Misalignment is one of the key reasons for vibrations in most of the rotating system. The present study focuses on interactions among speed, load, and defect severity by investigating their effect on the system vibration. Response surface methodology (RSM) with root-mean-square (RMS) as a response factor is used to understand the influence of such interactions on the system performance. Experiments are planned using design of experiments, and analysis is carried out using analysis of variance (ANOVA). It is observed that speed has a remarkable effect on RMS value in both parallel and angular types of misalignment and affects the system performance. RSM results revealed that a change in load has less impact on vibration amplitude in case of horizontal and vertical directions, but there is a significant variation in RMS value in axial direction for both types of misalignment. A slight increase in the RMS value with an increase in defect severity is observed in the axial direction. These observations will help to understand the misalignment defect and its effect in a better way.
{"title":"Experimental Investigation Using Response Surface Methodology for Condition Monitoring of Misaligned Rotor System","authors":"Shital M. Patil, A. Jalan, A. Marathe","doi":"10.1115/1.4051771","DOIUrl":"https://doi.org/10.1115/1.4051771","url":null,"abstract":"\u0000 Misalignment is one of the key reasons for vibrations in most of the rotating system. The present study focuses on interactions among speed, load, and defect severity by investigating their effect on the system vibration. Response surface methodology (RSM) with root-mean-square (RMS) as a response factor is used to understand the influence of such interactions on the system performance. Experiments are planned using design of experiments, and analysis is carried out using analysis of variance (ANOVA). It is observed that speed has a remarkable effect on RMS value in both parallel and angular types of misalignment and affects the system performance. RSM results revealed that a change in load has less impact on vibration amplitude in case of horizontal and vertical directions, but there is a significant variation in RMS value in axial direction for both types of misalignment. A slight increase in the RMS value with an increase in defect severity is observed in the axial direction. These observations will help to understand the misalignment defect and its effect in a better way.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"140 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2021-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76586986","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Three-point bending fatigue compliance datasets of multi-layer fiberglass-weave/epoxy test specimens, including 5 and 10 mil interlayers, were analyzed using artificial intelligence (AI) methods along with statistical analysis, revealing the existence of three different compliance-based damage modes. Anomaly detection algorithms helped discover damage indicators observable in short intervals (of 50 cycles) in the compliance data, whose patterns vary with the material and the number of load cycles to which the material is subjected. Machine learning algorithms were applied using the compliance features to assess the likelihood that material failure may occur within a certain number of future loading cycles. High accuracy, precision, and recall rates were achieved in the classification task, for which we evaluated several algorithms, including various variations of neural networks and support vector machines. Thus, our work demonstrates the utility of AI algorithms for discovering a diversity of damage mechanisms and failures.
{"title":"Machine Learning and Anomaly Detection Algorithms for Damage Characterization From Compliance Data in Three-Point Bending Fatigue","authors":"Subodh Kalia, Jakob Zeitler, C. Mohan, V. Weiss","doi":"10.1115/1.4051903","DOIUrl":"https://doi.org/10.1115/1.4051903","url":null,"abstract":"\u0000 Three-point bending fatigue compliance datasets of multi-layer fiberglass-weave/epoxy test specimens, including 5 and 10 mil interlayers, were analyzed using artificial intelligence (AI) methods along with statistical analysis, revealing the existence of three different compliance-based damage modes. Anomaly detection algorithms helped discover damage indicators observable in short intervals (of 50 cycles) in the compliance data, whose patterns vary with the material and the number of load cycles to which the material is subjected. Machine learning algorithms were applied using the compliance features to assess the likelihood that material failure may occur within a certain number of future loading cycles. High accuracy, precision, and recall rates were achieved in the classification task, for which we evaluated several algorithms, including various variations of neural networks and support vector machines. Thus, our work demonstrates the utility of AI algorithms for discovering a diversity of damage mechanisms and failures.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"29 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2021-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77890206","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Turbine blades are critical machine components in power plants and aerospace turbo engines. Failure of these blades in operation leads to catastrophic damages as well as high cost of maintenance and repair. Blades are often assembled in packets with lacing wire or shroud ring interconnections. Natural frequencies of the bladed packets are designed in a specific range to avoid possible resonant stresses. However, frequent damages during operation alter the stiffness of the blade-packet assembly and change the eigen-spectrum. A numerical study is presented in this work, where it is demonstrated that characteristic changes in eigen-spectrum can identify both severity and location of such damages. The work employs matrix perturbation theory on the eigen-value problem, formulated from the lumped-parameter modeling of the blade packet. Damage is considered as a perturbation in the stiffness matrix with damage severity acting as the perturbation parameter. First, a graphical pattern recognition method, and then, a damage proximity index evaluation method is suggested for damage identification. Further, an estimation algorithm for damage severity is presented with numerically simulated computations, which demonstrates that the methods can exactly identify the damage location and, with very little error, can estimate the damage severity.
