A nondestructive evaluation (NDE) technique based on highly nonlinear solitary waves (HNSWs) has been developed recently. The technique uses the propagation and detection of HNSWs along 1-D array of spherical particles in which one end is in contact with the structure to be inspected, and the particle at the opposite end induces the waves by means of a mechanical impact. Several studies have demonstrated that the dynamic interaction between the waves and the structure is dependent on the geometric and mechanical properties of the structure and such dependency can be monitored by sensing the waves reflected at the chain/structure interface. The NDE technique is typically performed using the so-called HNSW transducer, which indicates a portable device that consists of an array of particles, a device to trigger the waves, and a sensing element to detect the waves. In this study, the long-term performance of three transducers was investigated by placing them above a test object while the transducers triggered and detected thousands of waves. Any variability of the waves was quantified by extracting simple features such as amplitude, time of flight, and cross-correlation. Sixteen measurements were also captured with short videos at ~1,000 fps. The results of the study demonstrate that the traveling time of the solitary waves is the most reliable parameter. The findings of this study allow to frame a valid strategy to improve the design of the transducers in order to make the HNSW-based technique suitable for long-term monitoring.
{"title":"On the long-term performance of solitary wave-based transducers for nondestructive evaluation applications","authors":"H. Jalali, P. Rizzo","doi":"10.1115/1.4054391","DOIUrl":"https://doi.org/10.1115/1.4054391","url":null,"abstract":"\u0000 A nondestructive evaluation (NDE) technique based on highly nonlinear solitary waves (HNSWs) has been developed recently. The technique uses the propagation and detection of HNSWs along 1-D array of spherical particles in which one end is in contact with the structure to be inspected, and the particle at the opposite end induces the waves by means of a mechanical impact. Several studies have demonstrated that the dynamic interaction between the waves and the structure is dependent on the geometric and mechanical properties of the structure and such dependency can be monitored by sensing the waves reflected at the chain/structure interface. The NDE technique is typically performed using the so-called HNSW transducer, which indicates a portable device that consists of an array of particles, a device to trigger the waves, and a sensing element to detect the waves. In this study, the long-term performance of three transducers was investigated by placing them above a test object while the transducers triggered and detected thousands of waves. Any variability of the waves was quantified by extracting simple features such as amplitude, time of flight, and cross-correlation. Sixteen measurements were also captured with short videos at ~1,000 fps. The results of the study demonstrate that the traveling time of the solitary waves is the most reliable parameter. The findings of this study allow to frame a valid strategy to improve the design of the transducers in order to make the HNSW-based technique suitable for long-term monitoring.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"50 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2022-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82460900","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}
In the study of the remaining useful life (RUL) prediction of neural networks based on deep learning, most of the RUL prediction models use point estimation models. However, due to the influence of the measurement noise and the parameters in the deep learning model, the prediction results will be quite different, which makes the point prediction meaningless. For this reason, this paper proposes a multi-scale convolutional neural network based on approximate Bayesian inference to realize the credibility measurement of bearing RUL prediction results. First, in order to avoid the problem of insufficient single-scale feature representation, Parallel multiple dilated convolutions are used to extract multiple features. At the same time, the channel attention mechanism is used to allocate its importance, which can avoid the redundancy of multi-dimensional information. Then, MC Dropout can be used to describe the probability characteristics of the results, so as to achieve the measurement of the uncertainty of the RUL prediction results. Finally, the PHM data set is used to verify that the method has less volatility compared with the traditional point estimation prediction results, which provides a more valuable reference for predictive maintenance.
{"title":"Uncertainty Measurement of the Prediction of the Remaining useful Life of Rolling Bearings","authors":"Hongchun Sun, Chenchen Wu, Lei Zunyang","doi":"10.1115/1.4054392","DOIUrl":"https://doi.org/10.1115/1.4054392","url":null,"abstract":"\u0000 In the study of the remaining useful life (RUL) prediction of neural networks based on deep learning, most of the RUL prediction models use point estimation models. However, due to the influence of the measurement noise and the parameters in the deep learning model, the prediction results will be quite different, which makes the point prediction meaningless. For this reason, this paper proposes a multi-scale convolutional neural network based on approximate Bayesian inference to realize the credibility measurement of bearing RUL prediction results. First, in order to avoid the problem of insufficient single-scale feature representation, Parallel multiple dilated convolutions are used to extract multiple features. At the same time, the channel attention mechanism is used to allocate its importance, which can avoid the redundancy of multi-dimensional information. Then, MC Dropout can be used to describe the probability characteristics of the results, so as to achieve the measurement of the uncertainty of the RUL prediction results. Finally, the PHM data set is used to verify that the method has less volatility compared with the traditional point estimation prediction results, which provides a more valuable reference for predictive maintenance.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"20 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2022-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87861862","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}
M. Ross, David A. Najera-Flores, James E. Freymiller
This work provides a novel method for including mean stress in the ASME BPVC elastic-plastic fatigue procedure. Typically, the mean stress is accounted for by adjusting the stress-life cycle curve. A new approach is provided by adjusting the effective equivalent stress range with a mean stress correction. This approach is advantageous because this adjustment can be made at each cycle and efficiently implemented in the computation. This new approach allows for fatigue analysis of large-scale models, such as full spacecraft assemblies or a semi-truck trailer. The proposed method is verified by predicting the fatigue life of a test coupon and compared to experimental results.
