about 80% accuracy has been registered for both severe sliding and fatigue particles.

Machine learning-based wear fault diagnosis for marine diesel engine by fusing multiple data-driven models

Wear fault is one of the dominant causes for marine diesel engine damage which significantly influences ship safety. By taking full advantage of the data generated in engine operation, machine learning-based wear fault diagnostic model can help engineers to determine fault modes correctly and take quick action to avoid severe accidents. To identify wear faults more accurately, a multi-model fusion system based on evidential reasoning (ER) rule is proposed in this paper. The outputs of three data-driven models including an artificial neural network (ANN) model, a belief rule-based inference (BRB) model, and an ER rule model are used as pieces of evidence to be fused in decision level. In this paper, the fusion system defines reliability and importance weight of every single model respectively. A novel method is presented to determine the reliability of evidence by considering the accuracy and stability of every single model. The importance weight is optimized by genetic algorithm to improve the performance of the fusion system. The proposed machine learning-based diagnostic system is validated by a set of real samples acquired from marine diesel engines in operation. The test results show that the system is more accurate and robust, and the fault tolerant ability is improved remarkably compared with every single data-driven diagnostic model.

Research on a diagnostic method of gear wear fault degree based on ultra‐complete independent component analysis

In this study, a diagnostic method for gear wear fault is proposed, which is based on the ultra-complete independent component analysis. The ultra-complete analysis model is constructed, and the similar source signals more than mixing signals are separated. Based on the typical features of the fault source signal, the useful component of the mixing signals can be found. As the fault source signal and similar fault source signal are similar, the magnification range of similar shapes can be estimated by using the interval estimation method. The mapping between one-way varying magnification time domain and rotary component fault degree is identified, and the fault degree judgment standard can be established. Thus, the fault degree can be determined. The method is used to deal with the state monitoring data of the gear, and the analysis results show that the diagnostic method is of practical value.

Automated gear fault detection of micron level wear in bevel gears using variational mode decomposition

Gearboxes have an important role in power transmission systems. For such systems, vibration-based fault diagnosis techniques are frequently used to prevent premature failure and to ensure smooth transmission. We automated the fault diagnosis of gears having level of wear fault at micron using variational mode decomposition (VMD). VMD has been applied iteratively with specific input parameters. VMD decomposes the gear vibration signal into different narrowband components (NBCs) or obtained components (OCs). Various statistical features, namely kurtosis, skewness, standard deviation, root mean square, and crest factor, were extracted from the different OCs. Kruskal-Wallis test based on probability values was used to identify the significant features. For the automation of fault detection system, a comparative study was done using the random forest, multilayer perceptron, and J48 classifiers. The proposed method exhibits 96.5 % accuracy using random forest classifier with combined kurtosis, skewness, and standard deviation features.

Review on online inductive wear debris monitoring technology

Online oil monitoring technique is now an important development means to monitor wear condition and diagnose wear fault real time. Inductive wear debris monitoring is the research hotspot in the current online oil monitoring for its unique characteristics. Based on the current domestic and international online inductive wear debris monitoring, the relevant comments to the online inductive wear debris monitoring technology were given. Then, the positive and inverse problems of numerical analysis of inductive wear debris monitoring are analysed. Based on the above, the difficulties and shortages of online inductive wear debris monitoring are studied and the development trend and research focus of online inductive wear debris monitoring are discussion to provide for the oil online technology research.

Fault identification technology for gear tooth surface wear based on MPE method by MI and improved FNN algorithm

Multiscale Permutation Entropy (MPE) is a presented nonlinear dynamic technology for measuring the randomness and detecting the nonlinear dynamic change of time sequences and can be used effectively to extract the nonlinear dynamic wear fault feature of gear tooth surface from vibration signals of gear set. To solve the subjectivity drawback of threshold parameter selection process in MPE method, a joint calculation method based on the Mutual Information (MI) and improved False Nearest Neighbor (FNN) principle for calculating threshold parameters for MPE method was presented in this article. Then, the influence of threshold parameters on the identification accuracy of fault features with the MPE was studied by analyzing simulation data. Through the simulation analysis, the effectiveness of the proposed MPE method is validated. Finally, the wear failure test of spur gear was carried out, and the proposed method was applied to analyze the experimental data of fault signal. Meanwhile, the vibration characteristics of the fault signal are acquired. The analysis results show that the proposed method can effectively realize the fault diagnosis of gear box and has higher fault identification accuracy than the existing methods.

A Fault Feature Extraction Method for the Fluid Pressure Signal of Hydraulic Pumps Based on Autogram

Center spring wear faults in hydraulic pumps can cause fluid pressure fluctuations at the outlet, and the fault feature information on fluctuations is often contaminated by different types of fluid flow interferences. Aiming to resolve the above problems, a fluid pressure signal method for hydraulic pumps based on Autogram was applied to extract the fault feature information. Firstly, maximal overlap discrete wavelet packet transform (MODWPT) was adopted to decompose the contaminated fault pressure signal of center spring wear. Secondly, based on the squared envelope of each node, three kinds of kurtosis of unbiased autocorrelation (AC) were computed in order to describe the fault feature information comprehensively. These are known as standard Autogram, upper Autogram and lower Autogram. Then a node corresponding to the biggest kurtosis value was selected as a data source for further spectrum analysis. Lastly, the data source was processed by threshold values, and then the fault could be diagnosed based on the fluid pressure signal.