{"title":"Identification and Estimation of Damage Severity in a Turbine Blade Packet Using Inverse Eigen-Value Analysis—A Numerical Study","authors":"Animesh Chatterjee","doi":"10.1115/1.4051582","DOIUrl":"https://doi.org/10.1115/1.4051582","url":null,"abstract":"\u0000 Turbine blades are critical machine components in power plants and aerospace turbo engines. Failure of these blades in operation leads to catastrophic damages as well as high cost of maintenance and repair. Blades are often assembled in packets with lacing wire or shroud ring interconnections. Natural frequencies of the bladed packets are designed in a specific range to avoid possible resonant stresses. However, frequent damages during operation alter the stiffness of the blade-packet assembly and change the eigen-spectrum. A numerical study is presented in this work, where it is demonstrated that characteristic changes in eigen-spectrum can identify both severity and location of such damages. The work employs matrix perturbation theory on the eigen-value problem, formulated from the lumped-parameter modeling of the blade packet. Damage is considered as a perturbation in the stiffness matrix with damage severity acting as the perturbation parameter. First, a graphical pattern recognition method, and then, a damage proximity index evaluation method is suggested for damage identification. Further, an estimation algorithm for damage severity is presented with numerically simulated computations, which demonstrates that the methods can exactly identify the damage location and, with very little error, can estimate the damage severity.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"37 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2021-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84098014","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Naman S. Bajaj, A. Patange, R. Jegadeeshwaran, Kaushal A. Kulkarni, Rohan S. Ghatpande, Atharva M. Kapadnis
With the advent of industry 4.0, which conceptualizes self-monitoring of rotating machine parts by adopting techniques like data analytics, cloud computing, Internet of things, machine learning (ML), and artificial intelligence. The significant research area in predictive maintenance is tool condition monitoring (TCM) as the tool condition affects the overall machining process and its economics. Lately, machine learning techniques are being used to classify the tool’s condition in operation. These techniques are cost saving and help industries with adopting future-proof solutions for their operations. One such technique called discriminant analysis (DA) must be examined particularly for TCM. Owing to its less-expensive computation and shorter run times, using them in TCM will ensure the effective use of the cutting tool and reduce maintenance times. This article presents a Bayesian optimized discriminant analysis model to classify and monitor the tool condition into three user-defined classes. The data are collected using an in-house designed and developed data acquisition (DAQ) module setup on a Vertical Machining Center (VMC). The hyperparameter tuning has been incorporated using Bayesian optimization search, and the parameter that gives the best model was found out to be “linear,” achieving an accuracy of 93.3%. This study confirms the feasibility of machine learning techniques like DA in the field of TCM and using Bayesian optimization algorithms to fine-tune the model, making it industry ready.