{"title":"Including Mean Stress in the Effective Equivalent Stress of the ASME BPVC Elastic-plastic Fatigue Method","authors":"M. Ross, David A. Najera-Flores, James E. Freymiller","doi":"10.1115/1.4054357","DOIUrl":"https://doi.org/10.1115/1.4054357","url":null,"abstract":"\u0000 This work provides a novel method for including mean stress in the ASME BPVC elastic-plastic fatigue procedure. Typically, the mean stress is accounted for by adjusting the stress-life cycle curve. A new approach is provided by adjusting the effective equivalent stress range with a mean stress correction. This approach is advantageous because this adjustment can be made at each cycle and efficiently implemented in the computation. This new approach allows for fatigue analysis of large-scale models, such as full spacecraft assemblies or a semi-truck trailer. The proposed method is verified by predicting the fatigue life of a test coupon and compared to experimental results.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"21 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2022-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73752275","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}
Determination of the present condition of Grade 91 steels commonly used in fossil plants presents an ongoing challenge for the industry. Current individual methods of nondestructive evaluations (NDE) are challenged to reliably determine steel microstructures and are impractical for deployment over large areas such as piping systems. This work investigates the discriminatory potential of an integral assessment combining multiple NDE techniques: magnetic methods (including coercive field, incremental permeability and Barkhausen noise), thermoelectric power, and ultrasonic methods (including attenuation, backscatter and absorption). Using a sample inventory of manufactured representative microstructure conditions in thin-walled tubes, 5 of the 8 microstructure conditions could be identified by data fusion. Combined magnetic methods appeared as a primary technique, as this identify 4 out of the 8 conditions, while the ultrasonic methods can be seen as complimentary tests to increase confidence in classification and identify a fifth condition.
{"title":"Nondestructive Microstructure Characterization of Tempered Martensitic Steels through Data Fusion","authors":"M. Mawhin, T. Seuaciuc-Osório","doi":"10.1115/1.4054230","DOIUrl":"https://doi.org/10.1115/1.4054230","url":null,"abstract":"\u0000 Determination of the present condition of Grade 91 steels commonly used in fossil plants presents an ongoing challenge for the industry. Current individual methods of nondestructive evaluations (NDE) are challenged to reliably determine steel microstructures and are impractical for deployment over large areas such as piping systems. This work investigates the discriminatory potential of an integral assessment combining multiple NDE techniques: magnetic methods (including coercive field, incremental permeability and Barkhausen noise), thermoelectric power, and ultrasonic methods (including attenuation, backscatter and absorption). Using a sample inventory of manufactured representative microstructure conditions in thin-walled tubes, 5 of the 8 microstructure conditions could be identified by data fusion. Combined magnetic methods appeared as a primary technique, as this identify 4 out of the 8 conditions, while the ultrasonic methods can be seen as complimentary tests to increase confidence in classification and identify a fifth condition.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"31 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2022-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81731938","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}
V. Memmolo, E. Monaco, F. Ricci, Carmine Vaselli, N. Cimminiello, Pasquale Salvato
Electrical actuation concept fits perfectly with the worldwide strategy of more electric aircraft to reduce carbon footprint. However, the integration of linear electromechanical actuators is promising yet challenging in safety critical systems due to the jamming of the driven load. That fault is a critical mechanical transmission failure giving rise to concern in many applications such as primary flight controls or landing gears extension and steering. This paper critically reviews electric actuation solutions currently available for aerospace application, the limits for their upcoming deployment and the different solutions to achieve an on-condition maintenance to reduce any safety risk during lifetime. A particular attention is given to the jamming and possible strategies to avoid any hazard induced by this failure. The most promising approach relies on the establishment of a predictive maintenance by monitoring continuously the actuator to timely warn any structural alteration which is prone to induce jamming. Finally, a case study is presented, showing the ability of engine current monitoring to warn any alteration in the load maneuvering. The current spectrum analysis is found to be sensitive to both mechanical fault and control failure or malfunctioning. On top of that, no addition sensors are needed, making this technique affordable and cost-effective.