{"title":"A Bayesian Optimized Discriminant Analysis Model for Condition Monitoring of Face Milling Cutter Using Vibration Datasets","authors":"Naman S. Bajaj, A. Patange, R. Jegadeeshwaran, Kaushal A. Kulkarni, Rohan S. Ghatpande, Atharva M. Kapadnis","doi":"10.1115/1.4051696","DOIUrl":"https://doi.org/10.1115/1.4051696","url":null,"abstract":"\u0000 With the advent of industry 4.0, which conceptualizes self-monitoring of rotating machine parts by adopting techniques like data analytics, cloud computing, Internet of things, machine learning (ML), and artificial intelligence. The significant research area in predictive maintenance is tool condition monitoring (TCM) as the tool condition affects the overall machining process and its economics. Lately, machine learning techniques are being used to classify the tool’s condition in operation. These techniques are cost saving and help industries with adopting future-proof solutions for their operations. One such technique called discriminant analysis (DA) must be examined particularly for TCM. Owing to its less-expensive computation and shorter run times, using them in TCM will ensure the effective use of the cutting tool and reduce maintenance times. This article presents a Bayesian optimized discriminant analysis model to classify and monitor the tool condition into three user-defined classes. The data are collected using an in-house designed and developed data acquisition (DAQ) module setup on a Vertical Machining Center (VMC). The hyperparameter tuning has been incorporated using Bayesian optimization search, and the parameter that gives the best model was found out to be “linear,” achieving an accuracy of 93.3%. This study confirms the feasibility of machine learning techniques like DA in the field of TCM and using Bayesian optimization algorithms to fine-tune the model, making it industry ready.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"23 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2021-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89365294","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Online structural health corrosion monitoring in advanced molten salt reactor heat exchangers is desirable for detecting tube degradation prior to leaks that would either cause mixing of heat exchanger fluids or release of radiologically contaminated fluids beyond the design containment boundary. This program seeks to demonstrate the feasibility for a torsional wave mode sensor to attach to the outside of a long (30-m) heat exchanger tube in the stagnant flow area where the tube joins the heat exchanger plenum and where it is possible to protect a sensor cable from high-force flow connecting through a heat exchanger shell to a monitoring instrument. The envisioned sensor and cable management approach will be impractical to implement on existing heat exchangers; rather, sensors must be installed in conjunction with the heat exchanger fabrication. Initially, flaw surrogates of interest (50% notch and 50% flat-bottom hole) have been detected in a 3-m tube using low-temperature PZT piezoelectric crystals. The transducer consisted of multiple shear elements placed circumferentially around a tube. The program will continue to investigate higher temperature piezoelectric ceramics, long-term performance of high-temperature adhesives, and flaw sensitivity on long (30-m +) tubes.
{"title":"Feasibility of Molten Salt Reactor Heat Exchanger Online Monitoring","authors":"Samuel Glass, M. Good, E. Forsi, R. Montgomery","doi":"10.1115/1.4051486","DOIUrl":"https://doi.org/10.1115/1.4051486","url":null,"abstract":"\u0000 Online structural health corrosion monitoring in advanced molten salt reactor heat exchangers is desirable for detecting tube degradation prior to leaks that would either cause mixing of heat exchanger fluids or release of radiologically contaminated fluids beyond the design containment boundary. This program seeks to demonstrate the feasibility for a torsional wave mode sensor to attach to the outside of a long (30-m) heat exchanger tube in the stagnant flow area where the tube joins the heat exchanger plenum and where it is possible to protect a sensor cable from high-force flow connecting through a heat exchanger shell to a monitoring instrument. The envisioned sensor and cable management approach will be impractical to implement on existing heat exchangers; rather, sensors must be installed in conjunction with the heat exchanger fabrication. Initially, flaw surrogates of interest (50% notch and 50% flat-bottom hole) have been detected in a 3-m tube using low-temperature PZT piezoelectric crystals. The transducer consisted of multiple shear elements placed circumferentially around a tube. The program will continue to investigate higher temperature piezoelectric ceramics, long-term performance of high-temperature adhesives, and flaw sensitivity on long (30-m +) tubes.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"1 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2021-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78837839","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ultrasonic non-destructive testing traditionally uses a conventional monolithic transducer. An approach similar to this comprising of independent single transmissions but with reception performed by all the elements in phased array ultrasonics is known as Full Matrix Capture (FMC). The acquired data is processed by Total Focusing Method (TFM). Conventional FMC-TFM has limitations in the inspection at large depth in attenuating materials due to single element transmission. To improve the beam forming process, coherent recombination of the plane wave with specific angles is utilized in transmission and the same aperture is used for the reception in Plane Wave Imaging (PWI). A new methodology called Angle Beam Virtual Source FMC-TFM (ABVSFMC-TFM) is proposed to inspect thick attenuating materials such as nickel base alloys. The ABVSFMC method leads to improved Signal to Noise Ratio (SNR) as compared to the conventional FMC due to increased energy with directivity during transmission using a group of elements and improved divergence as compared to the PWI due to a small virtual source near the sample surface. In the present paper, FMC-TFM, PWI-TFM and ABVSFMC-TFM methods are compared for inspection of thick nickel base superalloy (Alloy 617) with slots at various depths in the range of 25-200 mm. Optimization of the incidence angle has been performed by beam computation in CIVA software. Results obtained by CIVA simulations are discussed and also compared for the three methods.