{"title":"Structural Health Monitoring of Electromechanical Actuators in Aviation - Challenges Ahead and Case Study","authors":"V. Memmolo, E. Monaco, F. Ricci, Carmine Vaselli, N. Cimminiello, Pasquale Salvato","doi":"10.1115/1.4054231","DOIUrl":"https://doi.org/10.1115/1.4054231","url":null,"abstract":"\u0000 Electrical actuation concept fits perfectly with the worldwide strategy of more electric aircraft to reduce carbon footprint. However, the integration of linear electromechanical actuators is promising yet challenging in safety critical systems due to the jamming of the driven load. That fault is a critical mechanical transmission failure giving rise to concern in many applications such as primary flight controls or landing gears extension and steering. This paper critically reviews electric actuation solutions currently available for aerospace application, the limits for their upcoming deployment and the different solutions to achieve an on-condition maintenance to reduce any safety risk during lifetime. A particular attention is given to the jamming and possible strategies to avoid any hazard induced by this failure. The most promising approach relies on the establishment of a predictive maintenance by monitoring continuously the actuator to timely warn any structural alteration which is prone to induce jamming. Finally, a case study is presented, showing the ability of engine current monitoring to warn any alteration in the load maneuvering. The current spectrum analysis is found to be sensitive to both mechanical fault and control failure or malfunctioning. On top of that, no addition sensors are needed, making this technique affordable and cost-effective.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"6 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2022-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78977764","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 guided wave based method for localization of breathing delamination is presented in this investigation. The proposed technique utilizes one-way mixing of a dual-frequency fundamental antisymmetric Lamb modes with judiciously selected central frequencies. The dual frequency interrogation signal, upon interacting with a breathing delamination, leads to additional frequency sidebands in the frequency response spectrum; strength of which is quantified in terms of the Combination Tone index. The numerical predictions of these sidebands are validated using an in-house experimentation. It is further exposited that the Combination Tone index depends strongly on the extent of the temporal overlap the two constituent wave-envelopes have as they propagate through the breathing delamination. Accordingly, for a synchronous passage (with 100% temporal overlap) the Combination Tone index is maximum while it reduces with the decreasing temporal overlap. By utilizing the dispersive nature of the chosen Lamb mode, a relation is then developed correlating the temporal separation of the wave-envelopes at the location of the actuator, the group speeds, and the distance between the actuator and the delamination. Based on these inferences, a technique for localizing a breathing delamination is proposed, which involves interrogating the component by systematically altering the temporal overlap in the input waveform and monitoring the Combination Tone index for its maxima. The efficacy of the localization technique (close to 90%) is demonstrated through an illustrative case analyzed numerically as well as experimentally.
{"title":"Localization of a Breathing Delamination using Nonlinear Lamb Wave Mixing","authors":"Yamnesh Agrawal, A. Gangwar, D. Joglekar","doi":"10.1115/1.4054100","DOIUrl":"https://doi.org/10.1115/1.4054100","url":null,"abstract":"\u0000 A guided wave based method for localization of breathing delamination is presented in this investigation. The proposed technique utilizes one-way mixing of a dual-frequency fundamental antisymmetric Lamb modes with judiciously selected central frequencies. The dual frequency interrogation signal, upon interacting with a breathing delamination, leads to additional frequency sidebands in the frequency response spectrum; strength of which is quantified in terms of the Combination Tone index. The numerical predictions of these sidebands are validated using an in-house experimentation. It is further exposited that the Combination Tone index depends strongly on the extent of the temporal overlap the two constituent wave-envelopes have as they propagate through the breathing delamination. Accordingly, for a synchronous passage (with 100% temporal overlap) the Combination Tone index is maximum while it reduces with the decreasing temporal overlap. By utilizing the dispersive nature of the chosen Lamb mode, a relation is then developed correlating the temporal separation of the wave-envelopes at the location of the actuator, the group speeds, and the distance between the actuator and the delamination. Based on these inferences, a technique for localizing a breathing delamination is proposed, which involves interrogating the component by systematically altering the temporal overlap in the input waveform and monitoring the Combination Tone index for its maxima. The efficacy of the localization technique (close to 90%) is demonstrated through an illustrative case analyzed numerically as well as experimentally.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"16 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2022-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85160619","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}
Guided wave mixing leverages mutual wave interactions to provide sensitive diagnostics of material degradation in plates and pipes and an early warning upon which maintenance decisions can be based. In some cases, the material to be interrogated may be otherwise inaccessible for nondestructive evaluation. The distortion of the waveform in nonlinear ultrasonics is typically quite small, often making it difficult to distinguish from nonlinearities in the sensing system. Mutual wave interactions are preferred to wave self-interactions in this respect because they can be designed to occur away from frequencies corrupted by sensing system nonlinearity. Furthermore, primary waves that generate secondary waves having a different polarity also provide a means to separate the material nonlinearity from the sensing system nonlinearity. Finite element simulations of wave mixing using a hyperelastic material model are conducted as a precursor to laboratory experiments in order to establish realistic expectations. In one case shear-horizontal waves are mixed with co-directional symmetric Lamb waves to generate back-propagating shear-horizontal waves at the difference frequency. In the second case counter-propagating shear-horizontal waves mix to generate secondary standing waves at the cutoff frequency of the S1 Lamb wave mode. In both cases the results indicate that the larger the wave mixing zone, the more measurable is the amplitude of the secondary waves. These results will be used to design experiments that demonstrate the utility of these novel wave interactions.