{"title":"TOTAL FOCUSING METHOD BASED ULTRASONIC PHASED ARRAY IMAGING IN THICK STRUCTURES","authors":"Sumana, Anish Kumar","doi":"10.1115/1.4050802","DOIUrl":"https://doi.org/10.1115/1.4050802","url":null,"abstract":"\u0000 Ultrasonic non-destructive testing traditionally uses a conventional monolithic transducer. An approach similar to this comprising of independent single transmissions but with reception performed by all the elements in phased array ultrasonics is known as Full Matrix Capture (FMC). The acquired data is processed by Total Focusing Method (TFM). Conventional FMC-TFM has limitations in the inspection at large depth in attenuating materials due to single element transmission. To improve the beam forming process, coherent recombination of the plane wave with specific angles is utilized in transmission and the same aperture is used for the reception in Plane Wave Imaging (PWI). A new methodology called Angle Beam Virtual Source FMC-TFM (ABVSFMC-TFM) is proposed to inspect thick attenuating materials such as nickel base alloys. The ABVSFMC method leads to improved Signal to Noise Ratio (SNR) as compared to the conventional FMC due to increased energy with directivity during transmission using a group of elements and improved divergence as compared to the PWI due to a small virtual source near the sample surface. In the present paper, FMC-TFM, PWI-TFM and ABVSFMC-TFM methods are compared for inspection of thick nickel base superalloy (Alloy 617) with slots at various depths in the range of 25-200 mm. Optimization of the incidence angle has been performed by beam computation in CIVA software. Results obtained by CIVA simulations are discussed and also compared for the three methods.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"27 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2021-04-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79676532","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
O. Mesnil, A. Recoquillay, T. Druet, Valentin Serey, Huu Hoang, A. Imperiale, E. Demaldent
In guided wave structural health monitoring (GW-SHM), a strong need for reliable and fast simulation tools has been expressed throughout the literature to optimize SHM systems or demonstrate performance. Even though guided wave simulations can be conducted with most finite elements software packages, computational and hardware costs are always prohibitive for large simulation campaigns. A novel SHM module has been recently added to the civa software and relies on unassembled high-order finite elements to overcome these limitations. This article focuses on the thorough validation of civa for SHM to identify the limits of the models. After introducing the key elements of the civa SHM solution, a first validation is presented on a stainless steel pipe representative of the oil and gas industry. Second, validation is conducted on a composite panel with and without stiffener representative of some structures in the aerospace industry. Results show a good match between the experimental and simulated datasets, but only if the input parameters are fully determined before the simulations.
{"title":"Experimental Validation of Transient Spectral Finite Element Simulation Tools Dedicated to Guided Wave-Based Structural Health Monitoring","authors":"O. Mesnil, A. Recoquillay, T. Druet, Valentin Serey, Huu Hoang, A. Imperiale, E. Demaldent","doi":"10.1115/1.4050708","DOIUrl":"https://doi.org/10.1115/1.4050708","url":null,"abstract":"\u0000 In guided wave structural health monitoring (GW-SHM), a strong need for reliable and fast simulation tools has been expressed throughout the literature to optimize SHM systems or demonstrate performance. Even though guided wave simulations can be conducted with most finite elements software packages, computational and hardware costs are always prohibitive for large simulation campaigns. A novel SHM module has been recently added to the civa software and relies on unassembled high-order finite elements to overcome these limitations. This article focuses on the thorough validation of civa for SHM to identify the limits of the models. After introducing the key elements of the civa SHM solution, a first validation is presented on a stainless steel pipe representative of the oil and gas industry. Second, validation is conducted on a composite panel with and without stiffener representative of some structures in the aerospace industry. Results show a good match between the experimental and simulated datasets, but only if the input parameters are fully determined before the simulations.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"1 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2021-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89301510","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Bearing remnant operational life can be determined by implementing a data-driven prognostics method. In this work, the bearing run-to-failure data from experimentation on test rig is used to extract time-domain features. The sudden change in time-domain information signifies the fault inception which led to failure stage promptly. The monotonicity metric is utilized to select the optimal feature set that best represents bearing degradation. Principal component analysis (PCA) is used for dimension reduction and fusion, and a unidimensional health indicator (HI) is constructed. Fluctuations of HI are smoothed by fitting it with a Weibull failure rate function (WFRF) and the corresponding parameters are estimated using nonlinear least-squares method. By inverting the model, the predicted time values are calculated, and hence remnant operational life of bearing is evaluated and compared with the actual life from experimental data. The performance assessment metrics utilized are mean absolute percentage error (MAPE), mean-square error (MSE), root-mean-square error (RMSE), and bias. Besides this, an online degradation state classification method using the k-nearest neighbor (KNN) classifier is implemented. The KNN model performance is assessed by constructing receiver operating characteristics (ROC) curve, which indicates the value of area under the curve (AUC) equal to 0.94, representing high accuracy of the KNN. The remaining useful life (RUL) is predicted within 95% confidence limits, and the predicted RUL almost follows the actual one with some fluctuations. The model performance is found promising and can be implemented to evaluate the remaining useful life of bearing.