{"title":"Mutual interaction of guided waves having mixed polarity for early detection of material degradation","authors":"C. Lissenden, Anurup Guha, Mostafa Hasanian","doi":"10.1115/1.4053959","DOIUrl":"https://doi.org/10.1115/1.4053959","url":null,"abstract":"\u0000 Guided wave mixing leverages mutual wave interactions to provide sensitive diagnostics of material degradation in plates and pipes and an early warning upon which maintenance decisions can be based. In some cases, the material to be interrogated may be otherwise inaccessible for nondestructive evaluation. The distortion of the waveform in nonlinear ultrasonics is typically quite small, often making it difficult to distinguish from nonlinearities in the sensing system. Mutual wave interactions are preferred to wave self-interactions in this respect because they can be designed to occur away from frequencies corrupted by sensing system nonlinearity. Furthermore, primary waves that generate secondary waves having a different polarity also provide a means to separate the material nonlinearity from the sensing system nonlinearity. Finite element simulations of wave mixing using a hyperelastic material model are conducted as a precursor to laboratory experiments in order to establish realistic expectations. In one case shear-horizontal waves are mixed with co-directional symmetric Lamb waves to generate back-propagating shear-horizontal waves at the difference frequency. In the second case counter-propagating shear-horizontal waves mix to generate secondary standing waves at the cutoff frequency of the S1 Lamb wave mode. In both cases the results indicate that the larger the wave mixing zone, the more measurable is the amplitude of the secondary waves. These results will be used to design experiments that demonstrate the utility of these novel wave interactions.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"17 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2022-02-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84694419","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}
Rolling element bearings are one of the most common mechanical components used in a wide variety of rotating systems. The performance of these systems is closely associated with the health of bearings. In this study, a nonlinear time series analysis method i.e. recurrence analysis is utilized to assess the health of bearings using time domain data. The recurrence analysis acquires the quantitative measures from the recurrence plots and provides an insight to the system under investigations. Experiments are performed to generate the vibration data from the healthy and faulty bearing. Eight recurrence quantitative analysis measures and five time-domain measures are used for the investigations. Three artificial intelligence techniques: rotation forest, artificial neural network and support vector machine are employed to quantify the diagnosis performance. Results highlight the ability of recurrence analysis to identify the health state of the bearing at the early stage and superior diagnosis accuracy of the proposed methodology.
{"title":"Fault diagnosis of bearings using recurrences and AI techniques","authors":"Aditya Sharma","doi":"10.1115/1.4053773","DOIUrl":"https://doi.org/10.1115/1.4053773","url":null,"abstract":"\u0000 Rolling element bearings are one of the most common mechanical components used in a wide variety of rotating systems. The performance of these systems is closely associated with the health of bearings. In this study, a nonlinear time series analysis method i.e. recurrence analysis is utilized to assess the health of bearings using time domain data. The recurrence analysis acquires the quantitative measures from the recurrence plots and provides an insight to the system under investigations. Experiments are performed to generate the vibration data from the healthy and faulty bearing. Eight recurrence quantitative analysis measures and five time-domain measures are used for the investigations. Three artificial intelligence techniques: rotation forest, artificial neural network and support vector machine are employed to quantify the diagnosis performance. Results highlight the ability of recurrence analysis to identify the health state of the bearing at the early stage and superior diagnosis accuracy of the proposed methodology.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"52 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2022-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74755234","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}
Zhaoyun Ma, Lingyu Yu, Y. Chao, P. Lam, R. Sindelar, A. Duncan, Thanh-Tam Truong, C. Verst, Peipei Sun, Andrew Campbell
Stress corrosion cracking (SCC) had occurred in early generation high level nuclear waste tanks constructed by welding carbon steel. This paper describes an ultrasonic inspection system and its fundamental ability to detect and quantify the length of SCC on thick welded steel plates. Finite element method (FEM) was applied to simulate the welding process to estimate the welding residual stress field. Growth of stress corrosion cracks are driven by crack stress intensities exceeding the subcritical cracking threshold intensity. The subject plate was experimentally inspected with ultrasonic nondestructive evaluation (NDE) techniques to characterize the extent of SCC. The NDE system employs a piezoelectric transducer to generate guided waves in the thick steel plate, and a scanning laser Doppler vibrometer (SLDV) to measure multidimensional time-space wavefield data over a user-defined scanning area in the plate surface. The measured wavefield data can show wave interactions in a localized area in the plate due to the presence of the discontinuities of the SCC. To generate an inspection image that can precisely show the crack's location and/or the dimension, the wavefield data are further processed to generate inspection image that maps the entire sample plate so the crack can be clearly identified in the plate while its length can be readily estimated. The ultrasonic test results for crack length agree well with the visually estimated length, and are close to that predicted by FEM for cracks in the weld residual stress field.
{"title":"Nondestructive evaluation of stress corrosion cracking in a welded steel plate using guided ultrasonic waves","authors":"Zhaoyun Ma, Lingyu Yu, Y. Chao, P. Lam, R. Sindelar, A. Duncan, Thanh-Tam Truong, C. Verst, Peipei Sun, Andrew Campbell","doi":"10.1115/1.4053653","DOIUrl":"https://doi.org/10.1115/1.4053653","url":null,"abstract":"\u0000 Stress corrosion cracking (SCC) had occurred in early generation high level nuclear waste tanks constructed by welding carbon steel. This paper describes an ultrasonic inspection system and its fundamental ability to detect and quantify the length of SCC on thick welded steel plates. Finite element method (FEM) was applied to simulate the welding process to estimate the welding residual stress field. Growth of stress corrosion cracks are driven by crack stress intensities exceeding the subcritical cracking threshold intensity. The subject plate was experimentally inspected with ultrasonic nondestructive evaluation (NDE) techniques to characterize the extent of SCC. The NDE system employs a piezoelectric transducer to generate guided waves in the thick steel plate, and a scanning laser Doppler vibrometer (SLDV) to measure multidimensional time-space wavefield data over a user-defined scanning area in the plate surface. The measured wavefield data can show wave interactions in a localized area in the plate due to the presence of the discontinuities of the SCC. To generate an inspection image that can precisely show the crack's location and/or the dimension, the wavefield data are further processed to generate inspection image that maps the entire sample plate so the crack can be clearly identified in the plate while its length can be readily estimated. The ultrasonic test results for crack length agree well with the visually estimated length, and are close to that predicted by FEM for cracks in the weld residual stress field.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"112 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2022-01-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85867988","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 system health index is a measurement of the health condition of complex systems. However, most of the health indices are developed based on strong assumptions. Consequently, existing health indices are not capable of measuring the actual deterioration behaviors with high accuracy. To address this issue, we introduce a probabilistic graphical model to examine the probabilistic relationships among sensor signals, remaining useful life (RUL), and health indices. Based on the graphical model, three types of conditional probabilistic autoencoders are combined to develop the health indices of a complex aero-propulsion system. The proposed method is demonstrated on an engine dataset. The experimental results have shown that the proposed method is capable of constructing robust health indices as well as improving the accuracy of RUL prediction.
{"title":"Constructing Robust and Reliable Health Indices and Improving the Accuracy of Remaining Useful Life Prediction","authors":"Yupeng Wei, Dazhong Wu, J. Terpenny","doi":"10.1115/1.4053620","DOIUrl":"https://doi.org/10.1115/1.4053620","url":null,"abstract":"\u0000 A system health index is a measurement of the health condition of complex systems. However, most of the health indices are developed based on strong assumptions. Consequently, existing health indices are not capable of measuring the actual deterioration behaviors with high accuracy. To address this issue, we introduce a probabilistic graphical model to examine the probabilistic relationships among sensor signals, remaining useful life (RUL), and health indices. Based on the graphical model, three types of conditional probabilistic autoencoders are combined to develop the health indices of a complex aero-propulsion system. The proposed method is demonstrated on an engine dataset. The experimental results have shown that the proposed method is capable of constructing robust health indices as well as improving the accuracy of RUL prediction.","PeriodicalId":52294,"journal":{"name":"Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems","volume":"74 1","pages":""},"PeriodicalIF":1.1,"publicationDate":"2022-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87020911","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：