采用数据驱动的预测方法可以确定轴承剩余使用寿命。在本研究中，利用试验台上的轴承运行到失效数据提取时域特征。时域信息的突然变化标志着故障的开始，从而迅速进入故障阶段。利用单调度度量来选择最优的特征集来代表轴承退化。采用主成分分析(PCA)进行降维融合，构造一维健康指标(HI)。采用威布尔失效率函数对HI的波动进行拟合，并采用非线性最小二乘法估计相应的参数。通过对模型进行反演，计算出预测时间值，从而评估轴承的剩余工作寿命，并与实验数据中的实际寿命进行比较。使用的性能评估指标是平均绝对百分比误差(MAPE)、均方误差(MSE)、均方根误差(RMSE)和偏差。此外，还实现了一种基于k近邻(KNN)分类器的在线退化状态分类方法。通过构建receiver operating characteristic (ROC)曲线来评估KNN模型的性能，曲线下面积(area under The curve, AUC)值为0.94，表明KNN模型具有较高的准确率。预测的剩余使用寿命(RUL)在95%的置信范围内，预测的RUL与实际的RUL基本一致，但有一定的波动。该模型性能良好，可用于评估轴承的剩余使用寿命。
{"title":"Prognostic Analysis of High-Speed Cylindrical Roller Bearing Using Weibull Distribution and k-Nearest Neighbor","authors":"M. Rathore, S. Harsha","doi":"10.1115/1.4051314","DOIUrl":"https://doi.org/10.1115/1.4051314","url":null,"abstract":"Bearing remnant operational life can be determined by implementing a data-driven prognostics method. In this work, the bearing run-to-failure data from experimentation on test rig is used to extract time-domain features. The sudden change in time-domain information signifies the fault inception which led to failure stage promptly. The monotonicity metric is utilized to select the optimal feature set that best represents bearing degradation. Principal component analysis (PCA) is used for dimension reduction and fusion, and a unidimensional health indicator (HI) is constructed. Fluctuations of HI are smoothed by fitting it with a Weibull failure rate function (WFRF) and the corresponding parameters are estimated using nonlinear least-squares method. By inverting the model, the predicted time values are calculated, and hence remnant operational life of bearing is evaluated and compared with the actual life from experimental data. The performance assessment metrics utilized are mean absolute percentage error (MAPE), mean-square error (MSE), root-mean-square error (RMSE), and bias. Besides this, an online degradation state classification method using the k-nearest neighbor (KNN) classifier is implemented. The KNN model performance is assessed by constructing receiver operating characteristics (ROC) curve, which indicates the value of area under the curve (AUC) equal to 0.94, representing high accuracy of the KNN. The remaining useful life (RUL) is predicted within 95% confidence limits, and the predicted RUL almost follows the actual one with some fluctuations. The model performance is found promising and can be implemented to evaluate the remaining useful life of bearing.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"16 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87852622","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Libowen Xu, Qing Wang, I. Ivrissimtzis, Shisong Li
The operation and maintenance costs of wind farms are always high due to high labor costs and the high replacement cost of parts. Thus, it is of great importance to have real-time monitoring and an early fault diagnostic system to prevent major events, reduce time-based maintenance, and minimize the cost. In this paper, such a two-step system for early stage rolling bearing failures in offshore wind turbines is introduced. First, empirical mode decomposition is applied to minimize the effect of ambient noise. Next, correlation coefficients between a reference signal and test signals are obtained and incipient fault detection is achieved by comparing the results with a threshold value. Through further analysis of the envelope spectrum, sample entropy for selected intrinsic mode functions is obtained, which is further used to train a support vector machine classifier to achieve fault classification and degradation state recognition. The proposed diagnostic approach is verified by experimental tests, and an accuracy of 98% in identifying and classifying rolling bearing failures under various loading conditions is obtained.
{"title":"Early Fault Diagnostic System for Rolling Bearing Faults in Wind Turbines","authors":"Libowen Xu, Qing Wang, I. Ivrissimtzis, Shisong Li","doi":"10.1115/1.4051222","DOIUrl":"https://doi.org/10.1115/1.4051222","url":null,"abstract":"The operation and maintenance costs of wind farms are always high due to high labor costs and the high replacement cost of parts. Thus, it is of great importance to have real-time monitoring and an early fault diagnostic system to prevent major events, reduce time-based maintenance, and minimize the cost. In this paper, such a two-step system for early stage rolling bearing failures in offshore wind turbines is introduced. First, empirical mode decomposition is applied to minimize the effect of ambient noise. Next, correlation coefficients between a reference signal and test signals are obtained and incipient fault detection is achieved by comparing the results with a threshold value. Through further analysis of the envelope spectrum, sample entropy for selected intrinsic mode functions is obtained, which is further used to train a support vector machine classifier to achieve fault classification and degradation state recognition. The proposed diagnostic approach is verified by experimental tests, and an accuracy of 98% in identifying and classifying rolling bearing failures under various loading conditions is obtained.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"70 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72497306","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A novel model-based technique is presented in this paper for the estimation of the fault size in different components of rolling contact bearings. A detailed dimensional analysis of the problem is carried out and an experimental methodology using the Box–Behnken design is applied to generate the experimental data set. First, the analysis of the vibration acceleration amplitude at fault frequency, its dependence on the bearing operating, and fault parameters using the obtained vibration data set are carried out by statistical analysis of variance. Numerical equations are developed then using the experimental data set for the correlation of the vibration acceleration amplitude in the frequency domain with the fault sizes based on the developed dimensionless terms. A hybrid backpropagation neural network integrating genetic algorithm is also developed to check the computational performance of the developed model equations. Validation of the proposed method is carried experimentally also for three seeded defect sizes on the outer race, inner race, and rolling element. The maximum model accuracy observed is for the inner race defect case with a predictive accuracy of 99.44% and for the roller defect case, it is 98.77%. The deviance observed for the model predictive performance is maximum for the outer race defect case with the least accuracy of 90.47% amongst all.
{"title":"Vibration Characteristics Diagnosis and Estimation of Fault Sizes in Rolling Contact Bearings: A Model-Based Approach","authors":"I. Jamadar","doi":"10.1115/1.4051176","DOIUrl":"https://doi.org/10.1115/1.4051176","url":null,"abstract":"A novel model-based technique is presented in this paper for the estimation of the fault size in different components of rolling contact bearings. A detailed dimensional analysis of the problem is carried out and an experimental methodology using the Box–Behnken design is applied to generate the experimental data set. First, the analysis of the vibration acceleration amplitude at fault frequency, its dependence on the bearing operating, and fault parameters using the obtained vibration data set are carried out by statistical analysis of variance. Numerical equations are developed then using the experimental data set for the correlation of the vibration acceleration amplitude in the frequency domain with the fault sizes based on the developed dimensionless terms. A hybrid backpropagation neural network integrating genetic algorithm is also developed to check the computational performance of the developed model equations. Validation of the proposed method is carried experimentally also for three seeded defect sizes on the outer race, inner race, and rolling element. The maximum model accuracy observed is for the inner race defect case with a predictive accuracy of 99.44% and for the roller defect case, it is 98.77%. The deviance observed for the model predictive performance is maximum for the outer race defect case with the least accuracy of 90.47% amongst all.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"84 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86905607","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：